Re: [Evolution] Wish: Uncached IMAP folders
- From: "Patrick O'Callaghan" <poc usb ve>
- To: David Woodhouse <dwmw2 infradead org>
- Cc: evolution-list gnome org
- Subject: Re: [Evolution] Wish: Uncached IMAP folders
- Date: Wed, 26 Apr 2006 07:32:40 -0400
On Wed, 2006-04-26 at 01:09 +0100, David Woodhouse wrote:
On Mon, 2006-04-24 at 22:52 -0400, Patrick O'Callaghan wrote:
I'm speechless. Was this done because some IMAP servers were buggy? If
not, there would seem to be no justification for it.
I believe it was done in order to fix inconsistencies in the unseen
counts on folders when the strange client-side Junk processing isn't
disabled. The Junk processing hides messages from a folder and pretends
that those messages actually exist in some other fake folder. And thus
the unseen counts in the real folder looked wrong, because some of the
unseen messages were hidden from view.
I see.
<sarcasm>
So even when I don't have Junk processing enabled, the rest of Evo Mail
is slower and causes extra load on the network *and* the server.
Brilliant!
</sarcasm>
The simple option might have been to mark the messages as read when we
decided they were junk.
That would work. Currently, when the user explicitly marks something as
Junk, it also gets marked as read (even if he hasn't read it). The only
slight downside is if the user decides something is not junk. If he
wants it to be Unseen he has to mark it as such (though a 90% solution
would be to have the Not Junk button do this automatically.)
That wasn't what was done, though -- instead of
just being able to ask the server "how many unseen messages are there in
this folder" we now have to fetch the flags for _every_ mail in the
folder and count the ones which are unseen but not 'junk'.
In fact we also download the _headers_ for every mail in every folder
too. That's just a side-effect of the above, I think; there doesn't seem
to be even a tenuous reason for that.
By "active" do you mean "subscribed"?
That is the definition of 'subscribed' in the IMAP specification, yes.
OK.
poc
[
Date Prev][
Date Next] [
Thread Prev][
Thread Next]
[
Thread Index]
[
Date Index]
[
Author Index]