Red Hat Bugzilla – Bug 253730
evolution message refresh loses focus
Last modified: 2008-04-30 08:05:59 EDT
Description of problem:
<@ajax> mbarnes: i have evo set to do ~30 minute checks for new messages, plus
automatic message filtering, and the folder view shows only unread messages.
when it checks for new messages, and some new ones get filtered into the folder
i'm currently in, the message i'm currently on gets marked as "read", and
Even when new messages don't filter into the current mailbox, the view panes
still flash. These might be separate bugs, or might not.
Version-Release number of selected component (if applicable):
I think the flashing is a separate bug. I've seen it reported elsewhere and can
readily reproduce it myself. It seems to only happen with longish messages. I
think GtkHTML is using a refresh timer while it renders the HTML, but it's
erasing itself and then redrawing the whole thing with each refresh. Hence the
For REALLY long messages you'll notice the flashing gets slower as more of the
message is rendered, I guess because it takes longer to draw.
Anyway, for this bug I'll focus on the message list behavior you described.
*** This bug has been marked as a duplicate of 247332 ***
Un-duping. 247332 is about the flashy refresh, this is for the message list thing.
Oh, whoops, sorry. I don't even remember closing this one.
Do you really think this is a bug? The updating folder thing. Evo just every ~30
minutes call update on each folder and then refreshes view. Which is probably
correct, right? I remember I did a patch, included definitely in 2.21.91, which
ensures there always selected any message, when the actual "disappear" (either
deleted, filtered out or anything else). It _tries_ to find the "nearest"
message from the old one.
Is this enough for this issue?
I think Adam's issue is the fact that a selected message in a folder gets marked
as read when new messages get moved into that folder via filters. If your view
is only showing unread messages, the message may disappear on you while you're
still reading it.
*** Bug 439432 has been marked as a duplicate of this bug. ***
Upstream bug for this is
Can we track it there? The problem is much complicated than it can seem to be.
Hmm, OK, I attached a proposed patch to upstream bug, we will see what they say.
I'm closing this as 'upstream'.
A patch for this has been committed upstream. Is it possible to port, or at
least make sure it is updated for F9?
Sure, it's committed to stable branch too, which means next update will include
it. I heard they plan next release on this Friday, so after that we will update
our package too.