478466 - (me-header) [meta] shortened e-mail addresses: "Me" in the message window header bar ambiguous or inappropriate in some cases

Status: RESOLVED FIXED

(Thunderbird :: Message Reader UI, defect)

Description

[Joachim Kock]

User-Agent:       Mozilla/5.0 (Macintosh; U; Intel Mac OS X 10_5_6; en-us) AppleWebKit/525.18.1 (KHTML, like Gecko) Version/3.1.2 Safari/525.20.1
Build Identifier: version 3.0b1: the about box says: Mozilla/5.0 (Macintosh; U; Intel Mac OS X; en-US; rv:1.8.1.18) Gecko/20081105 Thunderbird/2.0.0.18 Mnenhy/0.7.6.0

This is bug report against design, not against malfunctioning.

In the new header pane of the message windows, the name of the
main personality is replaced by the personal pronoun 

   You
    
I find this ridiculous for many reasons.  Briefly it is just
plain wrong to hide the exact information in favour of some
relative unstable reference like a pronoun, and it serves no
purpose.  A compute program should be objective and discrete.
The "You" is almost like pointing a finger at you.

Why "You" and not me "Me"?  This goes against the idea of unity 
between computer and user that is usually aimed at, for example
through names like "My Computer" or "My Documents".

What if you have more than one personality?  Or what if you sign 
yourself as

   Prof. Smith <[e-mail]>
   
towards your students and therefore expect their emails replies
to contain this same "To:" line, whereas among friends you use

   Patty <[e-mail]>
  
Replies to any of these will just show up as

   You

and if your students figure out you use Thunderbird 3.0beta,
they might even have fun of writing emails to you addressed to

   **** <[e-mail]>
   
and you would never notice it! --- even when you hover the mouse
over the "You", you will only see [e-mail].


There seems to be no justification for this change (and I suspect 
it is not a change that has been discussed very much before
suddenly showing up as part of a attractive new header bar.  Is
it an attempt of the program to be cool?  Or to create the
illusion that the program is talking to you or something?  I doubt
people enjoy that.  An email client should do its work and not try 
to be personal or smart.  Imagine a secretary doing something 
similar, putting a "You"-sticker on top of the addressee of the
letters you receive!

Furthermore, it is a disturbing interface element because it is
used only in the message header bar, not anywhere else in the
interface (neither in message listings, nor when you compose a
new message, even if it is from You).

Finally, have you considered the complications this design will 
have when localising Thunderbird into other languages?  In French 
for example, how will you determine whether the user prefers to be 
titulated "Toi" or "Vous"?  In fact both would look very strange as 
an interface element.  Or in German, in addition to this problem, 
there is the problem of case: supposing "To" translates into "Für" 
the corrrect case of the pronoun would be "Dich" or "Ihnen".  Both 
would look extremely awkward as an interface element.  I can only 
guess about the problems arising in more exotic languages...



Reproducible: Always

Steps to Reproduce:
1. Open an incoming message
2.
3.
Actual Results:  
The recipient is displayed as "You"

Expected Results:  
It should be displayed as in the header field.

This is bug report against design, not against malfunctioning.

Possibly it is also against the decision procedures that led to
the criticised design.

Comment 1

[rsx11m]

The "You" was introduced in bug 456818, but should only appear in the collapsed view, thus not when you have the detailed display (I'm seeing it with "show details" as well though, thus this is somewhat inconsistent or another bug). Anyway, that string is hard-wired in messenger.properties and the behavior not configurable in any way (at least not in the current state, maybe that's envisioned in the future).

The only way to avoid the "You" appears to be adding your own identity to the address book, then uncheck "Show only display name for people in my address book", which will display the full name and address with and without details shown. This is certainly not intuitive, and I can see reasons for not wanting to abbreviate your own address like this, specifically w/ multiple identities.

Adding the user-experience expert to the CC list for commenting.

Comment 2

[Joachim Kock]

Thanks a lot for the pointer to Bug 456818, it is very helpful.  I have read
the dicussions there, and found some remarks about the issue, like

> Comment#12 (Simon 'sipaq' Paquet):
> 
> This string may not work from an i18n standpoint since it is very
> English-centric. Other languages have different translations for "you" based on
> gender, politeness rate, etc. I guess we'll have to find out what localizers
> say to this.

followed by

> Comment #13 (David Ascher (:davida))
> 
> Agreed.  I wonder if using "me" forms might be easier to translate.  FWIW, it
> seems to be what gmail uses in a l10n email context.

Yet, the discussion contains absolutely no justification for the introduction 
of the pronoun.

I understand from the discussion that in the file

/mail/locales/en-US/chrome/messenger/messenger.properties

there is a declaration

   # second person direct object pronoun; used in the collapsed header view if
   # the user is in the To or Cc field of a message
   headerFieldYou=You

and later, in 

mail/base/content/msgHdrViewOverlay.js

this is referenced as follows

   -  var displayName = cardDetails.card.displayName;
   +  var displayName;
   +  var allIdentities = accountManager.allIdentities;
   +  identity = getBestIdentity(allIdentities);
   +  if (emailAddress == identity.email) {
   +    displayName = gMessengerBundle.getString("headerFieldYou");
   +  } else {
   +    if (!cardDetails.card) {
   +      return;
   +    }
   +    displayName = cardDetails.card.displayName;
   +  }

This is from a diff of a change occurring around 2008-11-24.

Notice in this diff that the absolutely clean, simple, logical, 
accurate, objective and neutral-in-tone solution of displaying 
the cardDetails.card.displayName is replaced by the convoluted,
unprecise, pointing-finger-against-user-tone, localiser-
unfriendly idea of _masking_ this information by displaying
instead the "You" string.

Hereby the program _overrules_ the user, effectively telling
him 

   I know better who you are than you do.  You are just "You".
   You cannot escape this; you have no influence on how to appear:
   change your personal information in the setup, yet you will 
   always just remain "You".  (This string is immutable, hard 
   coded into the program.)

I would very much suggest the change expressed in the diff is
reverted.

(Both in the expanded header and in the collapsed header.
The problem is the same in both cases.)

Cheers,
Joachim.

Comment 3

[rsx11m]

Based on the code snippet you are quoting, I'm amazed that my workaround from comment #1 actually works, given that the identities are tested first and then only the address book. Maybe I'm reading something wrong here...

Anyway, this splits up into two different issues:

(1) If an account has multiple identities, or if the Global Inbox is used for more than one accounts, or any e-mail read in Local Folders, the "You" is highly ambiguous and should be avoided. On top of that, problems similar to those reported in bug 457304 apply where the addressee may not match the address.

(2) The other issue is the one of localization, as addressed in bug 456818 comment #12, thus adding the "intl" keyword to the bug. Even in a region where a casual "You" would be considered appropriate, it may not hit the right tone, e.g., in a business environment. Is there any feedback yet from the localizers on this point?

For (1), I'd see a possible solution in using "You" only in a folder which is clearly associated with a single identity. For (2), this may be yet another preference setting per identity allowing/disallowing the "You" to be applied. The default may be localizable to reflect different expectations due to regionally different customs and etiquette.

Comment 4

[Joachim Kock]

In reply to Comment#3, where two different issues are singled
out:  I think one basic issue precedes those two, namely

(0) Why _ever_ employ the "You" string?  What was the reason 
for implementing this substitution in the first place?
So far I can see only drawbacks and no advantages.
Simply removing the substitution mechanism would make 
issues (1) and (2) moot.  

Re (1) and (2): assuming for a moment that there are situations
where it would be preferable to see the string "You" instead of
the real information, then it will seem necessary to convolute the
mechanism with several further conditionals and exceptions,
involving folder configuration, addressbook setup, localisation,
and possibly even preference flags.  For a programmer it is a
child's play to nest conditionals and exceptions.  However, it
is highly unlikely that such conditionals can be crafted in such
a way that the resulting behaviour matches the expectations of
every user.  (See Bug 457304 and Bug 474721 for related 
discussion.)  The more complicated the rules, the more difficult
it is for the user to understand the logic behind the behaviour,
and the higher is the risk that it will look like a whim of the
program.  Imagine writing the user manual for this "You" feature:
"We will now give an overview of the different cases in which 
your name will be replaced by "You"".  Is it really worth it?
In contrast, every user would understand the logic behind always
displaying the proper name -- in fact 99% of all users would not
even think about the possibility that any other string could be
substituted for it.  Nothing that could ever cause confusion, no
preference flags to clutter, no documentation to write.

Sorry for being long -- I'm afraid I already overstayed my welcome.

Cheers,
Joachim.

Comment 5

[Bryan Clark (DevTools PM) [:clarkbw]]

(In reply to comment #0)
> A compute program should be objective and discrete.

We disagree with this premise.  

There are issues that still need to be resolved with the use of You, some of which you pointed out.  It could be good to break out those issues mentioned into bugs that are actionable.  However I can't see forward progress being made in this bug as it's more of a discussion than actual objectives.

Discussion should take place on the mailing list/newsgroup.  I think your best course of action is to create an extension or patches that demonstrates the behaviour you want and then discuss those elements on the newsgroup.

Comment 6

[rsx11m]

Bryan, is there any Wiki page covering the plans regarding the use of "You"?

Thus far, I have to agree with the reporter that it is causing more problems than actually being useful...

Comment 7

[rsx11m]

I found https://wiki.mozilla.org/Thunderbird:Message_Reader as a concept page
for the headers and the message reader as such and briefly summarized the
points (0)-(2) on its Talk page. That link is also added as URL to this bug.

Comment 8

[Joachim Kock]

In response to Comment #5 (quoted only in part):

> We disagree with this premise.

Fair enough -- I was sort of suspecting this to be the case.

> Discussion should take place on the mailing list/newsgroup.

I wonder if there was ever a discussion of the "You" feature
prior to its implementation.  That would be a good starting
point for the present discussion.

(I searched the archive and found (and read) 52 messages in
the thread  "Experimental Message View".  Very inspiring
and interesting discussion.  I did not find mention of the "You"
feature though, and I am not sure it is part the big picture
discussed in that thread.)

> I think your best
> course of action is to create an extension or patches that demonstrates the
>behaviour you want and then discuss those elements on the newsgroup

The patch would be the converse of the diff quoted in Comment#2.  I just 
found the file in question, mail/base/content/msgHdrViewOverlay.js: 
the "You" is inserted at line 1062.

(I have now become eager to try to build my own Thunderbird -- this would 
be an interesting experience for me (a very recent Thunderbird fan, after 
15 years of Eudora).  Maybe I'll find the time to try over the weekend, then 
I'll report back.)

Cheers,
Joachim.

Comment 9

[rsx11m]

(In reply to comment #3)
> Based on the code snippet you are quoting, I'm amazed that my workaround from
> comment #1 actually works, given that the identities are tested first and then
> only the address book. Maybe I'm reading something wrong here...

I did - there is gShowCondensedEmailAddresses, being checked before the "You" if() statement is entered. Thus, unchecking "Show only display name for people in my address book" should avoid the "You" even without adding your own address to the address book. A workaround, yes, but a rather obscured one.

(In reply to comment #8)
> I wonder if there was ever a discussion of the "You" feature
> prior to its implementation.  That would be a good starting
> point for the present discussion.

I was unable to find any discussion on this as well, despite looking around quite a bit. A pointer to some previous arguments would be helpful here to determine which issues were already thought of and which ones not.

(In reply to comment #5)
> ... However I can't see forward progress being made
> in this bug as it's more of a discussion than actual objectives.

Well, the charge for this bug seems to be well defined:

1. Establish that "You" causes problems (can be checked off as "done" by now);
2. determine in which scenarios "You" is appropriate (I'd only see this in the
   single-account/single-identity case, otherwise it will always be ambiguous);
3. determine efforts needed to apply considerations in #2;
4. if efforts exceed benefits, consider rolling back respective changes made
   by bug 456818.

Nevertheless, discussions and concept presentations on the Wiki or elsewhere may be fine as long as those can be kept track of. This wouldn't be the first bug starting off with discussions leading to concrete solutions, by the way.

Unless anybody strongly objects, this appears to be a valid bug report.

Comment 10

[Sebastian Hengst [:aryx] (needinfo me if it's about an intermittent or backout)]

Setting to new.
The user should be able to distinguish in mixed folders like local inbox to which mailbox, person or identity the mail was sent. (And Gmail and Facebook aren't the philosopher's stone.)

Comment 11

[rsx11m]

Thanks for confirming. I'm making the summary a bit more descriptive and add a dependency to the tracking bug 456814. Due to the ambiguity factor, I'd still consider this a bug, though minor as a hidden but easy workaround exists.

Comment 12

[Dan Mosedale (:dmosedale, :dmose)]

It seems like there's another, easier to discover workaround: mouseover the word "You", and a tooltip showing the specific email address in question appears.

Comment 13

[Joachim Kock]

In reply to #12: that is not a workaround for any of the issues of 
the original report.  The tooltip only displays the email address, 
not the name! (as already observed in the original report).

But thanks for your interest in this bug.

Among the comments above several express their sympathy 
with this bug, agreeing that there are situations where the 
"You" may be inconvenient.

I would like to stress again that the whole concept of the 
"You" substitution is wrong.

If I choose some name for myself in the addressbook, it is not
for the email program to call me anything else.  
Why should the email program be allowed to override my choice?

There is even a flag in the preferences to check: "Show display name
only".  And the program respectlessly ignores this flag and shows
something else!

So far there seems to be no reason whatsoever to make this
substitution -- at least none expressed in the comments above.

If anybody can direct me to the discussion that led to the 
introduction of the "You" substitution I would be very grateful.
I am very curious to learn about the rationale, the pros and
contras discussed at that time, and more generally to get some 
insight in the decision procedures behind the development of
the software.

Cheers,
Joachim.

Comment 14

[Christoph Grether]

The main problem is, if you are using Thunderbird as default mail client for your company. I doubt, that a company wants to see "You" instead of the company name. It makes the message header childish and less informative. If I have multiple mail accounts under local folders, I have to move the mouse over "You" to see on which mail address the mail was sent. I need more time and work to get informations. With things like these, you cannot fight against Outlook. A serious mail client shows the full name, or the mail address, if no name is present. It's not up to the mail client to decide, who I am. I think a lot of problems are related to that "You" thing, including localization issues.

Comment 15

[Bryan Clark (DevTools PM) [:clarkbw]]

Dan: I think we could also have some success with always showing the email address when there is only 'You' in the to field.

Part of the design is to save space and be able to show more addresses on the single line that we have available.  You don't need to read your whole name or have it use up the small amount of space available when there are a number of people in the list.  However when there is only one person ('You') we have a good amount of space left and could show the whole email address.

Comment 16

[Jens Müller (:tessarakt)]

(In reply to comment #15)
> However when there is only one person ('You') we have a
> good amount of space left and could show the whole email address.


And, of course, the whole name ...

Comment 17

[rsx11m]

A possible alternative to finding some heuristics which tries to consider all scenarios and local etiquette rules may be to just make it easier to switch between the collapsed and expanded addresses. I could imagine an addition to
the right-click context menu toggling mail.showCondensedAddresses to switch between these two views. The default setting may/should be localizable.

Thus, a user may see the "You", doesn't like it, right-clicks on it to see what he or she can do about it, finds "√ Abbreviate Known Addresses" in the menu and unchecks it, then gets the full addresses. This would still link the display of "You" with the address-book preference though, but based on comment #15, the underlying intention for the condensed addresses is the same anyway.

Comment 18

[Alex Ihrig [:Thunderbird_Mail_DE]]

(In reply to comment #15)
> Part of the design is to save space and be able to show more addresses on the
single line that we have available. You don't need to read your whole name

If we want to save space, we could delete the whole "you". IMHO "You" is superfluous. For what do I need to read, the mail was sent to me ("You")? I received the mail - so it was sent to me, or not?
IMHO, if we show an information, we should show something meaningful -> the full email address or at least the display name.

Comment 19

[Mike Cowperthwaite]

I agree with Joachim and Alexander, this change is not good.  I'm not opposed to having the new behavior available, but I'd like the option to revert to the previous behavior, all the time.  The biggest problem with "You" is that there are several addresses that qualify, and I want to know which it is without having to hover.

However, I note that in the 0327-b3pre build I've started up to re-examine the behavior, I'm not seeing the "You."  If this persists, then all is well.  I wouldn't mind have full name + address even in "hidden details" mode, rather than name-only -- again, so I can see which address the mail was sent to without having to reach for the mouse.

Comment 20

[Mike Cowperthwaite]

(In reply to comment #19)
> However, I note that in the 0327-b3pre build I've started up to re-examine the
> behavior, I'm not seeing the "You." 

Duh.  I had the "Display name only" option turned off.

Comment 22

[Dan Mosedale (:dmosedale, :dmose)]

As per <http://weblogs.mozillazine.org/dmose/archives/2009/06/thunderbird_compact_header_mov.html>, the compact header is no longer part of the Thunderbird core, so this bug is no longer valid.

Comment 23

[rsx11m]

Reopening, this bug is not about the compact view of the header pane but the compact addresses when "Show only display name for people in my address book" is checked.

Still reproducible as stated for Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.9.1pre) Gecko/20090625 Shredder/3.0b3pre BuildID=20090625031837 nightly for the expanded header pane.

Comment 25

[Dan Mosedale (:dmosedale, :dmose)]

As was pointed out in bug 519675, this bug is more prominent now that we have Smart Folders on by default.

Comment 26

[Ben Bucksch (:BenB)]

Specifically, bug 519675 proposes to show the email address whenever there are multiple identities configured.

The current impl is a real problem when I have multiple roles (e.g. private vs. work, working for 2 companies, or customerservice@ vs. press@) and want them separated. In this case, I often want or *need* (e.g. press@) to know which address the email went to.

Comment 27

[Bryan Clark (DevTools PM) [:clarkbw]]

The suggestion to turn off the You when there are multiple identities makes a lot of sense.  The major design feature of the "You" is to save space and be more familiar.  With multiple accounts / identities it either needs to be improved to help identify the differences or more simply turned off.

Comment 32

[Jim Porter (:squib)]

I feel like I said this somewhere else, but I can't find it: we could use the address book to let people provide short names for themselves (i.e. they add their various email addresses to their address book and give them descriptive names). This is really just a matter of changing the order of a conditional somewhere (lower the priority of "You" so that addr book names are preferred). 

This also has the benefit of eliminating a bug where you can enter a name for yourself in the inline "edit contact" popup; it will change the "You" immediately, but it's reverted if you select another message.

On some level, this may be related to bug 474721.

Comment 33

[rsx11m]

Given the fairly significant number of duplicates I've nominated this for 3.1.

Some possible solutions have been pointed out here. The simplest one coming to my mind would be a context-menu option "Show full addresses" to toggle the mail.showCondensedAddresses preference to "false", then change that string to "Display name only" to revert to the default behavior of "true" again.

That would make the option at least more accessible for users who want to avoid this simplification (though not a fully satisfactory solution yet).

Comment 34

[Jim Porter (:squib)]

Attachment: Gives address book entries higher priority than "You"

Here's a patch on trunk that does what I mentioned above.

Comment 35

[Ben Bucksch (:BenB)]

While I think that's am improvement, please consider the suggestions in bug 519675 (dup of this). They would work without the user having to do anything, much less realize and resort to such tricks.

Comment 36

[Ben Bucksch (:BenB)]

Quote (part):
Easiest would be to just use email addresses [if there's more than one identity in total]: "You <[e-mail]>".

[Nicer would be to] generate a name, by looking at all identity email addresses and use only the local part, only the domain, or the full address, depending on what differs between them, e.g. [e-mail] and [e-mail] and [e-mail] -> "You (gmail.com)" and "You (verizon.net)" and "You (hotmail.com)" or [e-mail] and [e-mail] -> "You (sales)" and "You (info)".

But as said, just using "You <" + email address + ">", if there's > 1 identity, is easiest.

Comment 37

[rsx11m]

I have to agree with comment #35 to the extent that the necessary action needs to be easily discoverable by the user. Jim's suggestion is straight-forward for those who know but requires to associate "Add to my addressbook" in the context menu with "Click this option to replace the 'You' with something more suitable in the addressbook card" as the desired action for the unsuspecting user.

Comment 38

[Jim Porter (:squib)]

Attachment: As with first patch, but also shows email address when >1 identity

Ok, this patch takes the easy route and adds an email address when there's more than one identity. It also keeps the "address book is higher priority" from the earlier patch, since 1) the old way never made sense to me, and 2) it lets obsessive people like me provide shorthand like "You (Work)". :)

Comment 39

[Olaf Fraczyk]

(In reply to comment #36)
> Quote (part):
> Easiest would be to just use email addresses [if there's more than one identity
> in total]: "You <[e-mail]>".
> 
> [Nicer would be to] generate a name, by looking at all identity email addresses
> and use only the local part, only the domain, or the full address, depending on
> what differs between them
> (...)

Hmm, why would it be nicer?
I don't see any improvement over using full email address. Additionally the latter gives predictable results.
What would you do for:
[e-mail]
[e-mail]
[e-mail]
[e-mail]

In my opinion it is making simple things complicated and error-prone. 
Please tell me what problem solves using "you" in firts place? If something works OK, then why change it?

Comment 40

[Ben Bucksch (:BenB)]

> Hmm, why would it be nicer?

Shorter, thus less mental burden for processing.

> What would you do for:
> [e-mail]
> [e-mail]
> [e-mail]
> [e-mail]

The algo proposed in comment 36 would show the full address.

Comment 41

[Olaf Fraczyk]

If it is a mental burden to read an email address (my own in this case - sic!) then I need a visit in a hospital soon ;)

BTW. If something is shorter, it doesn't mean that the time needed for our brain to process it will be shorter.

Comment 42

[Joachim Kock]

The patches proposed recently are all a bit convoluted,
since they mix two independent issues (both concerning
only the case where the user has checked the prefs flag
"Show only display name for people in my address book").

The first bug is that if the owner of the addressbook
sets his display name to some string, then this string
is not displayed.  Instead the string "You" is displayed,
disrespecting the user's setting.

The second "bug" is independent of the "You" bug.  It is
simply the fact that display names are not unique, so if 
two email addresses are associated with the same display
name, you cannot immediately see in the header pane which
one it is (except by hovering over the display name).  It
seems to me that this is not really a bug, but rather an
evident tradeoff of checking the prefs flag in the first
place: the user asks for this ambiguity.


The first bug is solved very easily by removing the
clause

if (emailAddress == identity.email) {
   displayName = gMessengerBundle.getString("headerFieldYou");
}

from the file mail/base/content/msgHdrViewOverlay.js.

This simply avoids any substitution, hence respecting the
user's setting.

As already observed, this does not remove any functionality
from the interface whatsoever: the user who wishes to see
himself displayed as "You" will simply enter this string
as display name in his address book card.  This is as obvious
as the whole idea of display names, and the user will get
exactly what he asked for when he checked the prefs flag.

Furthermore this solves the issues with tone and localisation:
if a user likes the idea of a pronoun but is offended by "You",
he can write "Me", "Your Highness", or whatever he likes.


The second problem, to disambiguate display names, is about
any entry in the addressbook, not just the owner's.  It is
great if somebody wants to refine this feature, inventing
ingenious nested conditionals and append disambiguating
strings to the display names.

Whatever you come up with to disambibuate, it should not involve
substituting an immutable program-defined string for the user's
choice, so the two issues are really independent.

I hope the "You" issue can be fixed quickly, and not be held
back by the second issue.  Perhaps the second issue should be
filed separately.

Cheers,
Joachim.

Comment 43

[rsx11m]

(In reply to comment #42)
> The patches proposed recently are all a bit convoluted,
> since they mix two independent issues (both concerning
> only the case where the user has checked the prefs flag
> "Show only display name for people in my address book").

Actually, there are three different issues with the "You" substitution:

(1) Making the "Off switch" discoverable (currently tied to the address book);
(2) solving the ambiguity issue (the patch posted by Jim);
(3) ability for locales to opt out of this feature if culturally problematic.

It is not uncommon for a complex issue to be solved in steps rather than all at once. I've provide a suggestion for (1) in comment #33, Jim and Ben addressed the ambiguity issue (2) with the patch in attachment 422914 [details] [diff] [review]. Given a
fairly short time frame for 3.1 (feature freeze per current schedule in less than three weeks from now!), and that solving either of those issues would be an improvement over the current implementation, going forward with those which are feasible to be solved and deferring what's left to post 3.1 should be not the optimum but a reasonable way to go.

> The first bug is solved very easily by removing the
> clause ... from the file mail/base/content/msgHdrViewOverlay.js.

It's no longer there in that specific form, but anyway. While I fully agree with you that the "You" substitution is an unnecessary feature (and there apparently wasn't much discussion whether or not it should be introduced), historically there isn't much chance that the people at the review switches would accept a patch which is simply rolling back a feature that was recently introduced. Thus, the suggestions are intended to ease the pains it causes.

> Furthermore this solves the issues with tone and localisation:
> if a user likes the idea of a pronoun but is offended by "You",
> he can write "Me", "Your Highness", or whatever he likes.

That's issue (3), partially resolved by bug 481966 now allowing different strings for different headers. Those are still hard-wired though, and the localizers aren't given an option to /not/ specify a substitution string.
So, this would need further improvement, here or in another bug.

Either way, it's certainly good to see some activity in this bug again!

Comment 44

[Olaf Fraczyk]

(In reply to comment #43)
> (...)
> historically there isn't much chance that the people at the review switches
> would accept a patch which is simply rolling back a feature that was recently
> introduced. Thus, the suggestions are intended to ease the pains it causes.
Hmm,
If it was recently introduced and from the beginning here are objections, then it should be higher chance to get it removed. It would seem logical that a feature which is present for a long time would be harder to be removed, as it has proven its usefulness etc. :)
But I understand that "historically" doesn't have to be equal with logically ;)

Comment 45

[Joachim Kock]

> While I fully agree with you that the "You" substitution is an
> unnecessary feature (and there apparently wasn't much discussion
> whether or not it should be introduced), historically there
> isn't much chance that the people at the review switches would
> accept a patch which is simply rolling back a feature that was
> recently introduced.  Thus, the suggestions are intended to ease
> the pains it causes.

I can't believe this!  It would be a huge blow to the 
credibility of Thunderbird's development scheme: anybody with
write access to the repository could introduce whatever funky
feature he likes, and then it can never be removed!  Only
"improved upon" with endless patches...

By the way, who ever wrote that "You" substitution clause did
not really introduce any new feature: even without the clause,
anybody could achieve exactly the same effect by writing "You"
as display name of his addressbook record.  The clause only
caused an OBSTRUCTION to the feature, preventing the user from
having the string displayed that he set in his addressbook.
It was a bug to write this clause in the first place, and reverting is
the only reasonable solution.  All the complications with tone
and localisation go away by reverting.  This is a user string
not a system string.  What's the use of having the display name
feature at all if the system overrules the user's choice?

I can't believe what you say about the review switches.
I still have faith in Thunderbird -- it's a marvelous email
client.

Comment 46

[Phillip]

With all of the problems regarding this "feature" why isn't it simply undone? People understand their own email address and don't need the word "You" to comprehend that the email was sent to you. Showing "You" just hides relevant and important information. Meanwhile, I will uncheck "Show only display name for people in my address book" but I feel like the average user will not know about this option (I didn't know about this option until I saw this bug report). For the record, I hate how the GMail web interface shows "You" and Thunderbird was wrong in mimicking this behavior.

Comment 47

[rsx11m]

The workaround of unchecking that box isn't easy to find, hence my suggestion to add a respective option to the context right-click menu. That's probably the first place a user would look for a way to change the display.

(In reply to comment #45)
> I can't believe this!  It would be a huge blow to the
> credibility of Thunderbird's development scheme: anybody with
> write access to the repository could introduce whatever funky
> feature he likes, and then it can never be removed!  Only
> "improved upon" with endless patches...

Though theoretically possible, it's hopefully not quite like that, and it wasn't my intention to stir that discussion but to point out how it has been
in some other cases of recently introduced features. Fact is that the module owners and peers have the final say what does and what doesn't go into the code, and previous code was removed in cases where it was considered obsolete or no longer adequate. There have been features (e.g., related to the header pane and the toolbar buttons) which were introduced to present them to the user
as novelty. While those changes were not removed upon protests, a Migration Assistant was introduced to allow retaining customizations and old folder and toolbar defaults, etc.

It appears that Dan's and Bryan's comment #25 and comment #27 would support a partial solution of the issue as proposed in the patch along these lines.

Comment 48

[Jim Porter (:squib)]

Honestly, as someone who uses multiple identities extensively in TB, I've never had a real problem with "You", since all I care about is that I reply with the appropriate "From:" address. I do, however, think it makes more sense as a fallback from user-entered data (hence my patch).

To get something like this removed, you'd have to convince people that it *can't* be fixed (as opposed to merely not having been fixed yet). Since I wrote a patch "fixing" it, I think my stance is clear. :)

However, it's also an issue that the "Show only display name
for people in my address book" checkbox is difficult to find. Clearly, not everyone wants short names in the email header, and this setting should be discoverable by users. That's a separate bug, though.

Comment 49

[Olaf Fraczyk]

(In reply to comment #48)
> (...)
> However, it's also an issue that the "Show only display name
> for people in my address book" checkbox is difficult to find. Clearly, not
> everyone wants short names in the email header, and this setting should be
> discoverable by users. That's a separate bug, though.
Yes, I also found it by filling a bug reprt :)
Maybe changing the default to "off" would solve this problem?
Then the only affected users would be these who asked for it.

Comment 50

[Bryan Clark (DevTools PM) [:clarkbw]]

(In reply to comment #38)
> Created an attachment (id=422914) [details]
> As with first patch, but also shows email address when >1 identity
> 
> Ok, this patch takes the easy route and adds an email address when there's more
> than one identity. It also keeps the "address book is higher priority" from the
> earlier patch, since 1) the old way never made sense to me, and 2) it lets
> obsessive people like me provide shorthand like "You (Work)". :)

Jim, you should put this up for review.

Comment 51

[Jim Porter (:squib)]

Comment on attachment 422914 [details] [diff] [review]
As with first patch, but also shows email address when >1 identity

Oh yeah, I knew I forgot something!

Comment 52

[Bryan Clark (DevTools PM) [:clarkbw]]

Comment on attachment 422914 [details] [diff] [review]
As with first patch, but also shows email address when >1 identity

Seems like a good step forward for this feature.

Comment 53

[Dan Mosedale (:dmosedale, :dmose)]

If this were the last bug standing for Tb 3.1, I don't think we'd hold the release for it.  That said, I'm excited that there's a patch here and I'd love to see us land this in 3.1.  Jim, care to whip up a MozMill test for this baby?

Comment 54

[Jim Porter (:squib)]

Sure.

I notice that there's mail/test/mozmill/shared-modules/test-folder-display-helpers.js, which has code for injecting local messages/folders. I need to inject imap (or pop) messages, though. My plan was to add a parameter for options to the installInto function that (eventually) passes {mode:"imap"} to configure_message_injection (part of the xpcshell tests, I gather). However, it doesn't work ("load is not defined" at messageInjection.js line 143). I don't actually have any idea how xpcshell tests work, let alone how the MozMill tests talk with them. Is there something obvious I'm missing?

Comment 55

[Jim Porter (:squib)]

Nevermind, I figured it out. I just need to finish the test case for when you're in the address book.

Comment 56

[Jim Porter (:squib)]

Ok, I did find one problem with my patch: if you send an email to yourself (and have addresses you send mail to go into your collected address book), the "You" is replaced with the display name in the address book or the email if there is no display name. I guess there are two ways to handle this:

1) Go back to the old way where "You" is higher priority than the address book (this still seems strange in my opinion)

2) Change it so that if your address is in the address book but there's no display name, it still displays "You".

Any ideas? I prefer (1) from a development POV (the test is easier), but I'd rather have (2) as an end-user.

Comment 57

[Jim Porter (:squib)]

Attachment: Patch + MozMill tests

After thinking about it, I went with (1) above. I think a better way to handle multiple identities would be to give each identity a name (not the display name) and use that to disambiguate. For example, you might have an identity named "Work", so you'd get "You (Work)" in the message header. You could also apply this to the From field in the message composer.

But that's a different bug. :)

Comment 58

[Joachim Kock]

Concerning precedence:

If the user has taken the trouble of specifically entering a 
display name for himself in the addressbook,
please do not override his choice with the "You" string.

(If "You" is what the user wanted to be displayed,
he would just have entered that string as display name.)

Comment 59

[Jim Porter (:squib)]

I haven't really managed to convince myself either way is the "right" way. My main concern is that a user could end up with one of his identities in an address book without intending it (sending an email to himself, clicking the star). That part is fixable by falling back to "You", since there's no display name.

However, if the user has two identities and someone, say, sends an email to both of them (with display names filled in), the default reply action is "reply all", meaning he'll send from identity A to 1) the original sender, and 2) identity B. That means identity B ends up in the address book with a display name determined by someone else, and through no fault of the user (except not paying enough attention when he replied).

This is really part of a larger issue of whether to trust the display name in the address book when there's a naming conflict: bug 474721.

That said, I did technically get UX approval for the case where the address book display name is prioritized; maybe the example I described is rare enough that we don't have to worry about it. clarkbw, do you have any thoughts about this in light of what I've mentioned?

Comment 60

[Olaf Fraczyk]

I haven't really managed to convince myself that the "you" thing gives any benefit  besides growing bugzilla database :)
I still can't understand the idea of adding unnecessary things and later finding out how to make them not obtrusive.
If soemething is added and it causes problems, it should be reverted, redesigned and then added again.
The current process is like law creation in my country, first we vote a law, next we find that it is idiotic, next we spend 5 years making amendmends, after 5 years we cancell this law and then say very loud - look people - we make your life easier, there is no longer need to follow this stupid law!

Comment 61

[Dan Mosedale (:dmosedale, :dmose)]

Comment on attachment 427218 [details] [diff] [review]
Patch + MozMill tests

Since this is really a UI question, what to do here is clarkbw's call to make.  Requesting ui-review from him.  Note that he's out of the office this week, so it'll probably be a bit before he gets to it.

Comment 62

[Jim Porter (:squib)]

See bug 474721, where I have a patch that should fix both that and this bug, and which hopefully resolves people's issues about the priority of "You" vs. the address book.

Comment 63

[Joachim Kock]

The proposed patch in Bug 474721 solves important issues
by giving the user fine grained control over how display names,
provided names (or no provided names) interact.  The user can
even control this independently for each addressbook entry.

There are many different cases.  It would be great if you could
explain exactly in which cases the "You" string is meant to
appear.

Comment 65

[Mark Banner (:standard8)]

Comment on attachment 427218 [details] [diff] [review]
Patch + MozMill tests

Sorry for not checking in on this earlier.

As the patch here and in bug 474721 appear to need Bryan's input, I'm temporarily cancelling the review request until we get that. I've also contacted Bryan direct to remind him to look at these patches.

Please feel free to re-request review once you've got his ui-review (or ping me direct if you're still not heard anything in a few days).

Comment 66

[Bryan Clark (DevTools PM) [:clarkbw]]

Comment on attachment 427218 [details] [diff] [review]
Patch + MozMill tests

Just reviewed attachment 427448 [details] [diff] [review] of bug 474721 so removing from this as they seem to do the same thing.

Comment 67

[Ludovic Hirlimann [:Usul]]

(In reply to comment #66)
> (From update of attachment 427218 [details] [diff] [review])
> Just reviewed attachment 427448 [details] [diff] [review] of bug 474721 so removing from this as they
> seem to do the same thing.

setting the dependancy so nothing is lost when bug 474721 lands.

Comment 68

[Michael Steel]

Since after more then one year this is still open (don't you love open source? :) ) I wanted to highlight that there is a discrepancy in the behaviour of this "useful" feature.

- I am using Thunderbird 3.1
- My email is NOT in the address book
- Enabling the Recipient column in the message list panel I see the actual email address eg. [e-mail]
- Messages in the preview panel show "To: You", standing on it with the mouse cursor a tooltip would show the actual email address eg. [e-mail]

As for the discrepancy, if you search the message, the search results will show to: followed by whatever is in the header before the actual email address.

Is is a bug or a feature to spot the tricky "**** <[e-mail]>" mentioned in comment# 1? 

Regards

Comment 69

[Michael Steel]

Sorry, I forgot to mention the following:

If you click twice on the star besides "to You" and then click on delete, "to You" gets replaced by the actual email address temporarily.

While I wait for a more self explaining setting, I will stick with the workaround @comment#9.

Regards

Comment 70

[Jim Porter (:squib)]

(In reply to comment #68)
> Since after more then one year this is still open (don't you love open source?
> :) ) I wanted to highlight that there is a discrepancy in the behaviour of this
> "useful" feature.

We're working on it (in bug 474721), but things are a little bit held up due to some organizational issues that affect Seamonkey. If you want to keep up to date on what's happening with this bug, that's where to look.

> - Enabling the Recipient column in the message list panel I see the actual
> email address eg. [e-mail]
[snip]
> As for the discrepancy, if you search the message, the search results will show
> to: followed by whatever is in the header before the actual email address.

This is going to be hard to change, but in general, the goal is to make everything consistent with the header (see bug 312821 for the message list part of the problem).

> Is is a bug or a feature to spot the tricky "Asshole <[e-mail]>"
> mentioned in comment# 1? 

It's a feature, more-or-less. "You" are implicitly in the address book, so your addresses show the address book display name. If you really want to see what the email headers show, you should have "Show only display name for people in my address book" unselected in Options -> Advanced -> Reading & Display. (This could maybe use a better description and more prominent placement.)

Additionally, once bug 474721 lands, anyone who wants something other than the "official" Thunderbird way will be able to easily change how the addresses are displayed in the header (but it'll take a couple more patches to make that work everywhere).

(In reply to comment #69)
> If you click twice on the star besides "to You" and then click on delete, "to
> You" gets replaced by the actual email address temporarily.

This is also fixed in bug 474721.

Comment 71

[Michael Steel]

>Additionally, once bug 474721 lands, anyone who wants something other than the
>"official" Thunderbird way will be able to easily change how the addresses are
>displayed in the header (but it'll take a couple more patches to make that work
>everywhere).

Oh I see what you mean now. I did read the thread for bug 474721 but at the time the last comment was #59: I was planning to highlight the preview panel/search discrepancy since the 2nd of June, but I have been busy with other matters.

Sorry for the excessive sarcasm in my previous post, it seems I am getting old faster then I was expecting. :)

Regards

Comment 72

[Jim Porter (:squib)]

(In reply to comment #71)
> Oh I see what you mean now. I did read the thread for bug 474721 but at the
> time the last comment was #59: I was planning to highlight the preview
> panel/search discrepancy since the 2nd of June, but I have been busy with other
> matters.

If you're talking about the new global search tab, I don't think there's a bug report for that one yet. I keep meaning to write a patch for it and then I end up forgetting.

Comment 73

[Jim Porter (:squib)]

This should be (almost) fixed by bug 474721. Bug 590700 might be needed to fix it on Windows, or it might be a bug in the test itself.

Comment 74

[Ben Bucksch (:BenB)]

Not really. I have about 10 identities. I don't want to add an address book entry for each of them (manually), with the email address in the name, just to make the email address appear. We could just show "You <" + email address + ">", if there's > 1 identity.

Comment 75

[Ben Bucksch (:BenB)]

Bug 474721 comment 38 suggests that the patch took care of it and it should work. Unfortunately, it does not, in practice, from my testing. "Prefer display name" seems to be default on, which is good for remote recipients. However, that's also used for my own identities: if I send a mail to myself (e.g. another account) and my own address is added to Collected Addresses. In effect, the "You <address>" is almost never active for me. Can we change that angle of the logic?

Comment 76

[Jim Porter (:squib)]

Do you just get "You" in that case, or do you get something else (e.g. the name+address used by the sender)? In the latter case, that's bug 543701. I'm not sure what would cause the former case, but if you have an address book entry for yourself, that will override the "You" handling entirely.

Comment 77

[Ben Bucksch (:BenB)]

Jim, I get only my name, no email address. Your case 2.2.1 (not in AB, [two identities] -> "You <[e-mail]>") would be my desired outcome, and I think it should be the default.

Comment 78

[Jim Porter (:squib)]

(In reply to comment #77)
"You <[e-mail]>") would be my desired outcome, and I think
> it should be the default.

It is the default, modulo the two things I mentioned above. Anytime "You" is shown, it will append the address if you have more than one identity in Thunderbird. I'm not sure what's causing your issue, but it sounds like it's not using the "You" shorthand to begin with.

Comment 79

[Ben Bucksch (:BenB)]

Jim, I tested again with a fresh profile and 3 email accounts at different providers. I used "ben %provider%" as real name when setting up the account, i.e. as "From". When sending mail to myself (other accounts), I used "ben %provider% sent <myaddress@%provider%.com>" as To. As expected, these "sent" names and addresses get added to my "Collected Addresses" AB.

Then, after sending, I click on the messages, both in Sent folder of outgoing account and Inbox folder of incoming account. In all cases but one, I see in the message header display:
"from  You <myaddress@%outgoing-provider%.com> * (star)"
"to    ben %incoming-provider% sent * (star)"
e.g:
"from  You <[e-mail]> * (star)"
"to    ben yahoo sent * (star)"
That's true for both Inbox and Sent folders, of all accounts, with the exception of one mail. In that one case - folder Inbox>Sent of t-online account -, it's reverted:
"from  ben tonline * (star)"
"to    You <[e-mail]> * (star)"
Note it doesn't say "Sent", so it's taking my "From" realname here, not the AB. 

Anyway, what I would expect is that it would use "You <[e-mail]>" for both "from" and "to" (and cc etc.), in all cases (on all accounts, in Inbox and Sent folder).

Comment 80

[Ben Bucksch (:BenB)]

Tried to send a few more mails. I have at least one more "reverted" case, and also many cases where I see:
"from  ben %outgoing-provider% * (star)"
"to    ben %incoming-provider% sent * (star)"
So, this all looks very odd and unlogical. Just always show "You <m@e>" when it's one of my identities.

Comment 81

[Ben Bucksch (:BenB)]

"Collected Addresses" have the "prefer address book over header" off in the contact card.

Comment 82

[Jim Porter (:squib)]

Yup, this definitely sounds like bug 543701.

Comment 83

[Jim Porter (:squib)]

My comment in bug 543701 that this only applies to addresses in the same domain is probably wrong, although it does *sometimes* work when you have addresses in different domains. Doing "if(getBestIdentity() == emailAddress)" is probably a really bad idea in that function anyway; it should be something more like "if(isIdentity(emailAddress))".

All of these problems look like they boil down to problems with figuring out if a given address is one of your identities.

Comment 84

[Ben Bucksch (:BenB)]

Cool, sounds easy to fix. Do I hear you volunteering? *fishing* :)

Comment 85

[Jim Porter (:squib)]

I uploaded a patch (sans tests currently) to bug 543701 if you want to try it out.

Comment 86

[Jim Porter (:squib)]

Now that all the dependencies for this are checked in, I believe this bug is fixed. I'm going to leave it open for a bit so that people who had specific problems (e.g. :BenB) can comment on whether the issue is resolved for them. If it looks good to other people, we can mark this resolved, I think.

Comment 87

[rsx11m]

(In reply to comment #3)
> (2) The other issue is the one of localization, [...]. Even in a region where
> a casual "You" would be considered appropriate, it may not hit the right tone,
> e.g., in a business environment.

I don't see this issue addressed by any of the patches, i.e., provide a simple way (e.g., by leaving the strings empty in the .properties file) for localizers to opt out of the "You" for their region.

> Is there any feedback yet from the localizers on this point?

On the other hand, nobody from that community commented (at least not here).

Comment 88

[rsx11m]

Comment on attachment 427218 [details] [diff] [review]
Patch + MozMill tests

This patch does no longer apply (Hunk #1 FAILED at 1137) and was apparently covered by bug 543701, thus I'm marking it obsolete.

Comment 89

[Ben Bucksch (:BenB)]

rsx, German (my mother tongue) is one of those regions. Neither "Du" (2. person singular) nor "Sie" (formal respect form) really fits.
But "Ich" (= "Me", i.e. 1. person singular) would work and be a nice workaround.

Comment 90

[John Hopkins (:jhopkins)]

Another issue with the "You" label is in my case where I'm receiving emails for multiple addresses (via forward-only addresses) so I'd like to see - without clicking or hovering - which address the email was sent to.  Currently Thunderbird says "You" for all my email accounts which I find inconvenient.

Comment 91

[rsx11m]

(comment #90) That part works for me on trunk builds, it still shows "You" but with the address following in "<...>", so this ambiguity is gone now.

Comment 92

[Bryan Clark (DevTools PM) [:clarkbw]]

(In reply to comment #89)
> rsx, German (my mother tongue) is one of those regions. Neither "Du" (2. person
> singular) nor "Sie" (formal respect form) really fits.
> But "Ich" (= "Me", i.e. 1. person singular) would work and be a nice
> workaround.

Don't we have another bug for switching from "You" to "Me"?  If it makes for better translations I think it's a good change.

Comment 93

[rsx11m]

That's bug 535349 you are thinking of, but that's strictly speaking not the topic of this bug. While there are various workarounds to replace the "You" and the fix in bug 543701 makes it less ambiguous by adding the e-mail address, the remaining question to address here is whether or not a general "off" switch based on locale default or user preference is necessary.

Comment 94

[Jim Porter (:squib)]

As a workaround, you can add yourself to the address book and put whatever name you want in for "You", and it will show up in the message header like that.

Comment 95

[rsx11m]

Jim, I know about the workarounds, I'm trying to bring the topic back to the original reason this was filed and confirmed for... ;-)

Comment 96

[rsx11m]

(In reply to comment #94)
> As a workaround, you can add yourself to the address book and put whatever name
> you want in for "You", and it will show up in the message header like that.

So I've tried this workaround, pretending that I'm a novice user:

 1. Set up your account;
 2. receive a message or look at one you've sent yourself;
 3. this shows up as "You <address>" with an empty star;
 4. let's say I don't like that and right-click on the "You" address;
 5. there is an item "Add to address book", so I try and click on it;
 6. nothing changes, except that the star for the "You" turns yellow
    (I see the same non-change when clicking on the star, too);
 7. only after going to a different message I see the new display name.

Following issues with this procedure:

 a. It is virtually impossible to discover this workaround, given that there is
    no hint that adding to the address book will avoid the "You" and that the
    effect isn't immediately visible.

 b. If I don't click on the "From" address of my own message, I'll pick up any
    version of the header sent to me by someone else as replacement for "You"
    for that identity (which may be different from my original display name).

So, following suggestion: If someone right-clicks on an own identity replaced with the "You" display name, make "Add to address book" rather "Use full name" and add the configured name of that identity to the address book rather than whatever is associated with that address in the received message. This would likely require an extension of the binding in some sort to allow feeding back a boolean value that for this address the "You" display name has been triggered.

Comment 97

[David E. Ross]

What is the difficulty in implementing this RFE as originally stated?  

As already noted, some of us have several E-mail addresses.  In my case, I might receive a message for D. E. Ross at one address and David E. Ross at another, because that is the way my outgoing messages indicate From addresses.  I want to see what address was used on a message to me WITHOUT having to fiddle with additional buttons or menus.

Comment 98

[Jim Porter (:squib)]

(In reply to comment #97)
> What is the difficulty in implementing this RFE as originally stated?

To fix the problem stated in comment 0, this has always been possible (Tools -> Options -> Advanced -> Reading and Display -> uncheck Show only display name for people in my address book).
 
> As already noted, some of us have several E-mail addresses.  In my case, I
> might receive a message for D. E. Ross at one address and David E. Ross at
> another, because that is the way my outgoing messages indicate From addresses. 
> I want to see what address was used on a message to me WITHOUT having to fiddle
> with additional buttons or menus.

Again, you can see the address without fiddling with anything. If you have multiple identities, they will show up as "You <[e-mail]>", so there is no ambiguity (note that this only applies to Thunderbird 3.3). The only reason this bug is still open is to address possible I18N issues with "You" being harder to translate than "Me".

If you'd prefer to see "David E. Ross" or "D. E. Ross" without the address, you can do this too, but it requires writing a (simple) extension. One of the primary goals in bug 474721 was to ensure that, if people didn't like the way Thunderbird handled display names, they could easily override this by overriding the "FormatDisplayName" function.

I don't see a lot of value in providing options for *every* variation on display names as Thunderbird core code; there's a point at which the added testing and maintenance cost overshadows the added utility.

Comment 99

[David E. Ross]

Whether I have "David E. Ross" or "D. E. Ross", the actual address is always <[e-mail]>.  So if I see "You <[e-mail]>" on a reply to a message I sent, I still can't tell which display name (as described in RFC 5322) was used.

Comment 100

[Ben Bucksch (:BenB)]

look you dep bug, look you attach, you see will:
<https://bug474721.bugzilla.mozilla.org/attachment.cgi?id=427450>
Worky worky or no worky that?

Comment 101

[Jim Porter (:squib)]

(In reply to comment #99)
> Whether I have "David E. Ross" or "D. E. Ross", the actual address is always
> <[e-mail]>.  So if I see "You <[e-mail]>" on a reply to a
> message I sent, I still can't tell which display name (as described in RFC
> 5322) was used.

Again, this can be addressed by unchecking "Show only display name
for people in my address book" or overriding FormatDisplayName if you absolutely need to see the display name but not the address.

If you think there should be a separate setting for "Abbreviate my email addresses as 'you'" (I'm not convinced either way), it would probably be best to make a new bug report to talk about just that issue and mark it as blocking this one. The original report is really about several issues (multiple identities, localization, UI inconsistency, seeing what display name people use for you), and trying to resolve them all here gets unwieldy.

(In reply to comment #100)
> look you dep bug, look you attach, you see will:
> <https://bug474721.bugzilla.mozilla.org/attachment.cgi?id=427450>
> Worky worky or no worky that?

With the checkbox *unchecked*, that will show "You" for your identities. This was intentional, since I didn't want the "You" to be disabled just because you sent an email to yourself and ended up adding your email to the collected address book.

Comment 102

[Ben Bucksch (:BenB)]

> With the checkbox *unchecked*, that will show "You" for your identities.

That doesn't make sense to me. It says "always prefer display name over name in message headers". Default is true (isn't it?). For me, "You" is a display name rather than "message header name".

In any case, this doesn't matter, as long as there is *some* way to disable that "You". A global pref would be preferred. If there isn't any, I agree this bug should add one.

Comment 103

[Jim Porter (:squib)]

(In reply to comment #102)
> > With the checkbox *unchecked*, that will show "You" for your identities.
> 
> That doesn't make sense to me. It says "always prefer display name over name in
> message headers". Default is true (isn't it?). For me, "You" is a display name
> rather than "message header name".

The "display name" in that text field is meant to refer to that specific field, not to a "display name" in general. I agree that the description could be a little better, but I couldn't think of anything at the time that wouldn't be overly long. Maybe an explanatory tooltip would help (like the "Allow remote content" checkbox has).

As for the default, it's complicated. It does default to true, but not for collected addresses for the reason above. It also falls back to the raw header value when it's true and the address book display name field is blank, but that's just a side effect of how the behavior works for non-"You" contacts, and not actually intended as some obscure workaround.

> In any case, this doesn't matter, as long as there is *some* way to disable
> that "You". A global pref would be preferred. If there isn't any, I agree this
> bug should add one.

I think this should be handled in a new bug report, though. If this pref gets added, it'll probably need some people not currently involved to take a look at it, and I don't want to force anyone to read through 100+ comments, most of which don't have anything to do with the specific issue of a toggle to disable "You". :)

Comment 104

[Jim Porter (:squib)]

Created bug 619461 to track this issue.

Comment 105

[Joachim Kock]

In reply to comment #98:

> there's a point at which the added testing and
> maintenance cost overshadows the added utility

It seems clear that this point was reached very very
early: the You feature represents zero added utility,
as pointed out in Comment #42: anybody wanting to 'save
space' or to 'be on cool familar terms with his mail
program' could just write whatever string 
"You ([e-mail])" in his addressbook entry.

In rely to Comment #101:

> The original report is really about several issues 
> (multiple identities, localization, UI inconsistency,
> seeing wat display name people use for you), and 
> trying to resolve them all here gets unwieldy.

That's not true: removing the You feature altogether
would resolve all the issues at once.  It could have
saved a lot of time for many people, and it could have
avoided introducing convoluted code trying to make
everybody happy.  Wasn't everybody happy before that
unnamed programmer suddenly invented that You thing?

It was apparently decided by some highest instance 
Review Switch that this feature cannot be undone.  
I admire the patch writers for their patience in
finding workarounds and patches to mimimise the effect
of the bug.  Thank you for your work.  I have not
tried the patches yet, but I am sure I could live
with the outcome, according to the descriptions I
read.  Cheers, Joachim.

Comment 106

[Jim Porter (:squib)]

(In reply to comment #105)
> That's not true: removing the You feature altogether
> would resolve all the issues at once.  It could have
> saved a lot of time for many people, and it could have
> avoided introducing convoluted code trying to make
> everybody happy.  Wasn't everybody happy before that
> unnamed programmer suddenly invented that You thing?

There's no reason to disable the "You" feature for everybody just because some people don't like it. Some people aren't affected by the problems in this bug, some people are but still want the "You" feature, and some people don't like it at all. The work in this bug so far has been to improve things for the second group of people.

Saying "I don't like the 'You' shorthand" is different from "The 'You' shorthand doesn't work right", which is why I spun off a separate bug to disable it for people who just plain don't like it.

Comment 107

[Joachim Kock]

> There's no reason to disable the "You" feature for everybody
> just because some people don't like it.

Anybody who would like any particular substitution, be it "You",
"You (s@b)", "Your Highness" or anything else could achieve
this (by setting his display name) without the crazy hard-coded
substitution forced upon us all by a programmer who certainly
(as is obvious by now to everybody) did not think his idea through.

It would be fairer to say that there was really no reason to
force this upon everybody just because a few people found it cute.

Comment 108

[Jim Porter (:squib)]

(In reply to comment #107)
> > There's no reason to disable the "You" feature for everybody
> > just because some people don't like it.
> 
> Anybody who would like any particular substitution, be it "You",
> "You (s@b)", "Your Highness" or anything else could achieve
> this (by setting his display name) without the crazy hard-coded
> substitution forced upon us all by a programmer who certainly
> (as is obvious by now to everybody) did not think his idea through.

As far as I know, it's always been possible to see the "proper" header value by disabling "Show only display name for people in my address book", so this certainly wasn't "forced upon" users. I'm sympathetic to people wanting to be able to customize how the shorthand works (this is why I filed bug 619461 and made sure that FormatDisplayName can be easily overridden by extensions), but this never eliminated a feature, it just added one that some people didn't like and that they had to turn off.

That said, this discussion really isn't helping your issue get resolved. If you want to disable the "You" part but keep the rest of the shorthand, vote for bug 619461 because that's where I (or possibly someone else) will be adding a pref to disable "You".

Comment 109

[Michael Steel]

Not to throw gasoline on fire, but since "Show only display name for people in my address book" is obscure, almost hidden but more important enabled by default, the whole "You" "Me", "Whatever" idiocy was actually forced upon us all.

Of course one could argue that if the option was turned off by default other people would have been complying about it. Still this discussion was opened on FEBRUARY 2009, now we are at December 2010 and a solution that works following commons sense and logic has not yet been found/finalised: you draw your conclusions.

Comment 110

[Michael Steel]

When I say "complying" I mean "complaining", my bad.

Comment 111

[Ben Bucksch (:BenB)]

Please keep it constructive. This is a matter of preference. Some people like it, others don't. The default is what most people like. It's not idiocy at all.

The preference would be bug bug 619461.

Jim, rsx, is there anything left to be done here in this bug?

Comment 112

[Michael Steel]

Ben, I understand your point, still Thunderbird aims at managing multiple mailboxes and identities.
Some people like to have separate structures for each mailbox (I do), some other like to throw everything in the "Local Folders".

What kind of commons sense/logic was debated in the first place when it was decided that "Show only display name for people in my address book" was to be turned on by default?

Just because this is "what most people like" does not seems good enough to me: when you have let's say 12 different mailboxes accounts in "Local Folders", it seems pretty idiotic to me not being able to properly identify what is what because everything gets labelled as "You".

I could understand when you have a single email account, not when you have more than one, that would be the most common case (I am thinking HOME/WORK).

To keep it constructive: at least the option should turn off itself when more than one account is detected.

Anyway it is not easy to be constructive. Long time ago I reported a regression related to junk filtering and data loss, I was very detailed and it was a critical bug. It took dozens of "me too" messages just to be accepted and it was debated forever, just like this "bug".

I have other cases where Thunderbird is still destroying data, but reporting them seems like a waste of time at the moment, knowing that it will be debated forever.

Sorry for stating the obvious.

Comment 113

[Joachim Kock]

> it's always been possible to see the "proper" header value by
> disabling "Show only display name for people in my address book",
> so this certainly wasn't "forced upon" users.

The "Show display name" is a nice feature.  The "You" bug **** up
this feature by OVERRIDING the user's choice of display name with
something fixed, whose multiple backdraws have been explained over
the past two years.

OK, it was not forced upon us: we coud just give up the nice
"Display name" feature; or, we could just abandon Thunderbird.
Thank you for pointing out the alternatives.  Currently (for two
years now) I am accepting not to use display names.  I will not
give up Thunderbird, because in spite of some glitches, it is
still far better than any other email client.

I am sorry if you feel I am abusing by continuing my comments;
I try to limit myeslf t oreplying when I read something 
directly misleading, like the quote above.

I look forward very much to having all these complications
stowed away in a preference flag "Override display name with
pronoun".

Comment 114

[rsx11m]

(In reply to comment #111)
> Jim, rsx, is there anything left to be done here in this bug?

It appears to me that this has morphed into a meta bug rather, and given that arguments keep being repeated, it's indeed getting long.

Thus, let's summarize what has and what hasn't been fixed so far:

 - There is no dedicated preference to switch off the feature (bug 619461),
   only a rather obscure workaround exists (current TB 3.0/3.1 releases) to
   switch off short display names altogether along with "You" (comment #9).

 - Localizers can work around to a certain extent by using "Me" rather than
   "You" (comment #89), which may be harmonized by bug 535349; if bug 619461
   gets implemented, localizers could switch it off by default in their l10n
   preference overrides.

 + Ambiguities for multiple own addresses have been resolved by adding the
   actual address behind the "You" (bug 474721, bug 543701; works on TB 3.3).

 + Users can override the "You" by any screen name they choose using the
   address book (comment #94, bug 474721; works on TB 3.3).

 - The option of applying the addressbook workaround is hardly discoverable
   and may pick up unwanted display names from incoming messages; comment #96
   lists the steps and suggestions to make it better discoverable (no bug yet).

Thus, my suggestion is to keep this open as a meta bug at least, and close it once all dependencies are satisfied.

Comment 115

[Wayne Mery (:wsmwk)]

If there are UX keywords which are relevant to any of these bugs it would be good for those who care to add them, as it adds clarity to the goal and the fix. (but don't just plaster arbitrary UX keywords) https://bugzilla.mozilla.org/describekeywords.cgi#ux-affordance // http://uxmag.com/strategy/quantifying-usability

FWIW - the bug query https://bugzilla.mozilla.org/buglist.cgi?query_format=advanced&short_desc=you&short_desc_type=allwords&type0-0-0=nowords&value0-0-0=count%20counts&component=Message%20Reader%20UI&component=Search&resolution=---&product=MailNews%20Core&product=Thunderbird

Comment 116

[Michael Steel]

Any chance that the "Show display name" will be moved in some more prominent place (eg. the address book status bar or the Options/display/formatting tab) before Thunderbird 5 ships?

Comment 117

[Jim Porter (:squib)]

(In reply to comment #116)
> Any chance that the "Show display name" will be moved in some more prominent
> place (eg. the address book status bar or the Options/display/formatting
> tab) before Thunderbird 5 ships?

Almost none. (Personally, I don't think it should *ever* be moved.)

Comment 118

[Seungbeom Kim]

(In reply to Ben Bucksch (:BenB) from comment #111)
> Please keep it constructive. This is a matter of preference. Some people
> like it, others don't. The default is what most people like. It's not idiocy
> at all.

So, where is the evidence, supporting that most people like the "You" thing, justifying this "feature"?

It has been stated more than enough times that this is causing a lot of problems for those who don't want it, when those who want it could achieve the same effect very easily with existing features, without this hard-coded overriding; I almost feel I don't need to add anything. But on the other hand, it seems to be never enough.

Comment 119

[Magnus Melin [:mkmelin]]

Bug 1790946 fixed most cases where it cause confusion.

Comment 120

[Wayne Mery (:wsmwk)]

This issue should be tested in the 115 environment becaues the message reader is being largely rebuilt for version 115. 115 becomes available in July. Testing beta 3-4 weeks from now is possible.

Note also this bug blocks meta Bug 456814 which is being closed.

Comment 121

[Wayne Mery (:wsmwk)]

(In reply to Magnus Melin [:mkmelin] from comment #119)

Bug 1790946 fixed most cases where it cause confusion.

Or, perhaps this can be closed?

Comment 122

[Thomas D. (:thomas8)]

(In reply to Wayne Mery (:wsmwk) from comment #121)

(In reply to Magnus Melin [:mkmelin] from comment #119)

Bug 1790946 fixed most cases where it cause confusion.

Or, perhaps this can be closed?

Yeah, Bug 1790946 looks like it will go a long way in solving the main concern of this bug for Thunderbird installations with multiple identities and some other cases where this matters most.

Thanks much to everyone who contributed their feedback here, which is appreciated. Sorry for taking long, but the core problem of this bug has finally been addressed in Bug 1790946, which is available on Beta and Daily channels. This improvement will be available on release channel in the next release, Thunderbird Supernova (115), expected around midyear 2023. For anything beyond that, kindly file a new bug.