36441 - Reply-To and Followup-To not shown

Status: VERIFIED FIXED

(SeaMonkey :: MailNews: Message Display, defect, P3)

Description

[Ben Bucksch (:BenB)]

If Reply-To or Followup-To headers are present in the msg, they should be shown
in the msg pane. They are not.

I think, this is a bug and a beta2 stopper. These headers carry very important
meta information. It happens (and is fully legal), that the From address is not
reachable.

Comment 1

[Ben Bucksch (:BenB)]

I think, John Moreno said, this were a regression.

Comment 2

[Phil Peterson]

[nsbeta2+]

Comment 3

[selmer (gone)]

Msg Headers to MScott.

Comment 4

[Scott MacGregor]

We haven't implemented view all headers yet.


*** This bug has been marked as a duplicate of 17001 ***

Comment 5

[fenella]

Mark it verified/dup.

Comment 6

[Matthew T (active 1999-2002)]

I don't think this is a dup. Reply-To and Followup-To should appear by *default*, 
that is, in Normal headers mode, not just in Full headers mode.

Comment 7

[Ben Bucksch (:BenB)]

Oh, if mscott meant that, I agree. As I said, this is important info. That's why
Followup-To is often even included in the body.
REOPENing.

Comment 8

[selmer (gone)]

Mass moving M16 to M17 - look for nsbeta2 before anything else.

Comment 9

[Ben Bucksch (:BenB)]

selmer, this *is* nsbeta2+

Comment 10

[Scott MacGregor]

With the implemtnation of View All Headers mode and the ability to click on the
name of a a haeder (in normal mode) to get a popup showing you all the headers,
I don't believe this is a beta2+ bug anymore. And it's NOT a regression. I'm not
sure how it got that tag.

Clearing regression keyword. clearing beta2 nomination so PDT can re-evaluate.
I'd like to assign this to M18 if PDT doesn't re-plus it.

Comment 11

[Ben Bucksch (:BenB)]

mscott, view all headers (in whatever form) just doesn't cut it. reply-to and
followup-to are occasional, but very important headers. I don't want to check
all headers of every msg, just in case there's a followup-to header or so.

I added regression, because I thought to remember having seen reply-to in
Mozilla's normal header view before.

Comment 12

[leger]

Putting on [nsbeta2-] radar. Not critical to beta2, putting on nsbeta3 radar.

Comment 13

[John Moreno]

I posted a test message to <news://secnews.netscape.com/[e-mail]>

And this is definitely a regression, earlier versions did show the appropriate headers.

Comment 14

[Ben Bucksch (:BenB)]

Readding regression keyword per John's comment.

Comment 15

[laurel]

possible itme for b2 release notes

Comment 16

[lchiang]

- per mail triage for Netscape release
adding helpwanted if mozilla.org contributor can fix for Mozilla.

Comment 17

[Ben Bucksch (:BenB)]

*yell* This bug has to be fixed.

(Assuming we honor the reply-to and followup-to headers, ) the user won't know
why suddenly a new address appeares inthe composer.

Also, I have been biten by a not displayed reply-to header when I added the From
address to my address book and tries to send mail to that address for weeks
until I catched that person on IRC.
I know that a Netscape employee recently didn't find his own posts, becauses he
didn't notice the followup-to.

Requesting reconsideration.

Comment 18

[Scott MacGregor]

I agree with the triage teams recommendation. we are passed the point of being
able to fix this kind of bug in this release. Work arounds include View all
header mode or using the all headers popup by clicking on a header. This should
be a nsbeta3-

Comment 19

[Ben Bucksch (:BenB)]

Quote from the introduction of GNKSA:
<quote>
I believe most of this anger is misdirected.  The new users aren't
really that different from the old-timers.  What _is_ different is that
many of the old-timers are using relatively well-behaved software,
typically `rn' or one of its offspring, while many of the newbies are
using `tin', `uqwk', `AOL', or various PC newsreaders.  Unfortunately,
these programs frequently violate assumptions that come naturally to
people used to well-behaved readers:

  - The user can see the essential header fields, including "Newsgroups"
    and "Followup-To".
[...]
</quote>
Later, in the requirements:
<quote>
1)  Display all essential header information
[...]
  c) The list of newsgroups the article was posted to.  This list MUST
     be displayed in full, never truncated.  [...]
  d) The article's Followup-To list, if this is different from the
     Newsgroups list.  This MUST be displayed in full, never
     truncated.
</quote>

Before I argue any longer, I fixed this. Taking bug. Will attach patch.

mscott, please review.

Comment 20

[Ben Bucksch (:BenB)]

Attachment: Fix

Comment 21

[Ben Bucksch (:BenB)]

Attachment: Fix, version 2. Also adds Bcc.

Comment 22

[Scott MacGregor]

I disagree with showing this in the basic header pane. Since the headers don't 
scroll adding these headers takes up even more real estate than we should be. 

I'm probably not going to accept adding these headers without talking it over 
with our UI folks first.

4.x didn't show these headers in the short header view either and I don't 
believe we need to start doing so now (unless headers scrolled with the body).

but thanks for the patch anyway =)

Comment 23

[Ben Bucksch (:BenB)]

> I disagree with showing this in the basic header pane.

It is a GNKSA requirement.

> Since the headers don't 
> scroll adding these headers takes up even more real estate than we should be.

They only take up space, if they exist in the msg. In this case, it is extremely
important to show them (see above).

> talking it over with our UI folks first.

Adding jglick

> 4.x didn't show these headers in the short header view either

It does show them in the normal header view (the default), which is the
equvivalent of our header pane.

Comment 24

[Ben Bucksch (:BenB)]

I am *so* sick of arguing about each and every patch I submit. Escalating -
Steve?

> adding helpwanted if mozilla.org contributor can fix for Mozilla.

What exactly did I misunderstand?


I wonder, if I'm just such an asshole that (almost) everybody wants to annoy me,
or what. Answers welcome :).

Comment 25

[Ben Bucksch (:BenB)]

(If it was Lisa's /personal/ decision to add |helpwanted|, then please excuse me
for throwing that in. I don't want to move the problem to Lisa.)

Comment 26

[John Moreno]

mscott wrote:
> I disagree with showing this in the basic header pane. [...] adding
> these headers takes up even more real estate than we should be. 

These headers are relatively rare, and for that reason, it is absolutely
necessary that they be shown when they ARE present.  These headers when
obeyed, drastically change the behavior of a reply/follow-up, and there
very rarity means that even experienced users aren't going to waste their
time looking for them.

(BTW -- I just did a check on a newsgroup with 8700 articles, it had 69
articles with a Fu2 header; and about a 3rd of them were just duplicates
of the Newsgroup header.  Ben, I hope your patch checks for that, the
headers don't need to be shown if they just duplicate the main headers).

Comment 27

[Ben Bucksch (:BenB)]

John, no, the attached patches always show the followup-to/reply-to headers, if
present. Since you said this, I implemented (just a one-line change) and tested
it, but I am not sure, this is the right thing to do.

If there is a followup-to header, and you hit "Reply" (which has the wrong
tooltip "Reply to Sender only"), only the newsgroups in the followup-to header
are used as recipients. If no followup-to is present, the sender is the only
recipient. If you don't show the followup-to header in the viewer, this
behaviour might be very confusing for the user. I'll attach the patch anyway, so
that you can at least apply it locally.

BTW: Confusion vs. predictability is also an argument for fixing this bug.

Comment 28

[Ben Bucksch (:BenB)]

It's easier to say it here. Apply the latest patch and change

if (currentHeaderData["followup-to"]

to

if (currentHeaderData["followup-to"] &&
currentHeaderData["followup-to"].headerValue !=
currentHeaderData["newsgroups"].headerValue)

in mailnews/base/resources/content/msgHdrViewOverlay.js.

Comment 29

[Ben Bucksch (:BenB)]

rather

if (currentHeaderData["followup-to"])

to

if (currentHeaderData["followup-to"] && (!currentHeaderData["newsgroups"] ||
currentHeaderData["followup-to"].headerValue !=
currentHeaderData["newsgroups"].headerValue))

Comment 30

[Scott MacGregor]

I didn't know they were in the 4.x normal header view. Given that information 
and the fact that according to John they are rare as well. That means you won't 
always be losing the real estate to these headers. So I agree I'm the wrong 
here. I'll look at reviewing and landing Ben's patch if he wants me to check it 
in (otherwise he can after I review it tomorrow).

I'm going to bed now =).

Comment 31

[Ben Bucksch (:BenB)]

mscott, thanks. If you want to save yourself somem time, review this bug and bug
40036 at the same time. I'll check in then.

Comment 32

[John Moreno]

Ben replied to my saying R2/FU2 should only be shown if different from From/Newsgroups with:
> but I am not sure, this is the right thing to do.

It is, the below is a bug, showing the headers it will use doesn't
change that.

> If there is a followup-to header, and you hit "Reply" (which has the wrong
> tooltip "Reply to Sender only"), only the newsgroups in the followup-to 
> header are used as recipients.

Ben, this is a separate bug -- I've looked and haven't seen it reported
elsewhere, but if trying to reply via email results in a news message
that is a bug; no matter where it gets it's info for filling out the
Newsgroups header.  This looks to be a variation of bug 32024 (and a couple
of others) which is assigned to [e-mail].

I'm going to mail him a copy of this message.

Comment 33

[Ben Bucksch (:BenB)]

John, I agree that *this* is a bug. I also agree that our Reply buttons are
broken in general.
In general, we have the following possible recipients of the reply:
- sender
- recipients
- newsgroup(s)
(of the original msg). We'd have to make sure, a followup-to header overrides
only the newsgroup(s) and doesn't change any other recipients.

Does it make sense (in general) to offer the option to reply to the current
newsgroup only? What should we do for such a command, if a followup-to header is
present? E.g. you have a crosspost, followup-to identical to newsgroups, but
user said "Reply to current newsgroup only". Does such a command make sense? If
yes, what should we do?

Comment 34

[Ben Bucksch (:BenB)]

BTW: Why is the followup-to header at all specified, if it matches the
newsgroups header?

Comment 35

[John Moreno]

I don't think that replying to just the current newsgroups makes sense, there's already 
a way of fixing that, just edit the headers after starting the followup.  That's not a
common enough decision to make it worthwhile to special case it.

As for why FU2 is there if it is just a duplicate of Newsgroups: generally it's a bug
in other software (old versions of tin particularly), but sometimes it's user error
(this is more common with Reply-To, but it happens for FU2 to).

Comment 36

[Ben Bucksch (:BenB)]

OK, will make the change and attach the patch to this bug or bug 40036.

Comment 37

[lchiang]

yes, it was my decision to add helpwanted since I knew that some of you wanted 
this to be in and Netscape wasn't going to address this in the first release.

But, I'm glad to see this has worked out for all.

Comment 38

[Ben Bucksch (:BenB)]

Lisa, ok, then sorry. Feel free to continue to add helpwanted were you find it
appropriate.

Comment 39

[Ben Bucksch (:BenB)]

em, s/were/where. *red*

Comment 40

[Scott MacGregor]

why did you add a bcc box to the overlay? This just slows us down 'cause it will
never get used. There's no such thing as a bcc header on incoming mail. bcc
values get stripped out when we go to send the message. If they were still in
the message header then it wouldn't really be a blind cc would it =).

Can you take that out?

Still reviewing the rest

Comment 41

[Ben Bucksch (:BenB)]

Odd. Am I too used too Lotus Notes? (You could always see yourself, and only
yourself, in the bcc header, if you appeared there.) But for SMTP, you seem to
be right. Will check.

Comment 42

[Ben Bucksch (:BenB)]

You are right. Will remove it.

Comment 43

[Matthew T (active 1999-2002)]

Hangonaminute. You see BCC fields if you copy your outgoing messages to a folder. 
In such a case, knowing who you BCCed a message to is just as important as 
knowing who you Toed or CCed it to, IMO.

Comment 44

[Ben Bucksch (:BenB)]

Also right. I vote for showing bcc. mscott?

Comment 45

[Scott MacGregor]

Sorry I'm against it =). It's a rare case and I don't believe it should be in
the normal header view. Go to view all header mode or use the popup would be my
answer. Compromise with me...we are already adding follup to and reply to the
basic header view.

Comment 46

[Matthew T (active 1999-2002)]

But in the normal case, that where there are no BCCs, it doesn't appear anyway, 
right? So the only case where it takes up room is where there *are* BCCs.
Changing to all-header mode (and seeing all the technical headers like 
`Received:' etc) seems to be a ridiculous length to go to, just to see who you 
sent a copy of a message to.

Comment 47

[John Moreno]

For what it's worth, I agree with Ben and Mathew -- I often view old
messages both those that I sent and those that I received,  when
viewing messages I sent, I want to know who I sent it to; that includes
people that I bcc'd.

I don't believe the check is processor intensive, and the header will
only be shown when it is present, it's important, possibly vital ("have
I sent this  to him before?") information -- requiring the user to go
through the steps to view the full headers in order to see who they
sent the message to, seems a bit excessive for the nominal gain.

Comment 48

[Ben Bucksch (:BenB)]

View All Headers is nothing for Joe User. Joe user will want to see BCC in Sent
copies (if he used it). -> View All Headers is no solution.

If you want to increase speed, maybe you can cut the hidden pref for showing
"UserAgent"? *g*

Seriously: IMO, showing all recipients (if possible) is a basic requirement of a
MUA.

Comment 49

[Ben Bucksch (:BenB)]

mscott, you changed the files I touch in this bug. Please merge the changes and
attach a new patch.

Comment 50

[Ben Bucksch (:BenB)]

mscott merged the changes (removing some cleanup I did and the bcc stuff) and
checked them in. Thanks!

Filed bug 50103 about bcc.

Comment 51

[laurel]

OK using sep11 commercial build linux rh6.0, mac OS 9.0 and NT 4.0
Reply-to, Followup-to headers shown in View All Headers.
bcc not shown.