Closed Bug 62429 Opened 24 years ago Closed 21 years ago

The "Start my reply above the quoted text" setting should ALSO allow to prepend the signature above the quote text. (top)


(MailNews Core :: Composition, enhancement, P3)



(Not tracked)



(Reporter: cppunish, Assigned: iannbugzilla)


(Blocks 1 open bug)


(Keywords: helpwanted, Whiteboard: PLEASE DO NOT SPAM THIS BUG. See comment 283 for specs.)


(5 files, 13 obsolete files)

4.57 KB, image/png
1.91 KB, image/png
76.88 KB, image/jpeg
73.08 KB, image/jpeg
21.84 KB, patch
: review+
: superreview+
Details | Diff | Splinter Review
It would be great if when you select "start my reply above the quoted text" in
the Preferences -> Message Composition window that not only would your reply
start above the quoted text, but also the signature file that you include.

For example, right now, if I select "start my reply above the quoted text" and
hit reply to any message, my signature file will be appended to the bottom of
the message, below the quoted text. I would love it to be appended above the
quoted text just like where my message text starts.

I dont know if I am being clear about this, but I will try to elaborate if
someone doesn't understand.
Marking NEW so someone will look at it. Seems like a good idea to me.
Ever confirmed: true
Summary: Enhancement about composing a message and where test starts. → [RFE] Replying to a message w/a Signature
change qa contact to myself
QA Contact: esther → sheelar
*** Bug 65057 has been marked as a duplicate of this bug. ***
*** Bug 65970 has been marked as a duplicate of this bug. ***
Jason, I understand what you're saying: if the pref "start my reply above the
quoted text" is chosen then we should also append the signature above the quoted

QA's: does this sound like an acceptable behaviour?
URL: na
Keywords: mozilla1.1
OS: Windows 98 → All
Summary: [RFE] Replying to a message w/a Signature → Replying with the "start my reply above the quoted text" pref on should append the signature above the quote text.
May I explain why it is so important?. Assume you're a member of a mailing list.
Every time you send replies (Reply All) to the list, you want everyone to read
your signature at once, not to have to scroll through a (sometimes very long)
uncommented quoted text. Isn't this is the default IE behaviour ? (afaik).
Current workaround at Nav 4.7 is to cut & paste the signature text above quoted
On the other hand, sometimes you need to comment the quoted text, so your
signature must lie below it. So a new pref "Place signature above, below quoted
text" is needed here. 
in general we'd prefer for reply above to die a quick death. however that's 
only a wish.  reporter is correct, we should prepend the signiature when 
replying above.
Summary: Replying with the "start my reply above the quoted text" pref on should append the signature above the quote text. → Replying with the "start my reply above the quoted text" pref on should prepend the signature above the quote text.
It would be annoying to have to remove the signature every time you reply to
someone's email in a thread/newsgroup.

I don't think this should NOT be done; wontfix.
Closed: 23 years ago
Resolution: --- → WONTFIX
No one objected in over one and a half month, I will now resolve this as a
wontfix because it would be a pain to have everybody and his uncle's signature
above all the other text in newsgroups.

Besides, no other email client does this. 
Outlook Express for Windows appends the signature wherever your reply is 
It's annoying anyway.
> sides, no other email client does this

Outlook Express 5.0 for Win32 does this. Btw, not all people in the world 
read newsgroups every day. So, why not making "put signature below quoted text" 
as default and leave us an option to "put signature above quoted text" ? A lot 
of people do not use inline replying. Why to force us scroll the whole quoted 
text in order to see the signature? Sorry for repeating myself but it seems 
either my previous comment was not well understood or that bug is 
wontfixed because of the number of votes.
Please, I would really appreciate an option to have signatures start above the 
quoted text. I dont think it would really be that hard of an option to add, and 
there is more than enough room in the options pane, so I dont know.

I will understand if its something not worth implementing, but I really would 
appreciate it, because with the hundreds of emails I reply to every day, it gets 
kind of annoying to copy and paste my signature every time...

Such a pref is not worth implementing, not least because very few people would 
understand it.
Since you find that worthless, let "reply above to die a quick death". There's 
not much sense in replying above when the signature is buried many lines below. 
Another solution would be to implement that pref without having a UI. This would 
be better than nothing. Otherwise we will be forced to keep doing this joyful 
cut & paste job every day.
I finally stop nagging here. It's your choice to realise if user 
friendliness is another reason (other than their OS monopoly) why MS currently 
dominates the  browser market.
Putting the signature above the quoted reply is just dreadful. Replying on top
is bad enough as it doesn't encourage people to prune the quoted text, but
having a signature coming between the text and the reply is even worse.
Particularly if the sig file is one of these awful long disclaimer files.

Just because Outlook Express encourages bad practices doesn't mean mozilla should.
 David, initially I didn't want to reply you through Bugzilla because I would
become the biggest spammer on earth (after all those my previous postings here).
Then I changed my mind. This is my last attempt, honestly, and I apologise for
the annoyance.

The problem with the signature below the quoted text comes from the writing
habits of many (most, imho) people. All that you said are correct *if* you use
to comment messages *line by line* (I mean you mix quoted text with your reply).
Most people don't want to do that. Usually they just put their replies above to
ensure that the recipient will read them first. They leave the quoted text lying
below just for the rare case one wants to verify exactly what is replied. From
this point view, subject is what matters, quoted text is a "luxury" component
that it must not interfere with the rest of the document. You can call it bad
practice but imho it isn't.
Another annoying problem is the long spamming ads added by all "free" mailing
list servers. They usually add this to the end of message.
Please also take into account what is the usual practice in more formal
messages. In-line commenting such a message (thus distorting the original text)
is considered inappropriate. I suspect this was the reason why Microsoft put the
signature above.
I think generalizing a specific way of messaging (technical newsgroups, internal
corporate group communication) to the whole extent of mail usage is wrong and
the number of duplicates/votes may prove it when Mozilla becomes more acceptable
to the public.
*** Bug 86101 has been marked as a duplicate of this bug. ***
*** Bug 87317 has been marked as a duplicate of this bug. ***
I too would very much like to have this ability because....

...if someone (like me) *always* replies on the TOP of a quoted message (easier
to read newest message), why should ones signature be *completely separated*
from my message by the (potentially very long) quoted text?

It makes sense to have the signature *directly* below the message - how else can
you really tell which signature belongs to which message? Anything else is a
mess and non-intuitive. At *least* make it a preference setting.

Please reopen this bug.
I'm in a weird mood. it's clear i want reply above to die. But while we have it 
do I want it to be vaguely useful?

From: Blah
To: Blah
Subject: Black Sheep
Have you any wool
No wrote:
> Sir
> No wrote:
> > Sir
> > Signed,
> > I don't remember the rest.
> Signed,
> I don't remember

Clearly you can follow who wrote what. and clearly no one cares. (And clearly 
no one will understand the reference, but I don't care about that either.)  
This is because the from line or the quote line allows you to match text, and 
the signatures match w/ the indentation of a body of text.

In an effort to teach someone a wasted English lesson,
hwaara wrote:
> I don't think this should NOT be done; wontfix.
> Insincerely,
> A foreign student of the English Language
which equals He does not think this should not be done. So he thinks this 
should be done. [reopen]

I don't think that we should be wontfixing stuff like this.
If Assigned To: has no interest in fixing this, it should 
be assigned to . If the module owner comes in and says that 
this will never be accepted into's cvs then we'll probably have a 
problem w/ the module owner, but for that time this bug can be resolved as 
wontfix _by the module owner_ (who is that?).

David verified that:
> Just because Outlook Express encourages bad practices doesn't mean mozilla 

Which is true, but it isn't a valid reason to verify a bug either.

If we want to discourage bad practices we should consider:
1. removing reply above entirely.
2. adding a wizard mode for replies. -- It would have to intelligently discern 
paragraphs and help the user reply to them individually.  Then it would have to 
offer to remove unused quoted paragraphs.  At first, we could make all of the 
paragraphs we don't like blink and stuff, and before sending we could nag the 
user about them... Oh how annoying we could be.
3. Add some way to alert users of violations of good email/news quoting 

2+3 are probably not worth anyone's time.  0 (this bug -- and how annoying we 
are...) might be worth someone's time (keyword: helpwanted) since a few people 
asked for it and this other userbase uses it. 1 is um... i dunno.
Keywords: helpwanted
Hardware: PC → All
Resolution: WONTFIX → ---
Just remember in the name of netiquette that having a complete reply (including
sig) above the quoted text encourages the horrible behaviour of not snipping out
irrelevant portions of the aforementioned quote.
Please remember that a significant portion of browser/mail/news program users
are not computer g..ks (like me) but business persons doing business
correspondence (also like me). Business persons frequently prefer to top post
and to keep the entire correspondence trailing below the relevant & current
message. This is also for the famous "cover your ass" reason, but mostly to keep
ONE record where the entire communique is stored without forcing the reader to
page through a potentially VERY long message (some correspondences can be quite
lengthy (e.g., a report text).

I therefore think it is quite advisable to allow top posting and to allow the
signature to be placed directly below one's post.

An info box informing of good & poor etiquette when the user selects the pref to
top-reply is a fine idea. But we should definetely not exclude a feature that
would surely alienate MANY business users.

OT: Standards are good, homogeneity is not - let's find the right balance :)
*** Bug 94050 has been marked as a duplicate of this bug. ***
Ok I dont know if there has been any progress on this, but I took a poll of a 
lot of the people I work with, and who have had the chance to work with Mozilla. 
Almost all of them said that for years they had been using Communicator (its a 
requirement where we work) and had just gotten used to copying and pasting their 
sigs to the top of the message.

I know its not good practice, and for a long time I too pruned quoted texts, and 
so on, and still do with my university collegues because its pretty standard 
here. BUT, for the business world, people just get too confused by it, and would 
rather have the huge chunk of text showing them exactly what was said in the 
previous email. So I dont know, i think its a worthwhile addition, even making 
it the default would make sense to me considering that you want Mozilla to be 
used more by the masses rather than just the geeks ;-) But thats just me ... so 
I dont know, just wanted to put my input in again.

As someone who uses Mozilla as their browser and email client to conduct a
business, I use reply-above to keep a transcript of the entire conversation
between myself and the client.

The option exists for reply-above, so I can't understand why you would want your
sig at the bottom?  Surely it is a simple code change to add the signature at
the top, and I don't understand the reluctance to do it.

When I am replying to 10s or 100s of emails a day, it is a hassle to copy and
paste the signature for every message.

The option to remove reply-above would force me to switch email programs.

Scott, which email client let you put your signature somewhere else than at the
Jean-Francois: Outlook Express 5 (build 5.00.2314.1300). As for the business
usage Scott referred to, the following people have already mentioned it before :
me, Peter Lairo and Jason. Let's hope somebody at will understand
that Mozilla is not intended for recreational purposes only.
Well I dont think they intended it for recreational, but apparently it is the
"proper" way of doing an email. Probably started a lot within the academic
community. Anyways, I just wish they would put it in because I know that lots of
people would use it. The business world just does things differently than the
academic community. There is no effeciency with how business does these things,
its just because it works out easier for us. Anyways, thanx for keeping this bug
open, I really hope someone attends to it soon.
Target Milestone: --- → Future
Why should we have to put up with badly formatted emails just because Microsoft
does so?

Before outlook came along then I'm sure most corporate users didn't see the need
for this, and when you consider the average length of corporate signatures then
they really do need to be at the bottom otherwise you have to do a lot of
scrolling to get to the original message.

If this is so vital to the business community I'd rather that they suffer the
consequences of Outlook viruses we should not encourage bad habits.

Look at is this way:
Mozilla uses the proper signature separator (dash dash space) to separate the
sig from the message (outlook doesn't). Mozilla then will delete the sig when
hitting reply. I like this feature, it helps keep the message size shorter when
replying and makes long threads on newsgroups easier to use. However if you put
the sig before the quoted text then this feature automatically breaks.

This is just plain wrong, I'd rather mozilla encourage good habits among its
adopters rather than bend over backwards to fit the needs of a few.

I don't understand your reasoning?

- we don't do anything because microsoft does so, we do it because *users* want it
- People have been copying their signatures from the bottom of their emails
since Netscape 4.x days.  See other comments
- I'd rather use Netscape, but it seems that stubborness is pushing (or pushed)
people to Outlook.  Would you rather that *everyone* went to outlook?
- I'm not sure where is says that "--" is to keep sigs away from messages.  I
can't find anything on the email rfc's.  Is this a Netscape thing?

My last point is that I thought the idea was to have more people migrate to
Mozilla?  Less evil Microsoft?  Less code red etc?  If you keep aruguing that
the masses should leave, then you'll end up with a very few people using Mozilla.

Argh - end of rant.  I just want this implemented (as a pref even).

This is an important feature for business users, it really is (unfortunately).
It should be an *option* - and therefore no one has cause to be alarmed. If
there is one thing that upsets me about noble projects such as this one, it is
arrogance and ignorance towards real-world user's wishes.
I don't think anyone has explained to us *why* companies consider this so
important, and I bet before MS Outlook became popular many people didn't see the
need for this. It is awful (particularly as many corporations use sigs over 4

Personally I think this solution is the best - fix bug 70478 because most of the
time that people reply on top then the original message is probably irrelavant
particularly if they have to scroll down a large signature file just to read it.

The message writer can then press the "quote" button to place the original text
before or after the signature as required. This would reduce message sizes and
because the author only pastes the original message if they feel it necessary
and can put the text wherever they want.

To counter this, I know a lot of people who don't bother to create a sig in
outlook because this stupid behaviour is the default. OK, not a lot, but all the
people I've asked tend to agree.
For David - THE WHY:
1. because it's convenient -> your entire message os in *one* location (incl. sig)
2. because it's intuitive -> messages are sequentially listed one after the
other - no "nested" messages.
3. because it maintains document integrity -> each message is kept the way the
author sent it (incl. sig).

If this feature is an option and not the default, then there is NO sufficient
reason to keep opposing it.

PS. I prefer not to use it myself, but I understand that many would be confused
and put off if it's missing. I DO use it at my office (Lotus Bloates).
Putting putterman and jennifer on the list for this bug. Since this seems to be
over flowing with comments and suggestions. Needs to be determined if this
should or can be resolved according to the requests made here. 
I am strongly opposed to see us fixing this. Just because other email readers
have this bug (which has turned into a feature for some users) doesn't mean we
should have it.   People using this feature in other email readers makes it hard
to read a newsgroup thread properly, without always being disturbed by their
signatures in the middle of the discussion.   Please don't fix this.
HAKAN: misrepresenting the facts ("...some users") is highly counterproductive!
The best solution is to have separate prefs for mail and news.
Ok here is the deal as far as I see it. We have the purists
(newsgroups/university/old school users) vs the business world. I myself fall
into both categories and can completely understand both perspectives on the
situation. Let me just reiterate some of the things others have said...

Purists: When quoting entire msgs + sigs above the message, it complicates
newsgroups, its considered to be bad style when emailing (in the education
community at least), and is something people are used to doing.

Business: It allows for a chronological succession with emails (customer
support, inquiries, etc), allows readers to easily identify the owners of a
message within a multimessage forward, and most importantly allows readers to
keep the entire history of an emailing in one message without having to decipher
who sent what to who and when.

I dont see either way of using it as a bug, they both address different
problems, and inherently I would assume that would be a case where an option
would be neccesary. I also would have no problem having the default be the way
it is so that only the people who really want to use it can change it to what
they want, leaving purists with those who like it that way and anyone who doesnt
really need to change it. I would have promoted a pref change that could be
changed through a text edit of a pref file for the business style, but that
inherently would be useless because those who would want that option available
arent usually technically ready to be editing pref files. Now sure I can (and
most probably will) continue to use Mozilla whether or not this option is
implemented, but it would be an amazing annoyance out of my hair. After copying
and pasting my sig up above the quoted message literally tens of thousands of
times, I would love to not have to do that. There are times I forget to either
move or delete the sig, leaving me with a sig on a message I dont want one to
have, or a sig in a place that most of my collegues, customers, and exployees
would not know where to look.

Either way, this topic should not have to degrade to making or not making this
option available because of anything to do with Outlook or MS in general. Sure,
I hate MS, but that doesnt mean the way they do things is inherently wrong.
Anyways, sorry this was so long, but I felt I had to put my final word in on
this subject before decisions were made.
If this bug gets as far as a patch, someone CC me so I can argue vehemently
against it getting checked in.

We should _not_ encourage bad habits among email users - this includes
ridiculously long. The history of an email conversation is the mails that made
it up, not the current mail. That's like saying when you write to me, I should
send a copy of your letter back to you so you know what I'm replying to.

Ok Gerv.  What you've just criticized is a matter of preference, and it's your 
preference, not mine, and not many of the people on this list.

Go wage war against something a little more important okay?  I've been waiting 
for this functionality for a long time and business users need it.
David Hallowell braught this topic to .mail-news:
<news://>. I reply there. For the
record, I vote for wontfix.
Ducarroz, I suggest you read the thread in npm.mail-news -- you'll notice that
most people tend to recommend this as "wontfix"...
Hey come on guys, is it so tricky or difficult to implement this and we need 10
months to discuss this (this bug is open in last December)? If not, just make it
an "Preference" (with UI). We want a product that attract people to use, don't we?

I can't see any "right" or "wrong" here. I just consider this as a user
"Preference". So please don't force the others to follow our own preference.

P.S. I personally do not need this preference but I do hope Mozilla is more
user-friendly to adopt different people usage pattern.
rúbbish, please read the newsgroup thread and reply there. Thanks.
Attached image My view on the subject
fwiw, gerv has an interesting point. this bug is 13 pages long, has 45 comments,
17 ccs and 5 votes.

supposing we actually played this bug out in written correspondence and always
included a copy of what was written before.  that's say 25[total recipients as a
corteousy] x 13 [pages] x 45 [letters] {46/2} (arbitrary estimate for average
message length) ~~ 336375 pages.  Wow i'm glad we don't do that, the trees
bugzilla would be responsible for killing would be more than the trees mozilla
developers have ever torched or setup or changed in
That's a silly comparisson. How many times do we have to aknowledge that the
users who want this are generally referring to MAIL (about 2-5 replies long, on
average) and not newsgroups (where there is a lot of back-and-forth
communication). We will get nowhere if we can't discuss this without twisting
and ignoring the facts.
I am definitely referring to email messages only and hardly ever use the
newsgroups any more.  However I do *not* think it is's job to
determine etiquette.

The only reason Microsoft implemented this feature is because they have so many
more business users than Netscape does, and therefore hear the request more often.
if this pref happens, please default the pref to *off*.

I can't stand it when folks send the whole original message.

Comment #6 talks about how that's especially handy in mailing lists. This is
when it is *worst* to do it! Folks who subscribe in digest now have to read
message A followed by B which replies to A and quotes the whole darned A!


If it *must* happen, please please please let it not be the default.

IIRC, nobody were able to defeat the reasons for WONTFIXing, which I posted to
.mail-news (see comment #41; for the record, they are also appended at the end
of this comment). Thus, I'll go forward and mark this WONTFIX.
J-F, if you disagree, please excuse me, feel free to reopen and say why.

<quote src="news://">
There are strong (and IMO prohibitive) reasons why this bug is a very bad idea:

   * As defined by generally accepted internet standards, a signature
     is anything incl. and after "-- " on its own line.
     So, if we fix the "bug", the quote would be part of the sig, not
     below the sig (since there can't be anything below the sig). In
     other words, the message would be malformed.
     As it is our goal to track standards as good as possible, this is
     a clear-cut reason for wontfix.
   * As said, many mailers will remove the sig (as defined above)
     during reply, and some might not even display the sig during
     reading, if the user chooses so. So, in many cases (e.g. with
     Mozilla), it will be impossible or very hard to quote the quote.
     That alone is for me reason enought to wontfix this bug. Also,
     some readers might not see the quote at all anymore, although they
     just wanted to (and did) hide the sig.
   * Many other UAs (IMO wrongly) do not remove the sig during reply.
     With sigs above the reply, it will mean that they are scattered
     thoughout the quotes and hard to remove. If they are added below
     the quotes, they will all accumulate at the end of the msg.
   * The sig is arguably less important than the quoted msg, because
     the sig is always the same and usually meta-data, while the quote
     is directly relevant to the content. The rationale of quote below
     is exactly that the most important stuff comes first.

Closed: 23 years ago23 years ago
Resolution: --- → WONTFIX
Slick move Ben. Present your arguments (which have *all* been rebutted here
already) in a NG thread that you define as all-deciding and then say that nobody
played your game, so you win - very clever. Well, if you choose to disregard the
way many users want to use mail, then you will be the one responsible for the
consequences (not that you even care about the consequences). Long live oligarchy.

In the end, it's about choices; or the *denial* thereof. :(
> Present your arguments (which have *all* been rebutted here

I didn't see at least the first one being "rebutted".

> NG thread that you define as all-deciding

NG discussions *are* all-deciding. Or what procedure do you expect?

> In the end, it's about choices; or the *denial* thereof.  [:(] 

Yes, the denial to allow the disrepect of accepted standards.
> I didn't see at least the first one being "rebutted".

I should have said "rebutted here *or in the many newsgroup threads on this
subject* --> e.g., I have suggested that sigs not only have a beginning (i.e.,
"--<space>") but also an end (e.g., "<space>--"). That way the quoted text would
not be part of the sig. Also "generally accepted standards" in not what the W3C
says but rather what a *significant* portion of users use.

> NG discussions *are* all-deciding. Or what procedure do you expect?

Fine, but then the *other* threads on this topic should be considered in this
decision; not just the one thread *you* define as "all-deciding".

> Yes, the denial to allow the disrepect of accepted standards.

Repeat: "generally accepted standards" in not what the W3C says but rather what
a *significant* portion of users use. This has been pointed out numerous time
before (don't ask where, I forgot).

PS. For the record: I don'r prefer this feature, but I must and do use it when
corresponding with clients, bosses and colleagues that do. It would confuse the
heck out of them if I suddenly started breaking the format of our communication.
1. This has nothing to do with the W3C standards as it has nothing to do with
the web. This convention has been around a long time in email because it is the
right thing to do, separating email from content.

2. I don't think having the sig at the bottom would confuse anyone, as long as
you followed the company policy of replying on top (eww) I think they'd not get
confused one bit.

As I keep saying if the sig comes before the quoted text it means that the
quoted text will very unlikely be read, therefore why not fix bug 70478 instead,
then by default you don't quote the text (which is just wasting bandwidth) but
in the times you feel it necessary you can put the quoted text where you
consider appropriate for the occaision, either above the sig or below.

If bug 48570 is implemented we'll probably see how unpopular this 'feature'
really is, we should think of the recipients of this mail, I hate it when I get
mail with a huge sig that's before the quoted text. I got used to reply on top
years ago, but inserting the sig on top just seems insane. When I reply to these
users who put their sig on top I reply in the same format that I always do and
I've *never* had one of these people confused at the way I reply.
When and if Netscape 6.x gets enough market share, and if they are able to break 
into the corporate market at all, there will be many more requests from 
legitimate users that want this feature.

I'm betting that the percentage of holier-than-though users will decrease and 
Netscape will force developers to implement this feature.
If you want this in Netscape tell them:

Netscape 6.x might be based on Mozilla but it's not Mozilla, this feature
doesn't meet the needs of Mozilla (standards compliance) and at least one other
Mozilla distributor (Beonex) is against this. There's nothing to stop Netscape
adding this in their distribution if demand shows that Netscape customers want this.
> I have suggested...

I must have missed that.

> ...that sigs not only have a beginning (i.e.,
> "--<space>") but also an end (e.g., "<space>--"). That way the quoted text
> would not be part of the sig.

Nonsense. As I said, there is no such thing as "after the sig". There is no
(standard) "end of sig" delimiter, and no mail/news reader would recognize yours.

> Also "generally accepted standards" in not what the W3C
> says but rather what a *significant* portion of users use.

This is the case here. Most mail/news readers implement a recognition of this
delimiter. Since 10 years or so. BTW: IETF and RFCs, not W3C.

> Fine, but then the *other* threads on this topic should be considered in this
> decision

I have not seen others. There is exactly one mentioned here in the bug - the one
I am referring to.
*** Bug 143048 has been marked as a duplicate of this bug. ***
Except the 'bad practice' argument that keep that bug wontfix'd, anybody that
voted wontfix are really annoyed to see sigs above quoted text? It can't be that
bad if that many users have and want it; Refusing to give people a choice is
atsounding to say the least. What if people couldn't choose themes because some
may pick really ugly ones?
Let's prevent users from having something that fit their needs, since the 5%
that post in newsgroups will complain.. It looks to me that it's a coders vs.
users debate here.
By refusing to reopen (and eventually solve) the bug, you:
-**** off users that want to use mozilla, by forcing them to cut/past (moreover,
you didn't solve your issue about standards)
-force users to use another email client; given the memory footprint of mozilla,
well, why not :)
-waste way too much time to debate and read this.

anyway, who am I to think I can get coders to review their mind on such a
trivial issue? Still, kudos for the rest.

PS: I took the time to open a bug and write the above, so it has to mean
something to me.
"The answer, my friend, is blowin' in the wind" la la la... But the wind will
probably invert its direction once Mozilla 1.0 and NS 7 come out. Expect a hell
lot of duplicates then. For the business vs newsgroups usage, see comments 17,
23, 25, 28.
As for myself, I was forced to stop using a signature because I'm tired of the
everyday copy & paste. Now I 'm manually entering the signature whenever I need it.
Just logged bug 147854 which may provide a solution to this issue.  Dexcription:

To add from multiple predefined text strings at the insertion point.

For example

modifier key 1  "price list"
modifier key 2 "delivery details file...."
modifier key 3  "Signature...." 

I like command keys but a drop down selection pane, contextual menu option or
other ui implemtation could also be considered.  Preferably account centric.

...... more detail in the bug.

Othewise my two cents worth for some of the detractors of this bug. This is
definitely a usability issue. If you reject it on the basis of standards, usenet
ettiquette, or bandwith reasoning, - anything that automatically adds text to a
message is more to point - not the location of the signature.
*** Bug 148053 has been marked as a duplicate of this bug. ***
*** Bug 144042 has been marked as a duplicate of this bug. ***
Maybe there's a way to solve the "-- " delimiter problem and to satisfy business


Add an option that tells Mozilla to _attach_ the original message (mime type
message/rfc822) instead of _quoting_ it inline. Mozilla already works that way
(by default) when you forward a message.


Use "--" (instead of "--<space>") when the signature comes before the quoted text

Back to the original discussion:

Personally, I'm against allowing "signature above quoted text" (mainly because
of the "-- " problem), but it seems that this is a genuine need of business
users. I understand their point: they treat the correspondence as a _sequence_
of messages, not as _nested_ replies. Indeed, this view is against the
conventions of the "traditional internet", but I think that during the last
years, as new types of people joined the internet (e.g. home users), many of the
conventions have lost.

At least, remove the WONTFIX status of this bug. Let the discussion continue;
maybe we can find more creative solutions to this problem. I already gave two.

Well my bug has come a long way since I first started it, and I guess it wont be
fixed by 1.0 but from what people have been saying in this thread I dont
understand how it became a wontfix. I think that there are more than enough
users who would want to see something like this changed. I wouldnt mind leaving
the default at the current format which should appease the purists because only
those who really want to use it will have it turned on, and for newsgroups maybe
it could be automatically forced off, but I really think this deserves a second
Another solution: disable per default automatically adding a signature when
replying a message. Add a option in the menubar (for example edit -> add
signature) or contextmenu. So the user can manually insert a signature at the
cursor position.
QA Contact: sheelar → esther
*** Bug 129929 has been marked as a duplicate of this bug. ***
*** Bug 151429 has been marked as a duplicate of this bug. ***
*** Bug 152692 has been marked as a duplicate of this bug. ***
Please see bug 141531. It would solve the business vs. newsgoups vs. purists
conflict, since it (together with this bug) would please everybody (newsgroups =
bottom post, company mail account = top post & top-sig, personal mail account =
top post & bottom sig). Therefore, please reopen this bug.

The only minor "tweak" that would be needed is: 
IF top posts AND top signature are selected,
THEN NO "-- " delimeter to prevent the following text from being treated as a sig.
Filed bug 167319 "[RFE] Don't add signature to replies and forwards" option in
preferences". One of the worst consequences of the current state is (for reply
above users) that the signature is located away from the reply. Since most times
I don't want a signature on my replies and since I see no willing for fixing
this bug, let's hope that my rfe has a better chance.
Alright guys I know this is a verified wontfix but I just read over all the
commentary on it one last time and I was wondering if there was any way to get
this at least looked at one last time. I think there are more than enough users
to justify a pref option for this. I dont want this to change the way news works
and Im not asking for it to be the default at all, but if you have to copy and
paste sigs as much as I do you'd understand what a huge boon it would be for me
and the rest of the business world. If this is still gonna be a wontfix can we
at least get some sort of official word from the module owners or whoever is in
charge as to why its not gonna be fixed? I know there are lots of other things
that need to be worked on in regards to the mail/news module but this seems like
such a simple fix that it shouldnt be hard to schedule in. Thanx again and I
promise this is the last time ill try to get this in there. If it doesnt work im
going to actually try and figure out if I can fix it myself ;-)
If this awful thing *really* is important to you, you can write an installable
xpi to add this "feature" to mozilla. You can hand it out to your business
colleagues etc. Or you can fork the mozilla code and start your own project,
it's free software. But please don't add Yet Another Preference. Mozilla already
is bloated with preferences.
I would not be "yet another preference". I would merely extend an *existing*
preference by *one* (popular and useful) item:

Edit > Preferences > Mail & Nesgroups > Composition > [x] Automatically Quote >
Then, [ Start my reply above quoted text (signature directly below reply) |\/]
      | Start my reply above quoted text (signature below entire reply)   |
      | Start my reply above quoted text (signature directly below reply) | <-NEW
      | Start my reply below quoted text                                  |
      | Select the quoted text                                            |

We shouldn't be rejecting preferences for the sole sake of keeping prefs minimal.
See I dont understand how you can be so closed-minded about this issue Arthur,
its not just important to me, its important a large portion of the Mozilla user
base, and even more so of the Netscape userbase. And no its not "really"
important, but it would remove a needless hassle from my email work without
adding any real bloat to Mozilla, both code and preference wise. Its not like
this needs a new preference section its just one more line in a drop down box
that wouldnt take up any extra GUI space. Either way, thanx for the
consideration guys, if its not gonna work I may have to take you up on that XPI
I can't believe this is a verified wontfix! there are already 11 dupes and 9
votes for this bug suggesting that is a popular enhancement request. This
feature was originally requested as emails I sent to people had my signature
appearing after the quoted text giving the impression to the reciever that i had
no contact details et cetera. This was especially misleading when quoted text
was long.
It appears that the general consensus of these users is that this bug was
wrongly WONTFIXed. I think it's time we re-evaluate the reasoning behind this
wontfix and decide if it needs to be reopened or remain a wontfix.

Reasons to leave wontfix:

- It's *possible* that having an option like this would encourage bad habits.
- Using a separator ("-- ") may result in some mailers removing the quoted
message (I personally think cropping sigs like that is reckless and
presumptuous... does a bug exist on that? And do any other well-known mailers
remove sigs in this way? There are several other circumstances where
sig-cropping like this is undesirable.).
- Signatures before replies would wreak havoc on newsgroups.

Reasons to reopen:

- This is critical for marketability of the browser (users definitely insist on
having this). If we don't fix this, people feel strongly enough about this issue
that it will stop them from using Mozilla.
- If we don't fix it, people will discard Mozilla and send us this type of
message from Outlook Express anyway. The people that will do this are the ones
who abuse it.
- For people who top-reply, the sig is more readable (if someone sends a message
referencing a URL in their sig, the sig will have to be right at the end of the
- Sometimes it just makes more sense to top-reply. One example is technical
support messages, where lots of diagnostic information is sent along with the
message. Another is spam reports, where a copy of the full message headers will
be included with the message. It's not appropriate to put a sig below all of
that junk, especially in the case of URLs like I mentioned above. Why include
the original message? When multiple people are handling messages (common in my
two examples) it's sometimes necessary to leave the quote in like that.

Weighing the pros and cons, I think this bug should be reopened, given an
appropriate time to be fixed (this does need to be fixed promptly, but crashes
and errors have much higher priority), and eventually, be patched such that it
is available by pref. This should not be the default behavior under any
circumstances. I do think that it would be more appropriate in e-mail than in
newsgroups, because you're less likely to refer someone to a URL in your sig in
newsgroup chit-chat, or have a valid reason to top-reply. I don't think anyone
will disagree that this bug should depend on bug 141531 which would allow the
end user to control where he top-posts and where he bottom-posts. If you really
want to be rude, you could make a restriction that news accounts can't select
this feature, only mail accounts, and that might satisfy the whiners that have
wontfixed this bug. Generally I would say that *I* should have control over a
program running on *my* computer, and it's not the software's job to enforce
these not-quite-universal preferences for message formatting. If a user abuses
top-posting we could just killfile (bug 17483) them.
Keywords: nsCatFood
Id like to thank Skewer for his support on this matter. I agree completely, this
should NOT be the default under any circumstance, and I too am highly opposed to
it being available to newsgroup users because the standard usage in newsgroups
goes against this and I cant see why anyone would want to use it under those
circumstances. Anymore feedback from people for or against this matter are

If someone can tell me why it shouldnt be a pref option, other than the old "we
dont need more preferences" or "its bad practice" reasons, I would be very
interested, because I too am very interested in making Mozilla a good program
that covers as many peoples interests as possible. Just as a side note, whether
this gets fixed or not Im going to continue using Mozilla for email, I just cant
stand the way a lot of Outlook works, and I want to support the Mozilla
movement. Either way thanx for commentary guys...
> If someone can tell me why it shouldnt be a pref option, other than the
> old "we dont need more preferences" or "its bad practice" reasons

Comment 50.

> (I personally think cropping sigs like that is reckless and
> presumptuous... does a bug exist on that? And do any other well-known
> mailers remove sigs in this way?

Removing sigs during reply and sometimes even during reading is very common and
IIRC even encouraged. *Many* common applications do that. See above.
I've read all comments carefully. I appreciate Ben's comment #50 as a good
reasoning for wontfixing but, mostly, Skewer's coment #77 as the best balanced
synopsis of the situation so far. Let's summarize the set of compromising
measures that have been proposed for resolving this issue:
- don't make "signature above" as the default pref.
- Fix rfe 141531 "RFE: move quote reply pref to account setup" (pref)
- Fix rfe 167319 "[RFE] Don't add signature to replies and forwards" (another
pref). Also supported in comment #66.
- Fix rfe 147854 " Standardised Text insertion feature". Then, business users
would be able to insert their signature text on demand, without the need to
specify a universal signature that must be removed in non-business mails.
- Peter's comment #50 about adding a leading space to quoted signatures might be
useful. Or, even better, Matthew's comment #64, about removing trailing space of
quoted signatures. In the replying compose window, of course.

Adding one or two more preferences isn't a big excuce for wontfixing. Generally,
more preferences are bad for performance and ui cluttering reasons (*if* the ui
for setting them is badly implemented). However, are you realizing how many
useless or almost prefs we have now (visible or not)? Wouldn't be better to
clean them up, in order to make room for a few useful ones? I could name a few,
and I 'm sure that others with better knowledge of the hidden prefs set would be
able to add quite a few more. (please don't ask me to do that here, that would
be way off-topic).

By satisfying *any* of the above suggestions (or, even better, all of them),
current situation would be greatly improved. I think they are enough for
re-opening this bug but, personally, I wouldn't want to do that without having
convinced the drivers first.
>> (I personally think cropping sigs like that is reckless and
>> presumptuous... does a bug exist on that? And do any other well-known
>> mailers remove sigs in this way?

> Removing sigs during reply and sometimes even during reading is very common and
> IIRC even encouraged. *Many* common applications do that. See above.

Can you back this up with some specific examples, for both newsreaders and mail
*** Bug 170340 has been marked as a duplicate of this bug. ***
*** Bug 172951 has been marked as a duplicate of this bug. ***
Yet again the open source community demonstrates pigheaded arrogance over actual
if an app hides the signature (quite common), the full quoted text would be hidden.
Also fullquotes if this order are very bad to read, I don't even know, why it
should be possible to answer before giving the context.

please don't be inpolite, and don#t forget, that everyting after the "-- " line
IS the signature.
Can anyone please show me any standards body issued document that says
signatures start with "-- " and end at EOF?

My issue is that this bug means that mozilla as a mail client says, "I don't
like the way you talk, so I'm not going to talk to you" which has somewhat of an
impact on it's interoperability.
btw: mutt is able to put signatures on top, AND:

according to mutt signature starts with "-- " and ends with first empty line!

if mozilla would behave this way, there is no problem to have signatures above
the quoted text (optionaly in prefs of course)
Comment #86 From Karl Palsson  2002-10-07 19:52 wrote:
> Can anyone please show me any standards body issued
> document that says signatures start with "-- "

RFC 2646 section 4.3 refers to this convention.

The "son of RFC 1036" mentions it in section 4.1

The GNKSA v2.0 hits this in point 15a
    (see also mozilla bug 12699 and bug 76449)

Nowhere is there a clear mandate that an RFC-compliant MUA must do such-and-such
regarding "\n-- \n", but there does seem to be a widespread consensus as to the
polite behavior.

*** Bug 175350 has been marked as a duplicate of this bug. ***
Matt (Comment #88)

All three of the spec you reference make specific mention that this is for usenet.  

It has been brought up repeatedly in the comments here that this is "A good
thing" for news.  I don't think anyone disputes that.  The seemingly endless
dispute here is "we should apply news rules to everything" vs "we should apply
mail rules to everything".  Is there not room for mail rules for mail, news
rules for news?

There's a reason mail clients and news readers have traditionally been separate
apps.  Why can't we follow in at least a similar vein?

Karl P

NB: "rules" in all references above mean any one of, rfcs, generally accepted
standards, or common/desired usages
Reply above will not and should not die.  (Choice is good, remember.)  I lost a
potential convert to N7 today (and potentially Moz at a later date) and one of
the things he cited was the signature not appearing with his reply.

I tried explaining the idea of reading the history of a conversation in
chronological order by replying at the end, and he explained the need to be able
to read a person's urgent information without having to scroll down to see it. 
So while I agree with most that replies should come at the end, the real world
where people just want their answers quickly asks for replies to be at the top,
and the signatures as well so they don't have to either cut/paste them up there
or expect the recipient to know to scroll to the end to find it.

As has been stated before in this thread, this is the default on Outlook Express
(what my associate went back to).

In this matter, there seem to be two possibilities...
1. The technically correct way.
2. What many regular users want.  (I.E.:  AOL users, office workers, mom & pop)  

Politeness may be an argument against the sig above the quoted text, but aren't
we just as rude to snub what so many people want in an email client?

If nothing else, this should be implemented as a way of smoothing the induction
of current Outlook Express users.  The less they have to change how they work,
the less resistant to the change they are.

I'm also curious as to what apps hide signatures.  I have yet to run across any
such apps myself.  (Not that I've used a large number of mail clients.)
I understand now how Outlook Express got around the problem of MUAs and various
mail clients suppressing signatures...

They don't include the "-- " at all.  Nothing, nada, zip.

Given the market share of Outlook and Outlook Express, wouldn't that make them
the current criteria for "Generally Accepting Practices"?
In other words, whatever stupid thing Microsoft does, we should emulate it, just
because MS does it, because people will get used to it?
No. That's not why we're here.
Ben, we all appreciate your work for Mozilla but, from the last comment of
yours, I assume that your anti-MS bias has become counter-productive for the
project. A lot of people had posted well reasoned arguments (productivity,
business practices, business etiquette, lack of standards that explicitly
prohibit reply above in *mail* etc.). They have also proposed numerous
alternative (partial or total) solutions. You have the undisputed right to keep
this bug wontfixed. Or you may not approve any of the alternatives. But you
cannot say those people are (essentially) stupid MS emulators. After the latest
well-intended comments, there was no reason for your post. I'm pissed off and I
am very sorry of you.
> In other words, whatever stupid thing Microsoft does, we should emulate it

Ben: No, because it makes sense in certain (very common) situations. I am very
disheartened by your continual ignoring of these (often stated) facts. <sigh>

Rule 11: Design and implement Mozilla for users, not for programmers.
Dimitrios, sorry.

I was replying to the following statement and many like it before, expressively
or implied:

> Given the market share of Outlook and Outlook Express, wouldn't that make
> them the current criteria for "Generally Accepting Practices"?

It's just that this bug and many others sound to me as if MS could do whatever
stupid thing they want with their products, people would get used to it, it
would end up being a "business practice/etiquette", with this bug and similar
ones as result. I am not against any idea that MS ever had. I just think that
their products have a lot of deficits, not only the instability. I count sig
before quote as a premier such deficit.

If any *programmer*, anybody reallying working (or previously working) on
Mailnews, thinks this is a good idea and preferably wants to implement it, feel
free. But all the others: please do not *ask* them to reopen it. They have more
important things to do.

I already regret to have said anything at all.
Microsoft didn;t get where they are by NOT providing what people want. Ben your
Opinion is just that - an opinion. And whilst the opinion of people who actually
do things and work towards the solution is very important, I hope that it is
well founded.  People have voted with their feet - by not providing this feature
they are going to use a product that does it - which makes your efforts fruitless. 

I am in Bali now and was **** off,  wondered why my ex wife didn;t call me
after the bombing.. here is her reply.

"I was not near work or a place that I could access my work e-mail where I could
have found your telephone number in Bali had I know that it was on your e-mails.
 And as I told you before I didn't even know that your contact details were on
the bottom of your e-mail.  You have never told me this and as I only read your
written words and not all the **** at the bottom, how was I suppose to know?.."

I hope you get the message..and the point of our requests.

IT i started a company and called it evolving systems. From my business degree i
learnt that often businesses are constrained by what their systems can support.
I had envisaged a systems that develops and evolves in response to the
business's change needs.  The argument that news group behaviour and ettiquette
should extend to mail behaviour, despite the obvious support for this feature,
suggests that the Mozilla org is no longer able to do this.
Good Luck,
Resolution: WONTFIX → ---
Ben, I have also regretted for my over-reaction. It was quite a bit intense for
a bugzilla report.
Yes, I understand your point that not every MS bad habit should propagate in
Mozilla. Thanks to Mozilla , I use a full featured email application without
active-x, vbscript and other security holes. But we have a real problem here
(not in any case of emailing, I admit it) and it would be nice if you could
fully realise it. And there is a lot of alternative workarounds (see comment
#80, hoping I didn't miss any). 
Regardless of fixing this one or any of alternative solutions (or even all of
them), some progress has to be done. Too bad that Mozilla hacking is beyond my
(very limited) programming skills in C++. Is it possible to do it in javascript?
Re-resolving. Greg: you need to work on your manners.
Closed: 23 years ago22 years ago
Resolution: --- → WONTFIX
I kind of hate to say it, as a strong supporter of the traditional netiquette of
bottom-posting following a carefully-trimmed quote, but the people on the other
side do have a good point too.  In fact, the typical nature and general needs of
business email vs. those of "traditional geek" uses of email/newsgroups/etc.,
have enough differences that it's entirely natural that the two very different
"standards" as to how quoting should be done in replies would develop... at
first as practical adaptations to the perceived needs, but eventually (out of a
tendency in human nature) turned into strongly-held cultural values that can be
the subject of religious wars (as spoofed by Jonathan Swift in "Gulliver's
Travels" when two countries went to war over which side of an egg to crack first).

The typical characteristics of open-discussion forums (newsgroups, bulletin
boards, email discussion lists, etc.), which for some "geeks" are the main
context in which they read and write electronic messages, include:

* Discussion is typically in the nature of a point-counterpoint debate between
two or more people, where they attempt to refute, or sometimes agree with, one
another's ideas one by one.

* The discussion is a free-for-all, where anybody can chime in at any time; it
doesn't always proceed linearly, but sometimes branches and digresses in all
sorts of directions.  Thus, a linear presentation of all the messages, in their
entirety, that led up to the current one, might not make any particular logical
sense compared to a selective quoting of the portions of recent messages that
are actually relevant to the current digression.

* Past messages are usually archived where anybody jumping into a thread late
can find the earlier parts of it if they wish to see the background, or try to
figure out just how the original topic managed to digress into what is being
discussed now.  A complete history attached to any one message is unnecessary.

* Some people read the messages in digest form or in an archive file, where many
messages are put together in one message or file, and any extraneous material
(quotes, signatures, etc.) must be scrolled through whether it is at the start,
end, or middle of the message.  For this reason, and to save bandwidth and disk
space for all the many people to whom the list messages are sent, brevity is
desirable, and unnecessary material should be trimmed.

* Users tend to have an extremely wide range of client programs with which they
read and write messages; some use plain text only, some HTML; some break lines
at 65 or 70 or 75 or 80 characters, while some use infinitely-long lines and
expect the reader to re-wrap them; some prefix quotes with angle brackets or
pipe characters or colons, or indent them, or do nothing at all to distinguish
quotes other than preceding them with an attribution line.  Some of these
formatting choices violate traditional standards, but people use programs that
do it anyway.  One result is that any long message thread where each writer
attaches the entire past history of the thread in full turns into a tangled mess
of mis-wrapped lines, chaotic masses of bogus code where HTML messages are
quoted back in plain text, and other craziness, caused by the whole history
being filtered through lots of different programs' styles.

This set of characteristics leads naturally toward a preference for the
traditional Internet (geek/academic/Usenet) style of using carefully trimmed
quotes followed by the reply.  This allows point-by-point inline rebuttals,
minimizes bandwidth use, is a format well-suited for digesting and archiving,
and minimizes the ability of various mail programs to mangle an entire thread.

On the other hand, much business correspondence has these different characteristics:

* A message thread has a single directed goal, like resolving a customer's
techncial support problem; its purpose is not to engage in philosophical debate.

* Threads tend to be fairly linear in nature, as the "ball" is passed to
different "players" to be dealt with in order -- the customer writes to the tech
support rep, who writes back to suggest a solution; the customer replies that
the proposed solution didn't work, and then the rep passes the message on to a
second-level tech who might be able to provide a workable solution.  Sometimes
it might get bounced around to a number of people before it's finally resolved,
but the ball is usually in one person's court at any given time.  Thus, the
thread goes in a straight line without branches and digressions, and when it's
dropped in a new guy's lap it's useful to provide him with the complete history
so far.

* There's a much greater uniformity of mail clients, as there tends to be a
"corporate standard" (unfortunately, usually MS-Outhouse).  Thus, there's less
"mangling" of thread history even when it's attached in a big mass every time.

This is an environment where the "top-posting" style was a natural development
(whether or not its introduction was actually an evil Microsoft conspiracy). 
It's not perfect (when a message thread is passed on to a new person, "getting
up to speed" requires reading the quoted messages from the bottom up unless you
want to try to understand the story back-to-front (a clever movie, "Memento",
proceeds that way, but usually the other way round works better)), but it gets
the job done, while the alternative style of carefully-trimmed quotes would
result in the "newbie" having no access to most of the history of the current issue.

Unfortunately, styles that make some sense in their original context can cause a
big mess when taken elsewhere, but people (myself included) who "grew up" in one
style don't like changing even when the context shifts, so you get these
inevitable culture clashes with no obvious resolution.
Summary: Replying with the "start my reply above the quoted text" pref on should prepend the signature above the quote text. → Replying with the "start my reply above the quoted text" pref on should prepend the signature above the quote text. (top)
Since we are really asking for a pref to place the sig immediately after a top
posted message *in addition to* the existing pref to have the sig placed at the
end of a message when top posting, I suggest to *change the summary* to reflect
this ( ?). The current summary states that we want to
change the *entire* way we handle top posted signatures, which is not accurate.

Suggested New Summary: 
Need the Option to prepend the Signature above the quote text when Replying with
the "start my reply above the quoted text" pref. (top posting)

This will make it even more obvious how unnecessary wontfix'ing this bug was.

Please see comment #74 and comment #70.
i guess to sum it all up there are two different users, both with different
needs from an email client. There is a Usenet user where discussions may branch,
and there is a corporate email user where discussion will not branch. The
corporate email users are asking that signatures are always placed under the
email text body they write no matter where the quoted text body is positioned.
Currently the Usenet users (with exception of a few) seems to be imposing
conventions and netiquette on the corporate email user who wants ease of use and
pracicallity. From the direction all these comments of this discussion (tho
somewhat heated at times) will continue along similar lines for as it is left at
a wontfix.

Ultimately, it would not solve the real problem for corporate email users. So in
suggesting a solution, is it possible that this bug be re-open and reassigned to
some one like until some one wants to fix it. That way the
people who do not want to fix it won't need to and leave it to the people who
want to fix it to find a person who is willing to act upon fixing it.
> Ultimately, it would not solve the real problem for corporate email users. So in
> suggesting a solution, is it possible that this bug be re-open and reassigned to
> some one like until some one wants to fix it. That way the
> people who do not want to fix it won't need to and leave it to the people who
> want to fix it to find a person who is willing to act upon fixing it.

Ben will just re-wontfix it. Don't forget that Mozilla is his project and it's
up to him, not the users, what features make it in.
And it is up to the users to decide whether or not they will use a browser/email
client that refuses to include the features they desire or believe they require
to do their job quickly and efficiently.
Since Mozilla programmers won't fix this, I'll have to repeat the suggestion to
the Netscape programmers.  The same suggestion will be made to the KMail and
Evolution developers.  (Especially the Evolution people since they have the most
Outlook-like appearance and look like the least disruptive transition for
Outlook/Outlook Express users to convert to.)
Sooner or later, I'm bound to find an email program suited for business
communications instead of usenet posting, at which point, I'll reevaluate if I
need bother with Mozilla as my browser as well.
I will be forced to vote with my feet also.
Why in the world is this a WONTFIX??!! This is a very reasonable request. Who
wants to scroll all the way to the bottom to get contact information? And who
cares if the reason the message is long is due to whether it's a reply to a long
message or a reply to a conversation? It's a reasonable request! I like to be
able to click on a message and then pickup the phone. And I want the recipients
of my messages to be able to do the same. Is it truly that painful to give users
a CHOICE? I'm a loyal Mozilla user, but reading some of the closed-minded,
self-righteous comments above leaves me with a bad taste in my mouth.
If you'd trim your quotes in the time-honored way instead of attaching the whole
message or thread to the bottom of your message, nobody would have to scroll to
get to the signature at the bottom.
Dan, I understand your point, and for the most part I agree. But my question is,
you think it's ok to FORCE me and the people I correspond with to trim quotes
"in the time-honored way"? I'm sorry but I feel that my freedom is being inhibited.

You may consider it lazy, wasteful behavior, to just quote the whole message,
but sometimes its necessary, and it ALWAYS saves time and effort. If I wanted to
be completely efficient with resources (other than time and effort), I would
chop my firewood instead of revving up my chainsaw (ok, maybe a bad example). At
any rate, the thread (and it's quoting) quickly peters out, life goes on, and
meanwhile the impact on the net was negligible. (I truly hope that is what you
are worried about and not the violation by other people of any "time-honored"

Time and effort are valuable commodities to me. In my opinion, the purpose of
software is to save me time and effort. Right now it takes LESS time and effort
to cut and paste my sig back to the top where it needs to be, than it does to go
through and trim out cruft. Sure would be nice if I could eliminate even that.

Well, when I'm reading a mailing list in digest form, it wastes some of my time
and effort to wade through the crud that people leave attached to messages in an
ever-expanding tail, including heaps of old quotes plus various footers inserted
by list managers and free email services (including ads)... and if some of the
messages somewhere in that chain were in HTML form, sometimes there's a heap of
raw, ugly code in there too.

On the other hand, as I pointed out myself earlier in this bug, there are
certain types of message threads, especially in a business environment, that do
work more naturally in the top-posting style, much as I hate to admit it. 
Anything that goes linearly in a "follow-the-bouncing-ball" way, like a tech
support issue that bounces between the customer, the first-line tech rep, and
escalating levels later, works all right with top posting and allows the
complete history to be retained and forwarded when the message passes on to
others in the chain.  So I don't think top-posting should always be banned; just
strongly discouraged in places where it's inappropriate (like almost all
open-discussion lists).
OK, point taken. I wonder if there would be a way to tie sigs to destination
addresses somehow. In other words, sigs for mailing lists could be handled
differently than sigs to your CEO. Of course this would delay everything until
after your email address had been specified...

However, I'm not convinced that adding this functionality would truly impact
your digests. I think that people can honor or ignore netiquette regardless of
where the sig is inserted.
Dan, you have good points with regards to discussion lists.  However we seemed
to have bogged down in an apples and oranges comparison.  The nature of
discussion lists is such that those who participate more in those would already
have their preferences set to start their reply below the original message.
So if discussion lists are the apple, business communications are the orange in
this.  These are the people who are generally replying to customers, vendors,
coworkers, associates, etc in a one-on-one manner where there is a high
likelihood of an invitation to call for further discussion if desired, at which
point the worker must then either type in their phone number, cut and paste
their sig (I'm assuming contact info is in the sig for this) from the bottom,
inform the person that they're corresponding with that contact information can
be found after their prior correspondence, or hope that the
customer/assoc/coworker/etc is so intent on finding the information that they
check over the whole message even though they didn't need the courtesy reminder
of what they'd written and ignore their prior discussions long enough to find
the number at the bottom completely orphaned from the latest conversation.
*** Bug 180046 has been marked as a duplicate of this bug. ***
Please re-open this bug; I think by virtue of the number of posts about it and
the number of dupes, it stands to reason that this would be a useful feature for
many users, however unruly it may be. Leave it off by default if you don't want
most users doing it; as most of us realize, defaults are what most users use
anyway, particularly if they look weird. I propose the following: a check-box
under the reply-above option that says "Insert signature above quoted text".
Anyone should be able to understand this, and if they don't, they'll probably
leave it alone anyway. As has been stated, there is plenty of space on this
screen for another option.
i agree
As stated before: You're free to "fix" this yourself. If you really want to
implement this, reopen the bug, assign it to yourself, come up with a tested
patch, probably even providing an installable xpi. Make sure to set it off on
default then you can discuss the issue again with the module owner and others
(preferably in the newsgroups). But you can't expect other developpers whom you
don't pay and who don't really want this feature to come up with a patch.
I wish I could do all that you said. As stated before, most of the people who
like the current status of this bug are developers (ie, the ones qualifed to fix
this). The rest of us are fairly helpless when it comes to fixing things. Our
contributions are limited to pointing out missing usability features and bugs
from a users perspective. This type of contribution used to have (a little)
value. All we can do after that is hope some developer takes pity.

Nevertheless, I understand how hard it might be to work on something that you're
just not interested in.
As a module owner, I reopen this bug to be sure we don't forget about it. I
havn't made my mind yet about if yes or not we should fix it and how. The only
thing I am sure off is that I have plenty matter to make my opinion...
Resolution: WONTFIX → ---
I never top post but I think it should be an option.
If someone implements it then don't forget to fix the "-- " stripping.
I like the idea of stopping at the first blank line.
As much as I sympathize with the so-called "closed-minded geeks," I think
etiquette/netiquette is something that should be enforced on a newsgroup/mailing
list level, not a program level. Programs enforce technical issues, not people

The preference should certainly default to quote below, but Mozilla should at
least allow this setting to exist. I frankly don't understand why Mozilla is
content to allow users to quote below reply, but somehow the sig placement is a
holy war.

As a side note, with the exception of a few comments, most replies to this
discussion are content-void. If you took out all the MS flames, geek flames, and
"you suck" messages, we'd have less than 10 comments.

Put the option in (non-default) and educate. That's the best compromise.
I cant belive this was at some point put as wont fix. 
I really dont understand the closed-mindedness of some people some times. 

True, proper etiquette should be the default.
But as many people request this enhancement then people should have the choice,
at least in the user.js.

I believe:
1.If most people want a feature then it should be developed and set as default.
2.If many but not most then it should be an option via the UI.
3.If some then i recon it should still be a user.js option ?
4.if only a few then, yes wontfix we go.

I recon this request is a 2 or 3 above, and if we include the end users then a
definte 1 ?

Many, if not most, Mozilla users will probably not toppost or at least wouldnt
want the sig at the top. 
On the other hand many ( if not most, again) Netscape users would prefer the
other way.  Which may be out of either pure ignorance or company policies. And
since the Netscape distributions will have the most users(probably) then
shouldnt it dictate what options mozilla develops ? ( I might be completly wrong
there :0).

I agree that if outlook or oe has a feature (or a bug!) then mozilla should not
necceserily include it, but if enough people request a feature, irrelevant of
whether outlook has it, then i believe it should be implemented.

One feature I do like in outlook is the option of two differents sigs. one for
new messages and one for replies/forwards. The new message one would be the full
sig, while the reply on could be a simple one liner to be put after your reply,
but not at the end.

What we do need to consider is the difference between signatures in mail and
news. Not just which signatures to use, which we currently can choose, but also
the behaviours per account or account type? .

I just have been flames from one of my directors for having installed Mozilla
mail because it puts the signature below the quoted text. He said that in all
his life he has always seen it done differently, and that this is non-standard,
since he has *never* seen it done. Of course, he doesn't also know what proper
quoted text is.

There is a world out there with users that don't understand it, and are more
than confused to see the signature at the bottom, because they tend to think
that the user wrote the whole message. 
If we are putting the quoted stuff underneath, which anyway IMHO sucks, there is
in fact no reason why we should not go the whole way and make it possible also
to put the sig above it. 

Current situation IMHO is mixing the approaches, and while I think it should
remain as an option, it's really confusing some users.
The Call Centre employees @ my work find this bug to be a real annoyance --
Honestly I would have to agree. Each Call Centre employee responds to 80+ emails
a day. Cutting and Pasting the signature from the bottum of the reply to the top
for this many emails can become very annoying, and a waste of time. Heck, I find
this an annoyance, and I only reply to 10-15 emails a day.
I have voted with my feet.....Bye..
8 months ago I was a "not so fond" user of Outlook Express 5.x/6. I found the
functionality of the client very limited. So I set out to find a better client.
I found and decided to use mozilla because of the functionality, stability, and
*numerous customizations.

I understand that there are those of you that dislike this feature, but there
seems to be many that desire this behaviour. Why don't you leave the option to
the user. 
Can we try some new rules? People comment only, if they have something new to
add, something that has not been said before on this page. If you just want to
voice your unhappyness with this bug, then vote for it, that's what the bugzilla
feature is for. The amount of bitching on this bug with only 18 votes is driving
me crazy (yes, even more crazy than I am normally).
Whiteboard: Read comment 125 before commenting
I also got a request for this feature from a manager...

Anyhow, a simple suggestion: Add an "Insert signature" function that will insert
the file specified for "Attach this signature:" at the current cursor location,
even when the "Attach..." option is switched off. Balsa does this (and also
allows singature to be prepended, and disabled for replies even when it is
enabled for new messages or vice versa...) 
This current bug is the single most annoying "feature" of Mozilla and it 
utterly floored me to see that this bug has been open for 2 years without a 
fix due to politics.

Programming is not about forcing your users to do the "right" thing, it's 
about writing software to meet the needs of users. Obviously a lot of users 
would like the ability to put sigs on the top of replies. It's common 
etiquette in the business world.

Those who mis-use it in newsgroups and mailing lists can be educated in those 
For Toralf: your suggestion is bug 147854 (extremely useful indeed, if I had to
choose, I'd prefer it instead of this one). For all people: so many "me too"
comments, so few votes, as Ben Bucksch already noticed.
If you simpathise or want this feature please take 30 seconds of your internet 
time to register your vote.

To make it easier, just follow the link below:;bug_id=62429

Make sure you check the box next to 62429 or else your vote does not count!
I'll turn it off but I agree that it should be an option.
I'd personally like to see this "feature" be implemented in Mozilla.  As a
previous academic user of the internet, and now a business user this is a very
viable and useful feature.

1.  I like to see the contact information for a responder with the text of the
2.  It's easier for me to shorten the email stream and still keep signature
3.  People want it.  Period.  That should be enough.
4.  I don't have time to try to match up sig's with message text in an average
day.  Perhaps the academics do, but I don't.

So - will this be implemented and when can we expect it?

*** Bug 195570 has been marked as a duplicate of this bug. ***
Hi All,

   This will really make writing a business reply a lot
easier.  I hope it get implemented soon.  :-)

I'll vote for this one too, and I disagree with many ppl here who don't find
this feature useful (I can understand that everybody have a different way of
replaying) and THEREFORE think it's not worth implementing, EVEN AS AN OPTION !! 

Are we living in a word where imposing a way of thinking is legitimate ? Is it
the way Mozilla has been designed ? I hope the answer is NO an NO.

I'm not a moron, but some comments on this bug really prove their writers are
thinking that their opinion is the only one that counts ... 
Enough insults already. Nice that everyone ignores comment 125 and keeps flaming.
FYI: I originally closed this bug mostly out of *technical* reasons, *after* a
public discussion.

Implementor: Whatever you do, please in no case put "-- \n" (the conventional
sig delimiter) before the quote, this would be a violation of the conventions
and would cause severe problems on other ends, where sigs (including quote,
then!) are removed, hidden, whatever (e.g. default config in Mozilla and other
This implies that the recipient mailer is unable to recognize the sig and to
remove it automatically during reply. But I guess nobody cares.
Ironically, VERY ironically, I think that if the signature is placed above the
quote, it would be proper to put two dashes and a newline like this:


in a fashion similar to Outlook Express. I always thought it was stupid of them
to add an incorrect separator, but it actually makes sense to do it that way
when top-replying since, if it was a proper separator, the whole quote would be
removed. To the naked eye it looks the same as a normal sig separator, but a
newsreader will not remove the sig in this instance. Of course if the sig is at
the bottom the separator should be correctly formed with a space.
Why the heck hasn't this been accomplished if it was opened more than a year
ago.  This is a GOOD bug since if you start a reply above the quoted message,
your signature should be above it too.

Right now, I have to CUT the signature from the bottom and then paste it between
my reply and the quoted message.  Very big pain in the ass.  

I'm too lazy to read the complete discussion of this but provided the issue is a
year old, it really needs to be added.

I'm using Mozilla 1.4a
>I'm too lazy to read the complete discussion of this but provided the issue is a
year old, it really needs to be added.
Surajit, the only thing that can be happen after adding so many useless comments
like the one above is pushing someone to close this bug as wontfix, again. You
feel too lazy to even add your vote, ie the minimum of action required to OFFER
something to this bug. Or to read comment #125 and comment #135.
------- Additional Comment #135 From Ben Bucksch  2003-03-06 02:26 -------
Implementor: Whatever you do, please in no case put "-- \n" (the conventional
sig delimiter) before the quote, this would be a violation of the conventions
and would cause severe problems on other ends, where sigs (including quote,
then!) are removed, hidden, whatever (e.g. default config in Mozilla and other

So, don't make it a "real" signature. :-)
I've voted for this bug a long time ago but there is nobody to work on it. :-(
People, if we ever want to see this thing fixed YOU have to vote for it!!!!
If the only reason this bug never gets fixed is because of comment number 8, and
9 (replying to newsgroup threads) then it would seem logical that this bug
fix/enhancement should *only apply to email -- *not newsgroups. I'm not a
programmer, but I imagine it would be relatively simple to create a simple
"IF... THEN..., ELSE..." statement to test if the reply is for a newsgroup
ie: IF reply = "newsgroups thread" THEN do nothing.... ELSE prepend signature...

In lue of this suggestion/comment I suggest changing the summary to inclued
something along these affects mentioned above... Therefore appealing to those
against, AND for this bug/enhancement....
Just out of curiosity...just how many votes does it usually take? i.e. I've seen
(and voted for) bugs with MUCH fewer votes than this one get fixed. Somehow I
get the feeling that even with 100+ votes, nothing would change.
Metacomment: Maybe should think about adding a feature where 1 vote
is automatically added for every month a bug remains open.  That would help
automatically prioritize old bugs that are lying around.  There seem to be lots
of bugs like this one that have been open since, say, 2000, have hundreds of "me
too" comments, but no real activity.  There are worse examples than this in the
bug database...
Bugs, unlike wine, do not get better with age.  :)
Of course, it's the hundreds of "me too" comments that make bugs less likely to
be fixed. You have to remember that any developer interested in fixing a bug
will first have to read through all the comments to find relevant information.
Most can spend their time more productively than reading those "me too" comments. 
Which is why the votes are there - they allow developers to see there's a lot of
interest _without_ other people needing to clutter up the bug. (Which, btw, is
the only thing votes are good for - they won't *force* anyone to go and fix

There, another good deed. I've just made this bug slightly less likely to ever
be fixed. :) (Apologies for the spam though.)
Jean-Francois Ducarroz, Peter Lairo:
   I'd be particularly interested in your opinions of the following.
   I think this proposal might appease all interested parties...

I propose a wontfix on this bug.
Before folks get upset, here's my reasoning...

nothing below a signature should be significant.
    this is why we have "-- \n"
    this is why smart MUAs can recognize "-- \n"

sig should be after your message, not after a big looooong quoted text when the
only reason the looooong quoted text is there is for archival purposes. If the
sig is way at the end, then users don't see it or know to look for it at the end.

folks who want to include the entire thread for archival reasons want to have a
sig, and that sig should be immediately after their message content... not after
the archival material.

why not have that archival material included as an *attachment*?

In most apps, the attached material will be *rendered* below the sig, yet the
sig will still be a "real" sig and can be ignored by readers, and the sig will
immediately follow the original content. It seems that this solution works for

So, it sounds to me like what folks voting for this bug really need is an RFE
for a new feature which allows them to "reply with history" (and "reply to all
with history") which will wrap the current message up in a single attachment,
attach it to their new message draft, and let them type their message.

Since MIME parts can be nested, this solution should support two like-minded
users "replying with history" at each other and building a larger and larger
attachment with each reply. If their employers encourage this behavior, then
they can deal with the increased bandwidth and mailspool sizes.

Whether or not the original message is also "> " quoted in the body of the new
draft would follow their normal mailnews prefs. So, reply options would include

    reply to all
    reply with history
    reply to all with history

Please don't everyone respond with "me too"s on this. If you like the idea, then
go make an RFE and reference the bug number here for folks to sign on. If you
think the idea sucks, say why.

If nobody thinks the idea sucks, then I propose that Ducarroz mark this bug
invalid. and let whomever is interested move over to the "...with history" RFE
which presumably someone who wants this feature would create. That RFE might
want to mention that provisions should be made for a future pref regarding
whether the "...with history" replies include non-email message attachments in
the archival "MIME-ball" attachment.

I've pretty much been CC'ed on this bug so I could watch it and my hope was that
I'd see the feature request die. Content should *not* come after a sig. 'Twas
just this morning that it occured to me that content *can* come in an attachment
and attachments are typically *rendered* after sigs, so that might make everyone


PS: if someone opens an RRE for "...with history" features, and this bug is
invalidated, all of you who have voted for this bug will want to remember to
move your votes to the new bug.

I've been slowly transitioning people at my work to mozilla, and just about
every person compains about the default "forward as attachment" setting, since
they can't see the message they're forwarding, and some recipients (with a
different client, don't know which ones they were) didn't like it either.

The three things I have to do on each mozilla transition is:
 o replies at the top
 o send as text & html
 o forward as inline

Now that I have responded to that, let me add that even though I think everyone
should reply at the bottom, I think we should add the sig at the top just like
this bug is requesting (when replying at the top, of course).  There has been a
standards compliant way to do it proposed, and it should just be done.

Luckily, I don't have anyone here that is using a sig, but I'm sure that once
someone does start using it, they will NOT like mozilla and it will not make any
good will to keep mozilla with this problem left intact.

Simply said, you WILL lose mozilla mail users if you don't implement this
feature.  And where will Mozilla world domination be then? ;)
A trouble with comment 145 is that, in some mail readers at least, getting to
the content of a multiply-nested MIME-attached message can be like peeling an
onion, opening up one of those Russian dolls-within-dolls, or taking off all 101
hats of Bartholemew Cubbins (in the old Dr. Seuss book) -- in other words, you
keep opening one only to find another inside it, and so on, until you get really
tired of it.  Pegasus Mail is one such reader; it shows attachments in a
separate pane where all top-level attachments are listed, and you have to keep
expanding it into a deeper "tree" by clicking on the right spot to open up
subnodes at each level.  It can be a lot of work, and carpal-tunnel-syndrome
inducing, to get to the bottom of a degenerate case of multiple nestings.

And re comment 146... Why turn on "send as HTML"?  That's another abomination,
even worse than top-posting!
regarding "can't see the attachment they're forwarding"
   you are correct to turn on "forward as attachment" for these users
   however, I think that issue is orthogonal to my proposal
      in the case of my proposal, they would
          - set prefs to quote message when replying
          - then use "reply [to all] with history"
      which would *both* put the full history as an attachment
      *and* quote the text for them to whittle down as appropriate
      to provide context to their reply in the traditional manner

regarding "russian doll" syndrome
    how about if instead of nested attachments, the "...with history"
    feature would explicitly serialize any nested parts, preserving
    the order so that
     would be sent in that a..h order *serialized* rather than as a
     single nested attachment containing 'a' and 'h' (and 'a' contains
     ...). This would still require Pegasus (and similar MUA) users to
     unpeel an onion, but at least it would only have one layer of skin.

Are there other problems with the proposal?

Are there new problems introduced with the above adjustments?

Ok, so now you're going to take several mime messages and condense them, each
one possibly having an image sub-mimed.  And all to avoid adding a sig at the top!

1. add sig at top
2. add "--\n"
3. add two lines above sig
4. move cursor to first line

And WOW!  No mime processing added!

And yes I know about html mail, that's why it set to include plain text as well.
 And I used to use mutt all of the time (a text only MUA) but I switched to
Mozilla Mail only because "Well, if my users are going to use it, then I will
too".  So, please let's not get into the html mail issue here.  Oh, and with
Mutt, I could see everything except for the pictures in the html mail (just
process it through lynx or links...) :)
Without getting into too much detail, I want to make it clear that the
suggestion to "forward as attachment" is an unacceptable "hack" IMO (sorry). (1)
I am forced to use Lotus Notes at work and just hate the "reply" vs "Reply with
History". (2) I also can't stand how "users" only "reply" and I have no clue
what they are referring to. (3) This bug is much more about *Replying* than
about *Forwarding* (your example merely makes some sense for forwarding). (4) I
want quoting that i can enter text in between, (5) I want to *see* what i am
quoting. (6) Nested attached *files* will get forwarded/replied forever. I will
never use attachment for FWD/RPLY and continue to consider this bug a major
deficiency for corporate use. 

If you want another behavior, please file another bug. Please don't discuss it
here (suggestion: newsgroup).

PS. Matt: You obviously put some thought into your suggestion, that is noted. ;)
Regarding Comment 145 -- You 'Comment' sucks!!! I hate it when people do a
"...forward reply as an attachment..." When I'm reading a reply to a thread I
want to be able to reference the previous message easily/quickly. Besides, this
would require another window to be openned. On a slower computer where Mozilla
is already *somewhat slow (yes, I agree it has improved greatly from the 1.01
release) all this would do is use more ram, and peole don't want to wait another
4-10 second waiting for an attachment to open just to read the previous peice of
the thread....You need to remember the average user doesn't have the latest
Pentium 4 processor with 512MB or more of ram. You need to remember that the
minimal install base for mozilla is a 266Mhz system. I imagine mozilla already
runs slow enough on that system....One more excuse for the average user to go
back to the 'slimmer' M$ LookOut Express

Two years ago when I did tech support so *many stupid users would call asking
why they got mail with the message as an attachments. M$ Look-Out Express has
this functionality, and unfortunately most ISPs encourage the use of LooK-Out

I'm experiencing the same issue as comment 146. I have transitioned about 45% of
the office over to Mozilla Mail, and everyone complains about this
functionality. Some users including the CFO (Chief Financial Officer) decided to
drop using signatures all together because of this bug.

Anyways, what's wrong with my comment 141
Hi All,

    At the risk of getting hate mail, I do believe the only way to get
this fixed is the same way bug 134492 got fixed.  See:

    Yes, I know, Open Source is not suppose to work this way, but
we may all have to hold our noses on this one.  (Bug #134492 was
a similar bone headed bug that the developers also refused to fix.
It is like they hate business users.  It is bad enough that they refuse 
to create an Evil Empire plugin so that Business to Business web pages
will start to work with Mozilla, but I think that they have a
problem with business users.  Very sad.  :-(  )

A legitimate bug to me.  Why?

Literally thousands of employees in my very large company use Lotus Notes.  I am
(perhaps) the only user who could convince the Lotus Notes admin to allow me to
use IMAP and thus any native Linux client (Mozilla) for email (Notes in wine is
still buggy).

EVERYONE in our company inserts their signature _above_ the rest of the mail
when replying; it's the Notes default, even if it's not netiquitte.  So almost
NO one will see my signature (which is required, by the way) if it's at the
bottom, nor will they see my reply if I put it right above my signature at the
bottom.  BOTH the reply AND the sig MUST be at the top.  I tested it on our MIS
PC tech (who is not a computer idiot), and he wondered why I didn't type
anything.  So I _must_ cut+paste the signature.

As a programmer, I'm familiar with stupid requests, but this is a legitimate
need by me and countless others.  It ought not be difficult to fix.

I voted for it here:;bug_id=62429

This really shouldn't be an issue at all.  The reality of it is, Mozilla is made
for the end user, and only through doing that will Mozilla ever regain the
market share it lost while standing still with IE running circles around it. 
(we all know the rants that start here, so lets skip them).

  The reality of it is, countless businesses put their signatures with the
message, because it adds credibility to what you said.  It lets people take a
close look at who is sending them this message.  In E-Mail, people know what
they wrote based on the subject and the fact that they just sent that message
which is being replied to.  People use it, and people want to see it- or this
bug wouldn't be here.  People shouldn't HAVE to use a 'workaround' from
Netscape.  Otherwise you keep the netscape users and annoy the rest.

  Outlook did it, Outlook Express did it, Lotus Notes does it, Eudora I recall
did it- but that might be a few versions ago.  Clearly some people like it that way.

  So make an option to do it.  There's no harm and those on each side are happy.
 I suspect it easy as pie to do and make an option in the config file and a
checkbox., line 789:
   int position = doc.getEndPosition().getOffset() - 1;

Add in an if with a configure option and if someone checks it:
   int position = doc.getStartPosition().getOffset() + 1;
and it's good to add a blank like up at the top as well, using something like a:
    doc.insertString(1, "\n", null);

Clearly this can be put in with about five lines.  This is the first of me
looking at the source code, so I might be missing something, but it sounds like
people are talking it up to be needlessly complicated.

I will try and make a patch if I can figure it out, but this is the first I've
looked at the source.  Otherwise someone should really just fix this once and
for all.  Bug opened in 2000.  A joke that it's not fixed in 2003.

But Outhouse and Outhouse Excess have the reverse problem... I don't think
there's any way to get the darned things to put the signature where it belongs,
at the very bottom, if you want to bottom-post following a carefully-trimmed
quote.  (They also won't let you put in a proper signature delimiter, since they
strip trailing spaces from the message.)

On the other hand, my usual mail program, Pegasus Mail, always puts signatures
at the very bottom, so it's in the same "boat" as Mozilla.  In fact, since the
quotes are added after you hit "send", you can't even cut and paste them if you
want to put them somewhere else.

Sure, I think this bug should be fixed... if Mozilla is gonna support
top-posting at all it ought to do it right.  But I wish more people would
consider using the more sensible bottom (interleaved) posting style, where a
bottom signature is in fact right next to your message where it belongs.  Yes,
there are situations where top-posting makes some degree of sense, like for some
sorts of tech-support and customer-service responses, but even in a business
environment there are many, many other cases where the time-honored Internet
style works great.  Every time I get a top-posted message with garbage trailing
off at the bottom (like tin cans tied to the bride's car after a wedding), I
wish somebody would at least trim off the multiple repetitions of ads and
disclaimers that pile up at the bottom of a top-posted thread.
I'm a business user *and* a long-time user of unix mail.  Cutting and pasting my
sig is tedious but absolutely necessary when sending email to customers and
colleagues alike.

I'll pay £40 (pounds) via paypal to the developer that fixes this bug.

If this is the best and fastest way to get this bug fixed then can others also
put their hands in their pockets to make this a great incentive for the Mozilla
developers.  (Thanks for all your efforts so far guys, it's a great product!)
I'll give € 10 to whomever fixes this bug (via PayPal).
I'll give 10 euros via paypal too for this bug to be fixed.
Dan Tobias wrote: "Yes,there are situations where top-posting makes some degree
of sense, like for some sorts of tech-support and customer-service responses,
but even in a business environment there are many, many other cases where the
time-honored Internet style works great".
I agree (though those cases are not as many as you suggest). I think the
ultimate solution would be the ability to place sig at will and at any position
(bug 147854). I would prefer to see that bug fixed instead of this one. Note
that most sensible sig users wouldn't like their sig to appear in personal
emails or in situations were a minimum of privacy must be preserved. Let's
restrain signature usage to cases where they are absolutely needed!
NO, let's fix THIS bug instead of diverting attention to another bug.

After looking at bug 147854, I like what it is describing, but fixing this bug
does not keep you from also fixing bug 147854.

People are already promising money to get this fixed, so let's get it done!

With that said, I'll add $50 (US currency) via paypal to the pot.
Since there are now lots of comments on this bug, I'd like to suggest that,
should this feature be added to Mozilla/Thunderbird, there be a click-box or
something in the compose window so that the user can, while writing the mail,
switch from top-posted to bottom-posted. Then the preferences setting would only
be a default. I think this would be a really nice feature for users who send
both top- and bottom-sig mail on a regular basis.

I imagine this adds quite a degree of complexity, though, so I won't complain if
whoever adds top-post sig's ignores this suggestion. :-)
I'll pledge US$10 for this to be fixed.  

PLEASE make sure you've voted for this bug if you want it fixed.  And recruit
your friends!   We have 49 votes and only need 1 more to get listed in Bug
163993 [META] Mozilla Bugs with Large Community Interest.  
Yippee!  We made it to 50 Votes.  Adding to Bug 163993!  

Can we get this on the list to be worked on for Mozilla 1.5a?  
Blocks: majorbugs
Keywords: mozilla1.1
While I personally find top posting, and sigs before quotes revolting, I see the
point some make for an alterante. I'm glad the start reply at top option is
there. I use it, since I start at the top and trim the unwanted text as I go
through the message replying to the parts I want to.

Untill one of the many suggested fixes gets implemented, perhaps this could
help. appears to have the ability to insert multiple
lines of text on request in the mail composer. I have not check to see if it
will inserts at the current cursor position, but it probably would be a simple
fix, one that the author might make for you.

The you turn of the sig feature of mozilla, and use the <Ctrl>+J hot keys to
insert the desired signature. It would also be easier to modify, since it's
written in js. If desired the required parts of JSlib could be added to a custom
version of tagzilla, to make a single xpi install file for just this function.

I also surprized that tagzilla hasn't been mentioned before. It's been around
for a while.
testing for bug 204280
I would also like to see this done. After much research into what IMAP client(s)
to run at our company. We decided on Mozilla 1.3. It's fast, stable, and has a
lot of nice features. This is the only thing I've gotten that even resembles a
complaint. If you didn't want to tie this to the Start replay above or below
quote option, maybe a seperate option. There are good arguments for both above
and below replying. One one hand if you've followed a discussion the whole time
you want to see the latest reply at the top, rather than have to scroll through
10 pages of text. If your catching up on a topic, you want to read top down.
Thats why the option is nice. It just makes sense to take it one step further
and do that with the signature also.
If you have to scroll through 10 pages of text to get to new bottom-posted
replies, then somebody is doing *way* too much quoting... learn to trim a little
(or a lot)!

This may be of interest to some of you; I've recently created Web pages
discussing the issue of top and bottom posting, and other issues regarding
e-mail quoting (as part of a site I'm working on, still under construction, on
the topic of e-mail formatting).  The discussion in this bug was very helpful in
getting my thoughts together on it, and some of the content of these pages is
based on a revised and expanded version of my own comments earlier in this bug.
Does the work to fix this involve more than one component? I am a developer and
am so interested in getting this *fixed" or behaving to suit both cases that I
am prepared to invest my own time. 

Could any of the developers comment, help me out on what I need to start work?

Here's what I propose:

In account settings the check box should be removed and replaced with options:
#1 I will attach as needed in composer, #2 Attach to replys above quoted text,
#3 Attach to replys below quoted text(default), #3 Never attach to replys.

A tool bar button/menu click should be added: Insert signature at cursor.

Maybe this mod wouldnt get checked in to the Mozilla cvs but seen as so many
want it its likely it will. If not then I'll just have to add my mod to every
release, and re-compile. I'll make the mod'd components available to anyone who
wants it.

Thats all....
Pete. UK

RE comment #168: I don't think that is a good solution because it could
contradict the preference "start my reply above/below quoted text". Your comment
is actually very similar to comment #74. Maybe you could do it that way. See
also bug 141531.
I agree, very similar, however consider the case where the signature insertion
behaviour needs to be different for each account news/mail. From what Ive read
here thats why this is such a busy discussion - personal/business email and news
all require slightly different behaviour.

Perhaps the Composition settings are just in the wrong place, and should be per
account and as comment #74? After all, accounts can be email(above/below) or
news(mostly below). 

Pete. UK
That is exactly what bug 141531 comment #5 is for. ;) thank you for intenting to contribute your programming
experience. The main obstacle for a patch to get r/sr is the issue described at
comment #135. Or, (perhaps) you could make it an installable xpi, thus saving
the bandwidth needed for downloading the whole binary.
"A tool bar button/menu click should be added: Insert signature at cursor" is
bug 147854.
Finally, adding dependency on bug 141531 (hoping you all agree with this).
Depends on: 141531
Comment #70 addresses comment #135 (if top-sig then no "-- " (perhaps "--"?)).
This should at the VERY LEAST be an option for the user.  I can't believe so
much discourse has occured for over two years without any action being taken.

If you start your reply above quoted text, the signature should follow right
underneath.  If you are replying to one who has wrote an epic in their message
and the signature is underneath their message but your reply is above, the logic
does not flow that people will know that a signature is underneath - I know I
wouldn't be willing to take the chance that people will check.

If the signature remains at the bottom, an automated "pre-signature" message
should be appended at the top along the lines of "signature at bottom of e-mail"
or something.

> If you start your reply above quoted text, the signature should follow right
> underneath.

Not, it should not. There might be a preference for this (depending on the
module owner), but I for one choose to start my reply above the quoted text, and
then work my way _through_ that quoted text, trimming where appropriate and
responding to individual points, and then having my signature placed at the
bottom of it all. Any change that would prevent that particular workflow would
be a really bad one. 
Not to mention the countless of people who'll want to continue to do things the
right way even while top-posting.

I know you probably didn't mean thing that way, but I think it's important to
make the point that putting the sig at the top should not be directly tied into
the pref for top-posting. (And it's not as if people who're annoyed by the noise
in this bug haven't long since removed themselves from the cc-list anyway.) :)
S T O P   S P A M M I NG   T H I S   B U G ! 


Sander: See comment #74. The option would be in ADDITION to what you want.
To the guys saying that it is good to have the sig on top, so that they won't
need to scroll through the entire message: How can you be sure that there is not
anything else written below? When I get a long message with some comments on top
I just don't read the comments I can see, I still have to go through the entire
mail. What business will let their employees care just about the first page of
corporate mail? 

Fixing this bug will only make it even harder to find the new information in a
mail converation. Quoting should not be used for history, that is what mime is
for. Quoting is for simulating a conversation where the author is replying to a
specific comment in the original mail. <SPAM>When is voting against arriving?</SPAM>
> To the guys saying that it is good to have the sig on top, so that they won't
> need to scroll through the entire message: How can you be sure that there is
> not anything else written below?
> Fixing this bug will only make it even harder to find the new information in
> a mail converation.

This is not the complete issue.  The issue I have is, though I prefer
bottom-posting, the large company I am a part of top-posts, and will miss my
conversations *every time* I bottom post (I've tested it).  Though I feel bottom
is better, I _cannot_ force thousands of my coworkers to change.

So I am required to cut and paste when I type any email.  Though this is minor,
it does slow me down.

If I understood C++, I'd fix it myself and publish the patches to other people
who have been eagerly awaiting a fix for 2+ years.  Open source rocks!  Too bad
I don't have the knowledge/time to help.

So does someone want to try?  The source is free and I'll bet Mozilla's license
allows you to publish modifications to source.  It would probably make thousands
of people happy.
See comment #168 and comment #170. 

I have downloaded the source and will be assembling it into Visual Studio some
time this week. Thanks to Dipa for the help provided...
I think the point is: you have to follow what happens in the real world...
people read the first few lines of an email and assume everything below is
attached text for referance only. It may be that bottom posting is what
"educated" people want everyone to do, but if you do, no-one will ever read what
you write! I'm afraid I'm on the side of going with the flow, give us the option
and let us set our own preferances, then you'll probably find most people who
want to get the message accross will be top posting.
When taking a stand for logical structure and standards compliance in things
related to computers and the Internet (as I do in many areas, from Web
development to domain name structure to e-mail format), I constantly face
comments like #180 that say that "the real world" is against me, so I need to
give up, give in, and join the rush to dumb everything down instead of trying to
uphold standards and logic against this tide.  However, I think "the real world"
is much like the mythical elephant being pawed by a bunch of blind men: your
perception of what it is depends very heavily on what parts of it you happen to
experience.  I took a look at recent messages in several Yahoo Groups I
subscribe to, to see the prevalence of top vs. bottom posting.  In one list
devoted to the internal politics of Mensa, of the messages I checked, there were
16 bottom posts, 2 top posts, and 5 with no quoting.  In another Mensa-related
list, about Web development, there were 12 top posts, 12 bottom posts, 5 with no
quoting, and 1 with quotes at both the top and the bottom (wishy-washy!).  In a
fan list about a musical performer, there were 11 top posts, 3 bottom posts, 18
with no quotes at all, and 1 in unreadable format (posted in HTML-only, which
was stripped in the digest version which I subscribe to).  My conclusion is that
if your "real world" centers on Mensa, you'll be used to bottom posting (but it
was a bit of a surprise to find the Mensa techies being evenly divided between
top and bottom posting while the Mensa politicos were overwhelmingly bottom
posters; I expected the opposite to be true), while if your world is full of fan
lists about singers, you might be less acquainted with that format (and
apparently less inclined to quote at all).  Others' worlds may vary.
Dan, as stated NUMEROUS times, this bug is more for E-MAIL and not newsgroup
posts. I hope your bugzilla priviledges are revoked for your insistence on
spamming this bug with lengthy and superfluous posts. :(
You can argue which is better until the end of time. The point is is your going
to give them the choice to top or bottom post, and I think you should, then you
should have their signature top or bottom post also. Why do it half way?
Yahoo Groups are e-mail lists, not newsgroups... so the formatting of messages
therein *is* relevant to e-mail format.
Dan what you say advocates choice. Personally I like top posting for a simple
reason: I read entrire threads in one session, therefore I know exactly what the
previous poster said and don't want to scroll down to find the current posters'

As far as I know the only argument for bottom posting is the signature stripping
issue: some archaic mail reader strip everything below the second signature
line. To me that's thes ame as advocating not using tables in html because some
old browsers don't render them properly.
It is funny how many people know what is best for me and how I should structure
the messages I write. Please stop these fruitless arguments. Let the users
choose whatever they prefer and leave your opinion to yourself. It is extremely
arrogant to assume that people should not be allowed to make bad choices (i.e.,
choices YOU think are bad). Most people have some sort of brains to make their
own judgement. If this bug is fixed, it will make Mozilla more user-friendly, it
will satisfy a lot of people who want to use the option and it will not restrict
those who hate top posting. So what's the fuss about?
  Sorry to bother you, but I'm too reading this thread for two years in hope
that somebody has posted a workaround for this bug. With every release I'm going
into prefernces and search for an option to put my signature before the quote. I
swear the god, this is THE ONLY THING keeping me to switch Mozilla/Netscape as
my primary mail client. 
   That's it, the rule of normal business correspondence is to put ALL of your
reply (including signature) ABOVE the quoted text. You can even think that M$
made this standart de-facto if you wish. But it is standart today. And every
modern e-mail client does it. So all this heated debate what "right" or "wrong"
is historic today. It just HAS to be implemented. Of course if we want to comply
basic user needs.
   P.S. I switched to Mozilla browser many friends of mine. But they also still
keep Mozilla Mail as primary news reader and use another products as main mail
Here's an interesting finding.  Apple Mail adds BOTH below the message (if you
are top posting) _AND_ at the bottom of the e-mail.  The result?  You have a
signature with your name and phone number signing your message, as well as
having it at the bottom to signify who sent each message.

  Again, the power should be with the user.
IMHO, there should be one more option in the preferences:
[] start my reply above the quoted text
[] start my reply bellow the quoted text
[] insert the quoted text at the end of my reply

This third option would put the signature with the reply, since the quoted text
would be at the end - like a foward to a email. 
  So this leaves the choice to the user. (I personnaly like my signature at the
end, and quoted text in the middle.) Everyone is then happy. The is probably a
few minutes fix/addition.

I'm blown away at how many developers in this thread think they can dictate
(even if it's only by omission) standard business communications etiquette. 
Congratulations on keeping Mozilla Mail a boutique application.  Glad to see you
don't want to be taken seriously in the business world.

I've voted for this bug, but I also want to reiterate here to the developers how
utterly ridiculous this discussion is for mainstream business users.  And that's
the market you have to target to gain widespread acceptance.
I like comment #190. Give the users the choice. There are great points for
bottom and top posting, thats not the issue here. The issue here is give the
user the choice as to where the signature goes. If people want to top post, then
most would want their sig. after their reply. I don't think this is a hard
solution to impliment. I think more time has spent been spent arguing about this
than would be needed to actully impliment a solution.
Hi jjongsma79 (comment #190),

   It is beyond my comprehension why the developers are being so
bone headed on this issue.  Give 'm hell!!

   A funny aside.  Another bug they won't fix:

On this one, the spell checker checks quoted text.  
This must drive the bottom posting, elitist, etiquette 
police OUT OF THEIR MINDS!   :-D :-P :-D

Please! No one contributing to Mozilla can be forced to do anything. Everyone
works on what he considers important. If you consider this bug important fix it
yourself or pay someone to do it. I also want to see this bug fixed but by just
saying it a hundred times it wont get done. This is the nature of Opensource.
Come up with a useful patch (like suggested in comment 190) and I'm sure
something can be done. But again: everything done in Mozilla is done on a
voluntary base, endless bitching wont help.
> ------- Additional Comments From  2003-07-04 23:59 -------
> Please! No one contributing to Mozilla can be forced to do anything. Everyone
> works on what he considers important. If you consider this bug important fix
> it yourself or pay someone to do it. I also want to see this bug fixed but by
> just saying it a hundred times it wont get done. This is the nature of 
> Opensource.

Well stated.  SOMEONE PLEASE post a patch.  YOU have the freedom to fix this.

Wish I could offer a patch; I'm no good at C-like programming, have no time to
learn, and can't pay someone.

A thought: if you are a developer and are adamant at not allowing top-posting
because you don't think is correct, at least provide the option but default to
bottom post... GENERALLY people leave a program's settings at their defaults
and more people will try the bottom style and perhaps you'll change the world. 
There's many people like me who prefer bottom but must use top at work (reasons
stated above).

Too bad I can't help nor change email clients... tried mutt but had
unresolvable IMAP issues, so I'm back.

Anyone read Gulliver's Travels?  How about Dr. Seuss?  These authors address
petty arguments like these.  "Crack your eggs on the big end!"  "No, the little
end is best!" -or- "Toast always falls to the ground on the buttered side!" 
"No, on the un-buttered side!"

Moral: Pick your battles wisely.

Anyway, if you're annoyed the developers aren't moving like you want them to,
don't wait around any longer.  Patch it or pay someone to patch it and publish
it to this list...

"GNU/Linux is free freedom" --Me
First off, apologies for the spam.

Second, could someone familiar with mail/news point us newbies who are
interested in fixing this bug, but who aren't familiar with mail/news, to the
approximate location of the code that does sig insertion?  Of course, we would
follow the Hacking Mozilla rules; it's just that it's hard to know where to start.
I think there's an issue with the patch from the users.  The techie community is
thinking standards are key and setting their own standards.  They are the ones
doing the bulk of the programming.  It is the managers and business community,
who doesn't have such skills, who want this problem fixed in order to gain
acceptance within the organization and make it work for them.  It will of course
increase the popularity and functionality of the software in business
environment.  Those who can fix the problem don't want to.  Those who want the
problem fixed can't.
> If you consider this bug important fix it yourself or pay someone to do it.

There's already $60 pledged via paypal in comments 160 and 162 above to the
person who fixes this bug.  Put your money where your mouth is, guys!
Its interesting. Im a developer and am very keen on getting this sorted myself.
I prefer the business way(top posting), and since I converted my wifes email
from Outlook (rraaaaaaaa!!) to mozilla I'm getting greif every day about this!
Poor me. She never complains about anything else in Mozilla mail except this. So
you see I HAVE to fix this or my life isnt worth living.

Standby! A fix or patch will come soon...
Please also post your upcoming patch to Thunderbird....

(yay- post #200!)
Personally, I think the developers are being anal retentive when it comes to
this. I had an all out war on the wish list forums for Mozilla and they said,
thats how the standard is defined. Well, just because something is a standard,
doesn't mean its not f@#king stupid. Just like someone said earlier here, this
fix would gain more acceptance in the business community. But because the
developers of Mozilla got a stick up their asses, they refuse to address this
issue. I am a professional developer by trade and even I realize this needs to
be fixed. Unfortunately I do not have the free time to spend on fixing this
It's the same here. For our business purposes it just makes more sense to reply
above the quoted message, so I and the user don't have to scroll all the way
through the old text in order to see the new text. I'm not sure about what kind
of standard the developers are talking about here. After all, who makes
standards anyway ? I really like mozilla but get aggrevated every day about this

Don't you all think Mozilla sets its own standards anyway ? 
I think people who want this bug as "wontfix" should think about 
this "motivation" of others to make it as always open. Don't you then think 
that it is a very (very 100x) important feature? 

Please just include it and let it as an option ("off" by default) and this bug 
discussion will be immediately closed (even this evening). Forget Outlook and 
stuffs reasons. Is it a very subjective one. 

Believe me, this bug will never be closed as long as it is considered 
as "wontfix". This will be the fact during the upcoming ten years.
Updating Summary and Whiteboard for better clarity:
OLD: Replying with the "start my reply above the quoted text" pref on should
prepend the signature above the quote text. (top)
NEW: The "Start my reply above the quoted text" setting should ALSO allow to
prepend the signature above the quoted text. (top)
Summary: Replying with the "start my reply above the quoted text" pref on should prepend the signature above the quote text. (top) → The "Start my reply above the quoted text" setting should ALSO allow to prepend the signature above the quote text. (top)
Whiteboard: Read comment 125 before commenting → PLEASE DO NOT SPAM THIS BUG! See Comments 70 & 74 for specs.
Assignee: ducarroz → ian
Attached patch Patch v0.5 (obsolete) — Splinter Review
This patch addresses both this bug and bug 141531, I have added three new per
account preferences for controlling how you want to quote in a reply and the
position of the signature when starting your reply above the quoted text. 

Currently the UI for the quoting preference sits a new retitled "Addressing &
Composition" panel and the UI for the signature position preference is on the
main panel for each account.

For quoting preference you can either use the global preference or have a
custom preference for that particular account. There is currently no global
preference for signature position but I'm not sure it needs it. The default for
the signature position is to force the signature to be at the very bottom of
your reply, unticking the checkbox lets the signature be positioned between
your reply and the quoted text.

At the moment "-- " is used to separate the signature from the rest of the
reply, as noted in other comments this may cause problems to certain mail/news
clients but should be fairly simple to fix if it needs to be.

There is currently no per account setting for Automatically quoting, again if
this is needed it can be added.

I've built and tested this on linux and it works on that, it will need testing
on other platforms.

The patch does still need some tidying up but any constructive feedback is
Attached image Account Settings Panel v0.5 (obsolete) —
How the main "Account Settings" panel looks with patch v0.5
Attached image Addressing and Composition Panel v0.5 (obsolete) —
How the "Addressing & Composition" Panel will look with patch v0.5
I think I read the entire bug but didn't see this addressed:

What's going to happen if the poster/emailer top posts with a standard delimited
signature complete with sig-dash above an existing reply and the next reader
sees nothing below the sig as is the "standard" behaviour??
I'm not sure we need the "Signature at bottom" checkbox. It is confusing (only
to me?) that this setting is in another screen from the "Start my reply
above/below" setting. Why? I would prefer it to be in the new "Addressing &
Composition" screen as a dropdown item, as described in comment #74.

Regarding Jay's question on what happens to quoted text when top post + top sig:
See comment #70: "IF top posts AND top signature are selected, THEN NO "-- "
delimeter to prevent the following text from being treated as a sig." I hope
this was implemented in Ian's patch.
The reasoning behind the location of the checkbox is that you only need to use
it when you are including a signature so it is on the same screen as that.

The "-- " issue hasn't been addressed yet (as noted in my comment#207), just
wanted some inital feedback on the patch as it stands now. I will look at
working up a new patch with the extra checkbox removed, having an extra option
to the quote preference drop down and removing the "-- " from the signature
above quoted text. Though instead of removing it do we want to use a different
delimiter in those circumstances? "== " perhaps? Depends what looks acceptable
and won't cause other problems.

Thank you Ian for picking this up and doing the patch.

I'm one of the mozilla mail users that promised money to the developer that
squashed this bug. 

I will happily honour my promise as soon as your patch is in the downloadable
binary for Windows & Linux. Any idea on which version that's likely to be? 1.5?

> do we want to use a different delimiter in those circumstances? "== " perhaps?

Sure. "== " looks good to me. "--- " would be my choice. It doesn't really
matter, as long as it's not: (1) some other official delimiter, (2) visually
similar to "-- ", and (3) unobstructive.
I suggest "__ " (double underbar then a space) as the delimiter.  It's the most
visually similar to "-- ", and "--- " has the problem that if some newsreaders
are misconfigured to use "\n--" rather than "-- " as the sig delimiter, they'll
get confused.
Answering comment #208, I think that the text following the checkmark is
confusing : a normal user will not understand why mozilla is asking him whether
the signature should go to the bottom, if mozilla doesn't precise when this
preference will be used ! Of course a signature should be appendend at the
bottom of a message, in normal circumstances ...

I suggest something like "Signature above the quoted text when replying", in a
shorter sentence (I admit this is not easy). As I'm not a native english
speaker, I may be wrong .
Maybe not a checkbox? maybe a radio button or list box? 
When replying:

Signature above quoted Text
Signature below quoted Text
No Signature 

This will prob take up more real estate on the preferences page. Just a thought.
The no signature is also just a thought, some clients (outlook) don't send a
signature on a reply because they assume the person already knows who you are
because your replying. This is probably true for buisiness clients, not always
when replying to a mailling list.

>The reasoning behind the location of the checkbox is that you only need to use
>it when you are including a signature so it is on the same screen as that.

A minor detail: if this structure is kept, the "signature at bottom" checkbox 
should be enabled only if the "Attach the signature" one is also so.

Anyway, thank you for working on this patch.


> I suggest "__ " [...] if some newsreaders are misconfigured to use "\n--" [...]

I believe that the spec is "\n-- \n", so this supposed newsreader would have to
be doubly confused (i.e., ignore the space AND the hard return). Nevertheless,
we shouldn't be taking virtual broken newsreaders into our considerations. ;)
> A minor detail: if this structure is kept, the "signature at bottom" checkbox 
> should be enabled only if the "Attach the signature" one is also so.

This is how the patch does work.

wrt broken mail/news clients I can see the posibility that they might just check
for "--". The only problem I might see with "__ " is whether it exists in all
character sets as "_", I suspect x5f might display different under certain
locales. Thoughts?
Suggest changing option label to "...prepend the signature above the quoted
text..." for clarity reasons only...
Will this patch show up in Thunderbird?
The suggestion from comment#74 was:

Then, [ Start my reply above quoted text (signature directly below reply) |\/]
      | Start my reply above quoted text (signature below entire reply)   |
      | Start my reply above quoted text (signature directly below reply) |<-NEW
      | Start my reply below quoted text                                  |
      | Select the quoted text                                            |

Now this is too wide to fit in the prefs pane, so I thought perhaps this:

Then, [ Start my reply above quoted text (signature at bottom) |\/]
      | Start my reply above quoted text (signature by reply)  |
      | Start my reply above quoted text (signature at bottom) |<-NEW
      | Start my reply below quoted text                       |
      | Select the quoted text                                 |

but even this is too big unless I move "Then," up to the end of the line before.
That doesn't look too good either.
If "Then," stays to the left of the drop down there are about 50 characters
available for a meaningful description. If "Then," can be (re)moved there is
room for about 56-58 characters. Any thoughts?
UI text/choices are fine, but I think the issue of what to do about delimiting
or not delimiting or how to delimit the top-posted signature above replies needs
to be finalized first and foremost. Menu choices are quickly going to get out of
hand simply because you're going to have to make two choices for every choice so
to speak.

Reply above, include signature with delimiter
Reply above, include signature without delimiter
Reply above, include signature at bottom 

So forth and so on ...
Shorter version (using a radio button):

Then, (o) Start my reply:
          [ below quoted text                    |\/] <-- default?
          | above quoted text (signature at bottom) |
          | above quoted text (signature by reply)  | <-- NEW
          | below quoted text                       |
      ( )   Select the quoted text
I like comment #225, except that I like "(signature below reply)" better than I
do "(signature by reply)".

Also, regarding the delimiter problem, perhaps you could have a field where the
user could edit a custom delimeter (with a sanity check against "-- " when sig
is not at bottom) or have none at all.
Peter: Shortened menu looks ok, but what do you suggest regards the delimiter
issue insofar as a menu option(s)?
Thanks Jay ;-)

This is with just a simple rewrite of the drop down choices:

[x] Automatically quote the original message when replying
    Then, [ reply above, include signature with no delimiter |\/]
          | reply above, include signature with no delimiter |<-NEW
          | reply above, include signature with delimiter    |<-NEW
          | reply above, include signature at bottom         |
          | reply below                                      |
          | select the quoted text                           |

This one rewrites more than just the dropdowns but makes signatures appear
optional and doesn't mention delimiters:

[x] When replying, automatically quote the original message
    which should be [ between my reply and any signature |\/]
                    | between my reply and any signature |
                    | above my reply and any signature   |
                    | below my reply and any signature   |<-NEW
                    | selected                           |

Thoughts? (and I'm getting a headache from all these collisions!)
WRT comments 207, 212, 214, 215, 220, and 224:

The actual solution to the sig-dash issue is surprisingly simple:  excise it


1. You can't use \n-- \n anyway.
2. \n--\n is problematic because of problems with other mail/news clients.
3. Inventing a new sig delimiter is pro'ly a bad precedent, especially in a
client that has aligned itself so strongly with existing web standards.

Further, the existence of a signature /as such/ is predicated on its being
preceded by a delimiter so that clients that understand it will strip it upon
reply (whilst other clients just mangle it :-( ).

That is, the delimiter itself is an integral part of the sig and how it should
be understood.  But what you're trying to accomplish here is simply to create a
kind of template with some boilerplate.  That's it.  There /is/ no need for a
sig-dash, since nothing's being delimited -- so why include one at all?

I actually did try this in OE -- set up a sig and replied to a message -- and it
does not by default include a sig-dash; it just sets it in as boilerplate.

Ditching the very idea of a non-delimiting sig-dash entirely will also likely
reduce the shrieks of outrage from certain quarters of the Mozillaverse.  The
moment this goes in -- in /any/ form -- , you /know/ there's going to be a bug
filed to back out the patch.  I would suggest that you stand a better chance of
keeping the patch in if you're not breaking or inventing sig delimiters.

Your one major obstacle is the UI:  ideally, this inserted boilerplate would be
distiguished from a signature as such; however, to try to do so would be a UI
nightmare.  But I suspect you can carry that argument if you're able to argue
that this factitious 'top-sig' isn't really a sig and that you haven't treated
it as such (by linking it to a delimiter).

Of course, someone could still decided to insert \n-- \n at the beginning of
their 'sig' file.  Of course, it'd look stupid with a properly delimited sig and
wouldn't be very effective as a way of preceding your boilerplate.

Note, too, that, AFAIK, this affects /only/ plain-text messages.  I'd like to
see sig-stripping behaviour extended to HTML messages, but as it stands, it doesn't.
Jay & Ian, the user doesn't care about the delimeter, and therefore doesn't need
to see, nor make a choice about it (Please re-read comment #70 & comment #211).
The delimeter will be placed (or not placed) based solely on the selection the
user makes in the UI I proposed in comment #225. There is no need for the user
to see (let alone understand) the word "delimeter", ever. So to make it explicit:

above quoted text (signature at bottom) <-- use delimeter
above quoted text (signature by reply)  <-- don't use delimeter
below quoted text (signature at bottom) <-- use delimeter
Select the quoted text                  <-- use delimeter

BTW: Ian, your second suggested UI in the last comment had my head spinning. I
think it is too convoluted.
wrt to comment#225 the problem with using radio buttons is that when it comes to
a per account preference a radio button is already being used to select if
you're using the global or custom quoting prefs.

Ouch, two more collisions on the way!
Peter: The only problem I have with non-delimited, top-posted sigs is if there
is no mechanism to strip the sig, we're going to have multiple replies with
multiple sigs which defeats the purpose of a delimited sig in the first place.

Getting my spare asbestos suit cleaned now in anticipation of posts in the
groups when/if this becomes active. :-(
It would be fairly easy to put a 
[x] Don't delimit signature

on each account's main panel
(where [x] Signature at bottom is sitting in patch v0.5)

This would let the user control, on a per account basis, whether the signature
is delimited or not.
>This would let the user control, on a per account basis, whether the signature
>is delimited or not.

...except that Real Live Business Users don't understand the purpose of a
delimiter anyway.  To tell you the truth, I was slightly confused by Mozilla
greying out other people's sigs when they were "properly" delimited.  A
preference will, IMHO, just confuse people.

I think that the solution proposed in comment 230 is the best one so far.
Dan: Yes, comment 230 addresses the UI choices but we're taking a step backwards
after having come this far with delimited sigs. Now that we're (sic) going to
allow undelimited sigs at the top, above replies, we're now going to see these
sigs that included every known company these people have ever worked for as well
as tons of useless info to look at over and over and over in *every* subsequent
reply. Carry on gents, good luck.
The suggested layout in comment#230 and in turn comment#225 means the use of
radio buttons and this would has several issues:
1. Radio buttons within radio buttons in the UI when having custom quoting
preferences on a per account basis
2. Not enough real estate in the Addressing pane UI of preferences
I suggest not having any elaborate preferences regarding sig delimiters... the
sort of people who care about the proper use of delimiters probably aren't the
sort who top-post anyway, and most top-posters don't know or care what a
delimiter is... thus, the delimiter style should probably be set up by the
developers to conform to the way those of us who *do* care want it: with a
proper delimiter for bottom-posted signatures, and without one when the
signature is up in the middle.  The only question then is whether to have
nothing resembling a delimiter in this case, or, for aesthetic purposes, having
a pseudo-delimiter like "==" that won't be treated like a real one by programs
that recognize them.

The other suggestion, of having some sort of delimiter so that the signature can
be snipped out on quoting, would require *two* delimiters -- a "start sig" and
an "end sig" delimiter, and no standard or tradition provides any such thing. 
If Mozilla uses one, they'd have to be trying to invent a new standard.  But I'm
not sure it would really make much sense to snip out a signature from the middle
of a message while quoting stuff both above and below it; if you're keeping a
humongeous fullquote of all messages in the thread as ballast weighing down your
reply, you should probably leave the signatures in place, while if you're making
some effort to trim your replies, you can cut them out by hand along with
everything else you trim.  At least, with signatures stuck in the middle, you
don't build up an ever-lengthening trail of garbage at the very end (unless your
message passes through a server that puts in ads or disclaimers there).
One possible way of avoiding adding to the UI in the global addressing is to
only allow the option for signatures to be between the reply and the quoted text
on a per account basis and not have it as a global option.

It's possible I will get the part of the patch for bug 141531 reviewed, etc and
checked in before tackling the signature/delimiter stuff. It might give time for
a solution, as far as the UI is concerned, to be arrived at.
I tend to agree with comment 237.  Really, we're not talking about a sig as such
at all; we're talking about boilerplate.  We'd actually find ourselves in the
position of having three delimiters:  the standard delimiter for signatures, a
starting delimiter for boilerplate (which can't be \n-- \n or \n--\n), /and/ a
delimiter to indicate the end of such boilerplate so that it can be cut out by a
conforming client (which, for the time being, at least, would be Mozilla and
Moz-derived clients).

Further, Peter's comments elsewhere indicate that some business users have an
un-Godly fear of anything in an ever-lengthening train of correspondence being
changed -- which would include such signature boilerplate (since some users are
apparently incapable of figuring out levels of quoting and have to look at the
bottom of a what they're reading to figure out who wrote it :-( ).

Jay -- In re. comment 235:  comment 234 actually gives a partial answer to your
concern.  The fact is that most people don't get the point of the delimiter. 
Sad, but true.  It would seem that the only solution would be to have beginning
and ending delimiters for such boilerplate . . . which is problematic in and of
itself (at least at this time).

Now, if (in re. comment 232) you could send that spare asbestos suit up
this-a-way, I'd be muchly appreciative. :-)
If this is gonna make it into Thunderbird, signature UI should be kept as simple
as possible. Summarizing, we have 3 options:

a) Automatically inserting a delimiter only with bottom signatures, as Dan
Tobias suggested. But there's no need to automatically insert pseudo-delimiters
for top posters, users can  define their own if they like. Otherwise, imagine
how many would protest for the right to define their own aesthetical delimiters.
b) Do not automatically insert delimiters at all and let users define their own,
a la OE. That would be the ultimate solution but, I'm afraid, bottom posters
would block it immediately.
c) Ian's comment 233 about having a simple delimiter exclusion option, on
account basis.

The simplest and less conflicting decision would be a) but it might be confusing
because of inconsistent behaviour. Having bug 141531 fixed, option c) would also
be affordable.
Ian: If we can agree that there should be simple-to-understand options (without
mentioning delimeters, as suggested in comment #230), then the only issue is to
find a way of phrasing/ordering the UI to address your comment #236. Personally,
I see little problem with using a radio button within a radio button (as
suggested in comment #225).

Jay: Yes, if top-reply *and* top-sig, then the sigs would *not* be stripped
(until there is a spec - initiated by Mozilla or whoever - for an "end
delimeter"), since adding "-- " would cause the quoted message(s) below the sig
to be stripped too (which is not desired).

For now, I think in the case of top-reply + top-sig, we should probably leave
out any delimeter. Although I think it would be great if we could establish a
new standard for an end-delimeter, this is probably not the time nor the bug to
do it in.
Dimitrios: I don't think we should have a "delimiter exclusion option" (your
option C and Ian's comment #233) as that would allow the user to disable the
delimeter in all situations (even those where there is no good reason to do so).
Not to mention that most users will be confused by the term "delimeter" and that
it would needlessly clutter the UI. I still think comment #225 is the best way
to go.

Ian: Could you put together some screenshots showing how comment #225 would look?
How about:

| [x] Automatically Quote the original message when replying        |
|        ______________________________________                     |
| Then, [ Start my reply below quoted text  |\/]                    |
|       +--------------------------------------+                    |
|       | Start my reply above quoted text     |                    | <-- (A)
|       | Start my reply below quoted text     |                    | <-- (B)
|       | Select the quoted text               |                    | <-- (C)
|       +--------------------------------------+                    |
|         ( ) Place my signature below the quoted text (recommended)| <-- (I)
|         ( ) Place my signature below my reply (above quoted text) | <-- (II)

I  - inserts the delimeter ("-- ")
II - does not insert the delimeter

If A is selected, then I & II are enabled (default: I)
If B is selected, then I & II are grayed-out (delimeter is used)
If C is selected, then I & II are grayed-out (delimeter is used)

Note: B would be the default selection for new accounts.

PS. I realize we would still have a radio button within a radio button in the
per-account "Addressing & Composition" preference, but I believe this would be
OK. We would also not be setting a precedence, as this already exists (e.g.,
under "Advanced | Proxies").

PPS. This solution only seems necessary, because of potential horizontal space
problems with the solution suggested in comment #74 (I still think a rewording
of that solution might be possible - any takers?).
re: comment#243
This is a possibility in the mail/news account settings but there is not the
vertical space to do this in the mail/news preference settings, hence my
suggestion in comment#238

I'll do some mockups for some of the suggestions.
I am one of those business folk who need top posting with sig. (while wearing
the popular asbestos suit).  Let us not make this more complex than needed:
  I want to top post, with sig following my top post.
  When business, eg top posting, I do not want/care to have prior sigs trimmed.
  When bottom posting, in tech newsgroups and mailing lists, I do want the sig
  Yes my business contacts have no clue what the sig delimiter is and whould
only get even more confused by 3 delimiters.  

My request/recommendation:  use delimiter for bottom sigs and ignore the
delimiter  for non-bottom sigs.  KISS
Hehe, glad I didnt get too far with my fix - nice lively discussion here.

As a mostly business user I have to agree with williams comment #245. Not
bothered about top posting removing previous sig's, and as for the delimiter, in
the words of my wife, "whats all these funny dashes for?". I doubt any top
posters care about signature trimming - I dont, my wife doesnt, anyone else
bothered about the trimming?


Ian, thanks for taking this hot spud!
Re. Comment #238: Although that would solve the "technical" dilemma (limite
horizontal or vertical space), I think it would be too inconsistent and
confusing to the user. I still think the UI suggested in comment #74 (with a
slight rewording to reduce horizontal space) would be best.
Attached image Mock up for comment 243 (obsolete) —
Mock up of Composition preferences panel as suggested in comment#243
Now that I see it, it's not as terribly complicated as I was thinking. Peter,
Brian, Dan and (most of all) Ian thanks for contributing to this bug!
A very minor suggestion on the mockup for comment 243: change the comma after
"Then," to a colon, or eliminate it completely.  A colon would be consistent
with the "Windows" and "Message Display" preferences, for example, where one
sees "When opening messages, display them in:" and "When viewing plain text
Attached image Another mockup for comment 243 (obsolete) —
This mockup for comment#243 uses a drop down instead of radio buttons
While you are at it; How about an option for NOT including a signature at all
when replying?
" How about an option for NOT including a signature at all
when replying?" : see comment #71
Wow, excellent work Ian! I never thought I'd see the day... :-D

One minor nit with the two recent screenshots: The secondary dropdown list
(which I prefer) or the radio buttons (whichever we end up using) should be
slightly *indented* from the "Then," because is a _sub-function_ of the line
above, and not equivalent in the decision hierarchy. (yeah, I know. We're
getting into horizontal space problems again...)

PS. Still no suggestions for rewording to get all settings into one dropdown
(comment #74)? Ian, would you be willing to make a screenshot implementing
comment #74, so we can see how much space-saving we would need?
Attached image Revised first mockup for comment 243 (obsolete) —
Revised first mockup for comment#243 with radio buttons indented further.
Attachment #128775 - Attachment is obsolete: true
Attached image Initial mockup for comment 74 (obsolete) —
Initial mockup for comment#74
One thing we have to remember for all these UI changes is that localisations
need to be able to fit the extra wording into the UI in their languages -
thanks to timeless for reminding me of that.
Attached image Initial mockup for comment 228 (obsolete) —
Initial mockup for second suggestion of comment#228
We shouldn't actually mention signatures in the mail/news composition
preferences because signatures are done on a per account basis. So maybe we
should concentrate on the per account composition UI and where the settings for
how we insert signatures there.
I agree with comment #258 ... possibly use a check box 
Ian: In Re: Comment #258

You shouldn't even include the reply choices in the composition UI at all as
that will be redundant in the account settings. Include *all* reply choices on a
per account basis only. What you're trying to accomplish in the first place is
moving away from a global setting.

Absolutely agree with Jay in comment 260.
Do we want to go down the route of moving the whole of the "Forwarding and
Replying to Messages" box to a per account basis or just parts of it?
Is there a need to move the "Forward messages:" part to be per account? The rest
I can see quite happily sitting at an account level.

Really this decision process should be moved across to bug 141531 as that is
it's bug.
In re. comment 260:  I believe the idea was that there would be a global setting
with the option to over-ride on a per-account basis (/à la/ LDAP).  I suspect,
however, that this is just gonna confuse the hell out of the much-maligned
business user. :-( Perhaps, then, shifting it entirely to individual accounts
would be /à propos/?

In re. attachment 128797 [details]:  This sucks horribly. :-( Well, OK, it's more accurate
to say that it's just very confusing.  The other two possibilities (either radio
buttons or a drop-down) are clearer.

As to comment 253, shouldn't this be dealt with by way of the per-account
setting to use a sig?
Actually I didn't think that 128797 was that bad, although I think my preference
was for 128794.
Blocks: bms
In re. comment 264:  The problem is that it's not at all clear; after all, why
on earth would I care about where the quoted text to which I'm replying is
place? -- I'm concerned about whether I'm top-/bottom-posting or not and whether
my boilerplate (whether a properly delimited sig or otherwise) is located.

My personal preference is for the drop-down; it saves a bit of space, which is
an issue to some.

One other thing just occurs to me, however:  I realise that this was s'posed to
be set up as something analogous to LDAP settings, where there is a global
setting that can be over-ridden on a per-account basis -- but this doesn't seem
to be doing this -- at least not in an efficient way.

The way it's currently set up, you have the global pref, but the per-account
over-ride is limited to a check-box regarding a sig (attachment 128662 [details]).  This
really should be switched over entirely to individual accounts; that is, you can
determine whether to position the cursor for top- or bottom-posts or simply to
select the text on per-account basis and can furthermose decide whether you want
a proper sig or some inserted boilerplate on a per-account basis.

To wit:  With some accounts, there is a strongly likelihood that I will either
top-post (some e-mail accounts) or bottom-post (some NGs); at other times,
that's not quite so determinate (as in servers where some NGs prefer one form of
posting and others the other).  To have a global pref for this accompishes nothing.

Once you've moved those settings over to individual accounts, the sig issue just
becomes an item that follows from the per-account setting for sigs:  If you
select sig text, you have to decide whether to use it as a proper(ly delimited)
sig or as after-reply boilerplate.

Sorry if this suggestion kinda screws things up from a UI perspective, but I
just can't see it being truly functional with the UI that's been proposed so far.

/me wonders why we can't just resolve this bug as WONTFIX and go back to the
drawing-board on this whole issue. . . . :-(

(That was a rhetorical comment; I'm not actually expecting responses to it. . . .)
Re. The question whether to _only_ use per-account settings: Since quote-style
is what users are likely to want to customize on a per-account basis (especially
NG vs. e-mail), as opposed to forwarding style (which will likely be the same
globally), I suggest to move (not copy) the "reply/quote" preference to the
accounts, and to keep the forwarding preference (excliusively) in the
"Composition" dialog.

A benefit of not having a "global" setting will be that we lose the "use my
global settings" radio buttons, and thus could more easily (i.e., have enough
horizontal space to) implement the UI suggested in attachment 128783 [details].

This will have the additional benefit that users would have to change _each_
account separately away from the default setting (bottom-reply, delimited sig at
bottom), thus reducing the chances of this occurring. ;)
*** Bug 214514 has been marked as a duplicate of this bug. ***
this patch (which could be cleaned up a little) does the simple thing of making
the sig come before the quoted text when reply on top is set. I'm not convinced
we need anything more complicated
Re. Comment #267: We really do need the ability to top post with *either* sig at
top and at bottom. This current suggestion would really **** off newsgroups that
top post (e.g., many macintosh NG's), since now the sigs would also always be at
the top and also no longer trimmed. It would also upset e-mailers that prefer to
top reply, but still want their sig at bottom. Additionally, some (like me) may
have their prefs to top-reply (with sig at bottom) so they can begin reading and
*responding* to the post beginning at the top and interspersing their replies as
they work their way *down* the post. I therfore think limiting top posting to
only have top sig (and not bottom sig) would be a bad idea. Please reconsider.

BTW, the screenshots in attachment 128902 [details] and attachment 128901 [details] (posted in bug
141531) look really good, but now unfortunately lack the ability to select
whether to have sig on bottom or on top on top-repies.
Peter: that is because bug 141531 is just to do with the moving of the current
quote settings from a global preference to a per account preference. Once that
is working and checked in, we can then look at how best to implement the
signature (and delimiter if need be) issues.
Instead of going to general prefs or something like that, why not just make the 

The "Attach signature" signature checkbox in the per account setting will be 
replaced by two checkboxes:

[ ] Attach this signature under my reply
______________________________   ___________
|                             |  | Choose...
-------------------------------  -----------

[ ] Attach this signature at the end of my e-mail
______________________________   ___________
|                             |  | Choose...
-------------------------------  -----------

With this (per account basis) solution, 
- users will always able to choose between a top and bottom signature and (the 
best), independently of the case that there is a quoted text (due to reply, 
forwarding, etc.) or not. 
- they can also choose both sigs if they like. In this case, the two sigs 
content syntax is generally somewhat different. For example, I've already seen 
an e-mail with something like this:

<e-mail text>

Best regards,
John Smith (<--- end of the top sig)
John Smith (PhD.)
Biological Department of ....
tel... (<--- end of the bottom sig)

For some people, they can also set the "John Smith (PhD.) ..." as their top sig 

Perhaps have I missed some previous discussions or comment about the technical 
constrainsts involved by this proposition (norm, etc.) but I think it will make 
everybody glad. 

this fix should *only* apply to mail -- *not* newsgroups..... this was the
reason of so much depate.... the logic should be something like:

   Allow prepend of signature above the quote text. 
   Deny prepend of signature above quoted text.
I would suggest that the UI should have the preference for placement of
signature (and delimiter if need be) in the same pane as the preference that is
used to say whether a signature is attached and where to find that signature. To
me that either means moving the signature preferences to the composition
preference panel or making room on the main pane.

Now in bug 141531 I've already suggested that the HTML composition preference is
moved to the new "Addressing & Composition" pane, that leaves the only other
logical thing to be moved as being the "Advanced..." button. I would suggest
that this belongs better, slightly relabelled, in the "Server Settings" pane. It
should sit quite nicely alongside the local directory text box and with a label
of "Advanced SMTP".

Wherever the signature position preference goes, it needs to reflect (and be
enabled/disabled) the fact that the quoting/signature attachment checkboxes may
or may not be ticked i.e. they can explictly say there's definitely a signature
or quoted text as this may be confusing to the user. If sits with the signature
preferences it will need to be non-commital about quoted text and if sits with
the quoted preferences vice-versa.

Any thoughts?
I think a lot of these concerns would be alleviated by making the signature
optional for forwards and replies.  I just switched from Microsoft Lookout - er,
Outlook -  and there was a feature there that let users omit the signature
entirely when replying or forwarding.  So you could have an option like this:

 When forwarding or replying, append signature ...
 - above quoted text
 - below quoted text
 - not at all - it just takes up space anyway

Threads can sometimes get a little clogged with signatures.  This might help
squash them, while appealing to both sides of the debate as well.
Re. Comment #274 : This bug is - err - "complicated" enough. Please file a
separate RFE for the "no sig" option.
Adding bug 109162 is a blocker, one that bug is fixed advanced button will no
longer be in Identity pane.
Depends on: 109162
*** Bug 214514 has been marked as a duplicate of this bug. ***
Attached image Mockup of delimiter UI
The delimiter option would be diabled/grayed out if the attach signature is not
Attachment #128662 - Attachment is obsolete: true
Assuming patch to bug 141531 is checked in UI for signature position now needs
to be in account settings rather than mailnews preferences, this is a mockup of
that UI as per comment#243
Attachment #128663 - Attachment is obsolete: true
Attachment #128783 - Attachment is obsolete: true
Attachment #128794 - Attachment is obsolete: true
Attachment #128795 - Attachment is obsolete: true
Attachment #128797 - Attachment is obsolete: true
I hope this hasn't been discussed yet, but I suggest that the signature
delimiter (--) should be *turned on* by default. After all, this is the "more
standard" behaviour.
Sorry if it's been told before, but : will it be possible to decide for each
message if a signature will be included or not il a particular message. This
could "easily" be done, in terms of UI, with a "Sign this message" or "Attach my
signature to this message" (by default it should be positionned to the value in
the preferences, so that one doesn't have to press it every time).

Of course, the choice for the position of the signature should remain in the
preferences only, like it's been suggested.
Having a visible _preference_ for using a sig delimiter seems like just a _wee
bit_ overkill to me. Although we're not Firebird throwing any and all prefs out
of the window, there should be some borders, and I think that checkbox is firmly
on the other side of them.
Simply do the Right Thing (TM) and don't bother the user with it.
Re. Comment #279: I don't think we should give the option to place a delimiter
or not, for two reasons: 

(A) Most users will not know what it is (confusion) and how important/useful it
is (This is exacerbated by its location in the *top* account prefs panel - which
should be reserved for required/important settings), and 

(B) the problem is solvable with logic, as outlined in comment #243 (and updated

| [x] Automatically Quote the original message when replying                |
|        ______________________________________                             |
| Then, [ Start my reply below quoted text  |\/]                            |
|       +--------------------------------------+                            |
|       | Start my reply above quoted text     |   <-- (A)                  |
|       | Start my reply below quoted text     |   <-- (B)                  |
|       | Select the quoted text               |   <-- (C)                  |
|       +--------------------------------------+                            |
|                                 _______________________________________   |
|         and place my signature [ below the quoted text (recommended)   ]  |
|                                +---------------------------------------+  |
|                        (I) --> | below my reply (above the quoted text)|  |
|                       (II) --> | below the quoted text (recommended)   |  |
|                                +---------------------------------------+  |

Delimiter Used:     A(II)

Delimiter Not Used: A(I)

NOTE: The wording of the preferences should be optimized for the user's
perspective, and not with what is going on "under the hood".

PS. Note the *order* of A/B/C and I/II to have the above/below settings visually
coherent with the order in which the settings are listed. ;)
Re. Coment #281: that's bug 73567#c22 (which I hope will also be fixed someday)
I think that what is proposed in Comment #279 is a good idea (i.e. explicitly 
asking to set a delimiter or not). Indeed, how is it obvious for a simple user 
to know that if the reply is above the quoted text, there is automatically no 
delimiter? And what if this user wants a top sig but with a delimiter?

I think only that the word "do not use a delimiter ....with my signature" 
should be replaced by something like "set a delimiter between my e-mail and the 
signature". The idea is the same than the former but it is more clearer for a 
simple user. 

Finally, for a space reason, why not put this checkbox on the right of 
the "Attach this signature" one, instead of below?

Re. comment #283: oops, this was in response to comment 278, not comment 279. :-[

Re. Comment #285: Because the user doesn't (and shouldn't) care about the
delimiter. Mozilla should place a delimiter whenever possible. We also probably
shouldn't allow to place a delimiter on a top-sig because it would cause the
*recipient's* program to then strip the quoted text. We shouldn't be allowing
one user to cause "breakage" in another user's mailer. Additionally, these would
be _edge cases_ that don't justify confronting the user with jet another (in
this case, confusing) UI pref (I can't even think of why anyone would want
top-sig with a delimiter).
Re: comment#280
That is the reason why the preference is for "Do not use..." rather than "Use..."

Re: comment#282
So hidden preferences for control of the delimiter would be useful? Perhaps:
0 (default) - follow standard behaviour
1           - always use a delimiter
2           - never use a delimiter

Re: comment#283 (A)
The logical place for the UI for the setting of the delimiter, if there is one,
is by the UI for where the signature is set.

I'm also not sure how easy the second drop down would be to understand for the
user wrt to when/how it is enabled. Perhaps we need to change the leadup text to
read "When starting my reply above quoted text place my signature..."
> Re: comment#283 (A)
> The logical place for the UI for the setting of the delimiter, if there 
> is one, is by the UI for where the signature is set.

True. But the point is that there schouldn't be a visible pref for the delimiter

> I'm also not sure how easy the second drop down would be to understand for 
> the user wrt to when/how it is enabled. Perhaps we need to change the 
> leadup text to read "When starting my reply above quoted text place my 
> signature..."

I believe that that text and the 2nd dropdown should be *grayed-out* unless (A)
is selected. That would make it quite clear, IMO. That would also save us from
using such a long sentence.

PS. I think it is time to check something in and let the community use it for a
while. That will help hone any rough edges, if there are any. In the same veign,
I don't think this is the bug to decide on hidden prefs, as the ability to
remove the delimiter entirely will surely enrage many (and I would agree with
them), and could thus hinder this bug as a backlash (definetely something I
don't want).
>I believe that that text and the 2nd dropdown should be *grayed-out* unless (A)
>is selected. That would make it quite clear, IMO. That would also save us from
>using such a long sentence.

The problem is the signature position preference also has to be greyed out when
no signature is selected to be attached no matter what the quoting preference is
set to, hence my suggestion.

>PS. I think it is time to check something in and let the community use it for a
while. That will help hone any rough edges, if there are any.

Nothing can really be checked until at least the patch to fix bug 141531 is
checked in, though I am working on patch assuming that bug 141531 is fixed.

>In the same veign,
>I don't think this is the bug to decide on hidden prefs, as the ability to
>remove the delimiter entirely will surely enrage many (and I would agree with
>them), and could thus hinder this bug as a backlash (definetely something I
>don't want).

Whatever is introduced as far as delimiter behaviour is bound to enrage certain
groups. What I'm proposing is, for those that want to, the delimiter behaviour
can be overridden by means of a hidden pref.
Can anyone tell me the bugid that tracks disabling the sig stripping in the
first place?

It seems that for many people at least, the biggest problem with anything to do
with signatures is the inadvertant stripping of real mail.

Who cares why or how, hell, I just have to use "-- "as a bullet and then move on
and the rest of mail gets junked on reply.

Bug 9202 mentions that this pref was originally meant to be committed, but it
seems to have slipped through.  Is anyone aware of any new bug for this?
Re. Comment #289: (A) The setting will be clear once "Automatically Quote" is
activated, before that, the purported confusion is negligeable (IMO
non-existant). User will see that it is grayed-out until a particular item is
selected above it. That is the normal way deactivated sub-options work. I see no
source of confusion here. (B) The "hidden pref should be filed as a *separate
bug* (Reason: My suspicion is that it will be marked "wontfix", and I don't want
that controvery to affect this controversy ;) ). 

Re. OT Comment #290: There will be no "inadvertent stripping of real mail" when
this bug is fixed, as there will be no delimiter when top-reply+top-sig.
Whiteboard: PLEASE DO NOT SPAM THIS BUG! See Comments 70 & 74 for specs. → PLEASE DO NOT SPAM THIS BUG, see comment 125
Comment on attachment 128904 [details] [diff] [review]
if reply on top is set, make sig come before the quoted text

David Bienvenu, if I read the code correctly, your patch would introduce
exactly the bug I mentioned in comment 50 and comment 135.

I guess they went under under all the useless comments. I am thus obsoleting
the patch, so that *this* comment doesn't go under - if you (David) disagree,
you can still undo that.

Apart from that, I'm not sure that always attaching the sig before the quote
for top-posters is a good idea either, for the reasons mentioned (sigs
accumulating in the middle, not the end; the quote is hopefully more important
than a long sig with address etc.).
Attachment #128904 - Attachment is obsolete: true
Comment on attachment 128904 [details] [diff] [review]
if reply on top is set, make sig come before the quoted text

I wrote:
> your patch would introduce
exactly the bug I
> mentioned in comment 50 and comment 135.

I *think* you could address this one problem by just something like
firstFourChars.Equals(NS_LITERAL_STRING("-- \r")) || GetReplyOnTop() == 1));
around line 3400 (ProcessSignature())
Re comment #292: Comment #283 (and many others) already addresses comment #50
and comment #135. It states that "if top-reply AND top-sig, then *no
delimiter*". So there will *never* be the case when a sig-stripping program
(like mozilla) will inadvertently also strip the quoted portion on a reply
(because there will *not* be a delimiter in those cases).

Ian, now that you fixed bug 141531 (thanks!), any chance you will be submitting
your patch for this bug too?
Whiteboard: PLEASE DO NOT SPAM THIS BUG, see comment 125 → PLEASE DO NOT SPAM THIS BUG. See comment 283 for specs.
> Re comment #292: Comment #283 (and many others) already addresses comment #50

Eh? The *patch* doesn't, and I commented about *that*.
Attached patch Patch v0.6b (obsolete) — Splinter Review
This patch implements both the UI as per attachment#129221 [details] and the backend to
support it. If the signature goes above the quoted text then the delimiter is
not added to the signature.
Attachment #128658 - Attachment is obsolete: true
Attachment #130103 - Flags: review?(
Blocks: 216728
Comment on attachment 130103 [details] [diff] [review]
Patch v0.6b

>+pref("mail.identity.default.sig_position", 0); // 0=below quoted 1=above quoted
IMHO this should be a boolean pref.

>     if (!(firstFourChars.Equals(NS_LITERAL_STRING("-- \n")) ||
>           firstFourChars.Equals(NS_LITERAL_STRING("-- \r"))))
>     {
>-      sigOutput.AppendWithConversion(dashes);
>+      if (reply_on_top != 1 || sig_position == 0)
>+        sigOutput.AppendWithConversion(dashes);
>       if (!m_composeHTML || !htmlSig)
>         sigOutput.AppendWithConversion(CRLF);
>       else if (m_composeHTML)
>         sigOutput.AppendWithConversion(htmlBreak);
>     }
Can't see this working - you'll get an extra break when you sign on top.

>+  <checkbox hidden="true" wsm_persist="true" id="identity.attachSignature"
>+        prefattribute="value" prefstring="mail.identity.%identitykey%.attach_signature"/>

There's one big oversight here, I'm afraid. When you compose a message, you
have a choice of identities. Each identity has its own signature. You can
change the selected identity, at which point the previous signature is deleted
and the new one is inserted - but only if it is at the bottom; it should be
possible to replace the signature in HTML compose because it's wrapped in a DIV
that identifies it, but I can't see how you can delete a top signature in plain
text compose without wrapping it in the same sort of way - a bit like the way
that plain text quotes are blue (well mine are, not that I use plain text
compose). You would also have to locate the quote so that you can paste in a
top sig when necessary. See nsMsgCompose::SetSignature for the gory details.
Attachment #130103 - Flags: review?( → review-
Could the signature be "wrapped" (say in: <moz-sig> bye </moz-sig>), so
switching identities would allow mozilla to know where the sig is (and swap it
around, as needed)? This wrapping (if it shouldn't/can't be in the final e-mail,
e.g., plaintext) could be removed once the user selects "Send".
Neil, those are good points about removing the old sig on change of identity in
the case of top quoting. But I don't think it should hold up this getting
checked in (once your other review comments are dealt with), since changing
identities is rare, and sig on top nor quoting on top are the default.
Neil, I've looked at making the pref a boolean but mozilla extracts the value
from menulist as a string rather than a boolean which means further down the
line in AccountManager.js (line 598) there are problems. When it tries to set a
boolean from the string "false" it always ends up as true, no matter what the
string is unless it is "". Using "" in menulists for a value causes an extra
blank entry to appear so it looks like it will have to stay the way it is.

For the same reason I'm pretty sure i'll need to stick to using a hidden
checkbox for the AttachSignature pref as that seems to be the only way of
guaranteeing it is a boolean. label turns it into a string and I expect data
would do the same.
Try genericattr="true" - it should convert "true" and "false" into boolean.
Target Milestone: Future → M3

some one did a boo-boo with future milestone setting, can some one please change
it back?
Target Milestone: M3 → Future
The problem with switching identities is that anything after the signature file
is deleted no matter where the signature is located. I'll do some exploration
this weekend.
A solution could be to disable identity switching when top-reply+top-sig is
selected in the prefs. I doubt anyone would notice. And then it would be a
separate bug.
I'm being spammed to death by the comments on this bug.. yet I don't see myself
on the CC list so I can't remove myself.  How do I get off this list?
Sam - you voted for this bug. You will need to change your email preferences:
genericattr="true" seems to change the behaviour of id's too e.g.
id="identity.sigBottom" creates in each identity a pref of identity.sigBottom so
I've changed the id to "identity.sig_bottom". Does this mean I no longer need
the attribute prefstring? Does genericattr have any other side effects?
I'm still investigating several issues with this patch and when the signature is
configured to be above the quoted text:
a) When composing a new message there is no delimiter
b) When replying switching identities does not work - the quoted text gets
deleted as well as the signature being changed.

I'll try and find time to look at both these issues this week but I know I'm
very busy with non-mozilla stuff.
When a patch will be available for Windows builds ?
Attached patch Patch v0.6c (obsolete) — Splinter Review
This patch fixes the problem with composing new messages.
If one or more accounts has the signature set to be above the quoted text then
there is still an issue with swapping identities, the new signature gets
inserted rather than overwriting the existing signature - this issue could
possibly be spun off into another bug but, seeing I've missed 1.5b anyway,
could be dealt with in this bug.
Attachment #130103 - Attachment is obsolete: true
Attachment #130529 - Flags: superreview?(bienvenu)
Attachment #130529 - Flags: review?(
*** Bug 217021 has been marked as a duplicate of this bug. ***
Above or Below? That is the question.
I have learned how to compose e-mails on my own with no one teaching me.  I see
posting above w/sig as the only logical option.  The most recent message is
there for immediate viewing.  The older messages below for archival purposes.
How is scrolling down to the bottom of the message to read the most recent post
"proper" e-mail composition?  I would appreciate if someone would explain this
to me.  I don't realy know how to utilize newsgroups since I have recently
discovered them, but I suspect this "posting above/below" debate has it's roots
in the evolution of normal email usage vs. newsgroups usage.

> Above or Below? That is the question.

No. As far as Mozilla is concerned, that question is answered: Leave the user
the choice. I personally think that quote-below is a bad idea, but it's offtopic
The question in this bug is: Does it make sense to have the sig above or below
the quote, *if* your quote is below the new text? And if so, can we make it work
right/well technically (sig-stripping, interop)?

> I would appreciate if someone would explain this to me.

I don't want to re-ignite discussion, but just to explain the logic behind

> The older messages below for archival purposes.

Quotes are *not* for archival purposes, but (1) to *shortly* re-introduce the
reader into the discussion again and (2) to show which specific point you answer
to. This often means that your replies can be shorter. Quotes are meant to be
skimmed over, not to be ignored. Because of the time difference, this may be
even more important for 1:1-email than for newsgroups. If you need an *archive*,
then save your mails - you can always look them up (by threading, sorting,
searching or even more easily, if bug 62033 is fixed - works with other clients).

> How is scrolling down to the bottom of the message to read the most
> recent post "proper" e-mail composition?

Quotes are supposed to be minimal, 1-3 lines each, as demonstrated with this
comment, so you usually don't have to scroll. Also, if the logic of my first
answer is used, you always have to skim over the quote first before you can read
and understand the new text.

Forgivr my ignorance but how do I install this Patch 0.6c?
Peter: unless you compile your own version of mozilla you will not be able to
apply the patch, if you do then you should be able to use the appropiate patch
Comment on attachment 130529 [details] [diff] [review]
Patch v0.6c

People who reply on top just shouldn't change identity, obviously :-)
Attachment #130529 - Flags: review?( → review+
Attachment #130529 - Flags: superreview?(bienvenu) → superreview+
Comment on attachment 130529 [details] [diff] [review]
Patch v0.6c

Requesting driver approval for 1.5, fairly straight forward patch which adds a
new feature and is fairly low risk.
Attachment #130529 - Flags: approval1.5?
Blocks: 218346
Comment on attachment 130529 [details] [diff] [review]
Patch v0.6c

ok, forget that request, leave it til 1.6
Attachment #130529 - Flags: approval1.5?
Comment on attachment 130529 [details] [diff] [review]
Patch v0.6c

Bad news: the patch has bitrotted.
Good news: fix is to remove the semicolons from this section.

>Index: mailnews/base/util/nsMsgIdentity.cpp
>RCS file: /cvsroot/mozilla/mailnews/base/util/nsMsgIdentity.cpp,v
>retrieving revision 1.61
>diff -p -u -d -8 -r1.61 nsMsgIdentity.cpp
>--- mailnews/base/util/nsMsgIdentity.cpp	12 Aug 2003 12:32:59 -0000	1.61
>+++ mailnews/base/util/nsMsgIdentity.cpp	28 Aug 2003 09:12:05 -0000
>@@ -458,16 +458,17 @@ NS_IMPL_IDPREF_STR(Email, "useremail");
> NS_IMPL_IDPREF_STR(ReplyTo, "reply_to");
> NS_IMPL_IDPREF_WSTR(Organization, "organization");
> NS_IMPL_IDPREF_BOOL(ComposeHtml, "compose_html");
> NS_IMPL_IDPREF_BOOL(AttachVCard, "attach_vcard");
> NS_IMPL_IDPREF_BOOL(AttachSignature, "attach_signature");
> NS_IMPL_IDPREF_BOOL(AutoQuote, "auto_quote");
> NS_IMPL_IDPREF_INT(ReplyOnTop, "reply_on_top");
>+NS_IMPL_IDPREF_BOOL(SigBottom, "sig_bottom");
> NS_IMPL_IDPREF_INT(SignatureDate,"sig_date");
> NS_IMPL_IDPREF_BOOL(DoFcc, "fcc");
> NS_IMPL_FOLDERPREF_STR(FccFolder, "fcc_folder");
> NS_IMPL_IDPREF_STR(FccFolderPickerMode, "fcc_folder_picker_mode");
> NS_IMPL_IDPREF_STR(DraftsFolderPickerMode, "drafts_folder_picker_mode");
Attached patch Patch v0.6d (obsolete) — Splinter Review
Revised patch which is unbitrotted and also fixes identity switching problem
with compose window when one or more of the identites being used has the
signature position set to be above the quoted text.
Attachment #130529 - Attachment is obsolete: true
Attachment #131078 - Flags: review?(
Comment on attachment 131078 [details] [diff] [review]
Patch v0.6d

Not sure why yet, but the identity switching code fails to insert an HTML
signature on top unless the message starts with a blank line.
Attachment #131078 - Flags: review?( → review-
Ian, I tried putting your patch in a thunderbird build and I noticed the
following regression:

When inserting the signature at the top, we are dropping the "-----" line from
the signature when replying to a message. For new messages, the dashes are
present. Just the replies when the signature gets inserted at the top. That's
the case where we now seem to bypass the mail compose code that inserts the
boundary text.
Scott, yes it drops the delimiter by design when replying with the signature set
to above quoted text, so that clients that strip the signature do not strip the
quoted text.
It the missing delim bothers anyone, you could put in "--\n" instead of the
normal delimiter "-- \n". If the space is missing, the signature detection of
recipients should not trigger, but it would still look like a sig delimiter to
Attached patch Patch v0.6e (obsolete) — Splinter Review
Patch adds ForwardAsAttachment as a type that has a delimiter when pref is set
for signature to be above quoted text.
Attachment #131078 - Attachment is obsolete: true
Comment on attachment 131492 [details] [diff] [review]
Patch v0.6e

Issues on identity switching will be covered by bug 218346
Attachment #131492 - Flags: review?(
Blocks: 219336
This is now working in Mozilla Thunderbird. Can we now mark this bug fixed?
No, as it's not been checked into the trunk yet (as far as I am aware).
Comment on attachment 131492 [details] [diff] [review]
Patch v0.6e

> function quoteEnabling()
> {
>   var quotebox = document.getElementById("thenBox");
>+  var placebox = document.getElementById("placeBox");
>   var quotecheck = document.getElementById("identity.autoQuote");
>+  if (quotecheck.checked && !quotecheck.disabled &&
>+      document.getElementById("identity.attachSignature").checked &&
>+      (document.getElementById("identity.replyOnTop").value == 1)) {
>+    placebox.firstChild.removeAttribute("disabled");
>+    placebox.lastChild.removeAttribute("disabled");
>+  }
>+  else {
>+    placebox.firstChild.setAttribute("disabled", "true");
>+    placebox.lastChild.setAttribute("disabled", "true");
>+  }
>   if (quotecheck.checked && !quotecheck.disabled) {
>     quotebox.firstChild.removeAttribute("disabled");
>     quotebox.lastChild.removeAttribute("disabled");
>   }
>   else {
>     quotebox.firstChild.setAttribute("disabled", "true");
>     quotebox.lastChild.setAttribute("disabled", "true");
>   }
Nit: That blank line should be between the var and the if lines.
Attachment #131492 - Flags: review?( → review+
Attached patch Patch v0.6fSplinter Review
Revised patch adding blank line between var and if.
Attachment #131492 - Attachment is obsolete: true
Comment on attachment 131996 [details] [diff] [review]
Patch v0.6f

Carrying r= forward and requesting sr
Attachment #131996 - Flags: superreview?(bienvenu)
Attachment #131996 - Flags: review+
Attachment #131996 - Flags: superreview?(bienvenu) → superreview+
Ok this has now landed on the trunk (thank you timeless), any remaining issues
to do with identity switching are in bug 218346.

P.S. All those promises of money for fixing this bug should be donated to the
Mozilla Foundation - for details of how to, look at
Closed: 22 years ago21 years ago
Resolution: --- → FIXED
I just sent my PayPal donation.  And I did not even promise.
Re: comment 283, comment 290 et seq, I'm still not sure of the idea behind
omitting the delimiter in top-reply, top-sig.

Have we established that, in the average business in which top-posting is house
style, it is house style just as well that all reply levels are left in, so that
the total amount of data transferred in a conversation increases quadratically
with successive replies?
Donation sent as promised. Thanks! :-)
Donation made of $65 (£40) as promised.

Thanks guys - keep up the good work!
My experiences are that is the way things are done.  Some people like to have
the entire conversation with every email.  This can be good for cases where you
CC someone in the middle of the "conversation".
Donation sent to MozillaFoundation as promised. Thanks! :-)
I take it this did not make it into 1.5?
Running the latest nightly build, 20031019 I notice this works when replying but
not forwarding. I know this isn't in the origional scope, since reply was only
stated, but maybe it should also be used in forwarding? Should I stary a new
bug? Also When I compose I get a -- delimiter but when I reply I don't.
Justin, all that you noticed are not new bugs:
"When I compose I get a -- delimiter but when I reply I don't". This is
intentional, to avoid problems with some mail clients that automatically cut
everything after the delimiter.
Afaik, forwarding + sig might not work because of the switching identities issue
(bug 218346). It will work with a new profile.
*** Bug 224333 has been marked as a duplicate of this bug. ***
*** Bug 230476 has been marked as a duplicate of this bug. ***
Product: MailNews → Core
No longer blocks: majorbugs
*** Bug 300249 has been marked as a duplicate of this bug. ***
Product: Core → MailNews Core
You need to log in before you can comment on or make changes to this bug.