324350 - False Indication of Collapsed News Threads

Status: VERIFIED WORKSFORME

(Thunderbird :: Mail Window Front End, defect)

Description

[David E. Ross]

User-Agent:       Mozilla/5.0 (Windows; U; Win98; en-US; rv:1.7.12) Gecko/20050915
Build Identifier: Thunderbird 1.5

View setup:  
  Sort by: Order Received, Ascending, Threaded
  Messages: All
  Threads: Unread

Threads with only a single unread message often show the box with +, indicating a collapsed thread.  When I select the + (changing it to -), there is no thread.  


Reproducible: Sometimes

Steps to Reproduce:



Expected Results:  
A thread with only one message meeting the "View" criteria should either have a box with a - or else not have any box at all.  

This is especially noticeable with the mozilla.support.thunderbird newsgroup.  On the other hand, I don't see this at alt.security.pgp on a commercial NNTP server.

Comment 1

[Lorenzo Bettini]

it happens to me as well, although it shows a - instead of a +; however it only shows one message... if I delete that messages, the other messages of that thread are not shown (although they are counted in the total number of messages).

The only way to make those messages show up again is to set "unthreaded" in sort, switch to another folder and then go back to the first folder.  If the folder is threaded again, the messages are still hidden.

This happens to me with IMAP, and after using offline for a while.

Comment 2

[Wayne Mery (:wsmwk)]

don't think I've seen this with imap in recent times, but I do see this in news. Needless click to open thread is easily avoided if one has the unread count column, but otherwise this might be annoying.

My comments from bug 232562
Circumstances might be when 
a) it's the last message A in the folder before I do ctrl-T to get new messages
b) and after new messages the subject B that follows A has multiple messages,
and thus correctly has a +, and A somehow also gets a +

Comment 3

[Wayne Mery (:wsmwk)]

still see this?

Comment 4

[David E. Ross]

No, I have not seen this lately.  However, I have changed the way I view newsgroups since I submitted this bug report.  Thus, I don't know if the problem was fixed or merely hidden.  I can't remember my old settings, so I can't attempt to recreate the problem.

Comment 5

[Lorenzo Bettini]

as for imap I haven't seen this problem lately

Comment 6

[Tony Mechelynck [:tonymec]]

(In reply to comment #3)
> still see this?

Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.9.1b3pre) Gecko/20090121 SeaMonkey/2.0a3pre - Build ID: 20090121000535 (with EarlyBlue theme).

yes, I see it in newsgroups, though normally I use View => Messages => All and View => Threads => All. Selecting Threads => Unread expands everything, most threads have a triangular twister pointing down; some have a triangular twister pointing right, and clicking that doesn't produce any change. I suspect these are threads with all messages read except one.

Note that I'm on SeaMonkey (this is a Thunderbird bug) so I don't know if Shredder still has it.

Comment 7

[David E. Ross]

Since there are no comments about this still being in Thunderbird, I'm closing this bug report.  

If Tony sees this in SeaMonkey, he should submit a new bug report for that Product, possibly citing this bug report.

Comment 8

[Tony Mechelynck [:tonymec]]

(In reply to comment #7)
[...]
> If Tony sees this in SeaMonkey, he should submit a new bug report for that
> Product, possibly citing this bug report.

Hm? If you mean me, in the current Sm2 trunk version I see all single-message threads with no twisty at all AFAICT, and even if I depart from my usual "All Threads/All Messages" view. Other threads have either twisty shape (in my current theme, triangle pointing right or pointing down) depending on whether they're collapsed or expanded. IOW the bug is gone for me too.

Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.9.1b3pre) Gecko/20090219 SeaMonkey/2.0b1pre - Build ID: 20090219000527

Marking VERIFIED since bug isn't seen anymore on either of the two builds (Tb/W32 and Sm/Lin) where it used to. If anyone sees this again, it might be better to report a new bug, preferably mentioning this one and possibly with the "regression" keyword.