Hello Kevin,
That makes sense, I completely misunderstood those few lines of code.
What really should happen then is that all messages in every thread
should be checked and if a !old message is found, then the thread it
belongs to should be uncollapsed.
I'm also interested in hearing what other devs/users have to say about
this. It may be because I'm relatively new to mutt but I don't really
understand the purpose of uncollapsing all threads when flags have been
externally modified.
Thanks!
Yoshiki.
On (Feb-19-15|12:39), Kevin J. McCarthy wrote:
> Yoshiki Vazquez Baeza wrote:
> > I think I found the reason why this happens, there's a block of code
> > that uncollapses *all* the threads in the current mailbox when it is
> > reopened. The comments around that block of code seem to indicate that
> > only new emails should be uncollapsed, so I've added an if statement
> > (see attached patch) checking for new mails that are unread. This fixes
> > the problem that I was having (all threads would uncollapse when flags
> > were externally modified).
> >
> > Not sure if this is really a bug or the intended behaviour, any feedback
> > is greatly appreciated!
>
> Hi Yoshiki,
>
> I haven't looked at the threading code in mutt at all (so it's possible
> I'm wrong), but the patch doesn't look quite right to me:
>
> > for (h = ctx->tree; h; h = h->next)
> > {
> > - for (j = h; !j->message; j = j->child)
> > - ;
> > - mutt_uncollapse_thread (ctx, j->message);
> > + for (j = h; !j->message; j = j->child)
> > + ;
>
> Here, we've located the first message in the thread.
>
> > + if (!j->message->old && !j->message->read)
> > + mutt_uncollapse_thread (ctx, j->message);
>
> So, here you're only looking at that one message and checking its
> old/read flags.
>
> I think the "reopen" occurs when messages disappear from the mailbox.
> The comment about uncollapsing "new" messages means messages that have
> appeared in the index, not necessarily those with !read.
>
> The sudden uncollapse is certainly not friendly, and there's some room
> for improvement here. It would be great if one of the more experienced
> committers would comment about the behavior and ways it could be
> improved.
>
> -Kevin
