29593 - [autoconfig] pre-populate the users fullname in the new mail account wizard

Status: RESOLVED FIXED

Description

[(not reading, please use [e-mail] instead)]

use nsIUserInfo to pre-populate the users fullname in the new mail account
wizard.

eventually, nsIUserInfo will be extended to pre-populate more fields in the new
account wizard.

note, the implementations of nsIUserInfo are not checked in yet.

Comment 1

[(not reading, please use [e-mail] instead)]

marking m15, so that the owners of 29594 and 29596 know when I'd like to check
in.

I have this working for unix with nsUserInfoUnix and I've got the changes to
aw-identity.js (in my local tree) so that the pre-population works on unix.

this is not a high priority bug, and it can be delayed if the mac and windows
implementations get delayed.

Comment 2

[Doug Turner (:dougt)]

the windows implementation of nsUserInfo is done, and I just have to change the 
makefile.win to have it build. 

Comment 3

[(not reading, please use [e-mail] instead)]

fix in hand.

Comment 4

[(not reading, please use [e-mail] instead)]

moving to m14.

Comment 5

[(not reading, please use [e-mail] instead)]

moving out to m17.

this is not a beta2 priority feature.

it works on unix right now, but not mac or windows.

Comment 6

[selmer (gone)]

Mail Review hopes this is post-beta2, easy, polish work.  Marking M18.

Comment 7

[(not reading, please use [e-mail] instead)]

updating status whiteboard.  just waiting for the windows impl.  

Comment 8

[Doug Turner (:dougt)]

I can try to take a look at this some time before m18.

Comment 9

[Doug Turner (:dougt)]

I stoled the wrong bug.  Reassigning to sspitzer.

Comment 10

[(not reading, please use [e-mail] instead)]

not going to get to the win32 implementation for 6.0

moving to future.

Comment 11

[(not reading, please use [e-mail] instead)]

adding mike to the cc list.

I guess I'm the on the hook to get this working for win32.

law, you interested in the win32 implementation?

Comment 12

[Bill Law]

Sure.  I'll add it to my/our list for the next train.

Comment 13

[Ludovic Hirlimann [:Usul]]

I'm pretty sure this is now properly covered by the new autoconfig dialogs.

Comment 14

[Magnus Melin]

Not on linux, at least.

Comment 15

[Ben Bucksch (:BenB)]

|nsIUserInfo.fullname| works for me on Linux.
Mac looks implemented, too, but can't test.
Win32 is still not implemented. :-(
<http://mxr.mozilla.org/comm-central/source/mozilla/toolkit/components/startup/src/nsUserInfoWin.cpp>
Filed bug 581508 about Win32.

Comment 16

[Ben Bucksch (:BenB)]

Bug 561531 is about username.
(Neither username nor email address should be filled in blindly from the OS, because it's very likely to be wrong (e.g. "ben@(none)" for me), and definitely wrong for users with multiple email accounts. That bug asks to allow *specific* configs for some domains to use the OS username as placeholder.)

Comment 17

[Magnus Melin]

Created attachment 690524 [details] [diff] [review]
proposed fix

Comment 18

[Blake Winton (:bwinton) (:☕️)]

Comment on attachment 690524 [details] [diff] [review]
proposed fix

Huh.  It knew my name.  That's equal parts cool and creepy.  :)

>+++ b/mail/components/newmailaccount/content/accountProvisioner.js
>@@ -249,16 +249,23 @@ var EmailAccountProvisioner = {
>+    if (!name) {
>+      try {
>+        let userInfo = Components.classes["@mozilla.org/userinfo;1"]
>+                                 .getService(Components.interfaces.nsIUserInfo);

I'ld prefer
        let userInfo = Cc["@mozilla.org/userinfo;1"].getService(Ci.nsIUserInfo);
or perhaps we could even add it to MailServices.js.

>+++ b/mailnews/base/prefs/content/accountcreation/emailWizard.js
>@@ -137,19 +137,27 @@ EmailConfigWizard.prototype =
>+      userInfo = Components.classes["@mozilla.org/userinfo;1"]
>+                           .getService(Components.interfaces.nsIUserInfo);

It looks like Cc and Ci are available here, too.

>-    this._realname = "";
>+    this._realname = (userInfo) ? userInfo.fullname : "";

If we have a userInfo object, is it guaranteed to have a non-null fullname?

r=me with those two nits fixed, and that question answered.

Thanks,
Blake.

Comment 19

[Ben Bucksch (:BenB)]

> That's equal parts cool and creepy.  :)

We try to please.

Comment 20

[Magnus Melin]

http://hg.mozilla.org/comm-central/rev/057219336b07 -> FIXED

I looked it up and it looks like the getters needed to be in trycatch too, at least in theory.

Comment 21

[Sebastian Hengst [:aryx][:archaeopteryx] (needinfo on intermittent or backout)]

This causes perma-oranges:

TEST-UNEXPECTED-FAIL | c:\talos-slave\test\build\mozmill\account\test-mail-account-setup-wizard.js | test-mail-account-setup-wizard.js::test_mail_account_setup
TEST-UNEXPECTED-FAIL | (runtestlist.py) | Exited with code 1 during directory run
Bug 661906 - Random Orange TEST-UNEXPECTED-FAIL | (runtestlist.py) | Exited with code 1 during test_order_of_address_books TEST-UNEXPECTED-FAIL | c:\talos-slave\test\build\mozmill\newmailaccount\test-newmailaccount.js | test-newmailaccount.js::test_search_button_disabled_if_no_query_on_init
Bug 707105 - Near-perma-orange: TEST-UNEXPECTED-FAIL | test-newmailaccount.js | test-newmailaccount.js::test_get_an_account
Bug 710912 - Permanent orange on Linux: TEST-UNEXPECTED-FAIL | test-newmailaccount.js | test_can_pref_off_account_provisioner TEST-UNEXPECTED-FAIL | (runtestlist.py) | Exited with code 1 during directory run
Bug 661906 - Random Orange TEST-UNEXPECTED-FAIL | (runtestlist.py) | Exited with code 1 during test_order_of_address_books

Comment 22

[Magnus Melin]

backed out http://hg.mozilla.org/comm-central/rev/c8bbd750e1b4

Comment 23

[Magnus Melin]

Created attachment 706858 [details] [diff] [review]
proposed fix, v2

Tests needed updating.
Successful try run: https://tbpl.mozilla.org/?tree=Thunderbird-Try&rev=35ff96164f13

Comment 24

[Blake Winton (:bwinton) (:☕️)]

Comment on attachment 706858 [details] [diff] [review]
proposed fix, v2

Given the successful try run, this is more of an rs=me.  ;)

Thanks,
Blake.

Comment 25

[Magnus Melin]

http://hg.mozilla.org/comm-central/rev/4e99a6a55f80 -> FIXED