TRAIN_00555.eml

NoneNone> From:  Scott Lipcon <slipcon@mercea.net>
> Date:  Mon, 23 Sep 2002 22:35:05 -0400
>
> The speed is a problem for sure - It takes a long time to do the
> rescanning of sequences (I associate it with hitting the 'flist'
> button, or when my background flist goes off).    I'm running on a
> pretty fast system (Athlon 1700+, 512Mb RAM, 10k RPM ultra SCSI disk)
> and hitting flist used to take no more than a second.  The big
> difference might just be perception, because the the old code just
> updated all the folders (count + color) all at once, instead of making
> it look like there is 0 unseen, then counting its way back up.  
> 
> I doubt I'll have much time in the immediate future to hack at this,
> but if I do - can you suggest areas that might be the best to
> optimize?   If not, do you think we can put in some preferences to
> disable some of the more intensive features?   I'd rather disable all
> the sequence support (except unseen, of course) and have reasonable
> speed.   I suspect people on slow machines would find the current
> state unusable.

If I knew where the problem was, I'd fix it myself.

Finding it is probably more work than the actual fix.

> > 
> > This is because of your Hook_MsgShow_update_unseen which is calling a fun
> ctio
> > n 
> > which no longer exists.
> > 
> > I suspect you need
> > 	Seq_Del $exmh(folder) unseen $msgid
> > now instead of 
> > 	Mh_MarkSeen $exmh(folder) $msgid
> > 
> 
> Thanks.  I'm not sure I'll need it with the new sequence code, but I
> might.  Does your new code commit sequences immediately?  The old code
> didn't, so I put that in to help keep my mh and exmh windows in sync.

Yes it does.

Chris

-- 
Chris Garrigues                 http://www.DeepEddy.Com/~cwg/
virCIO                          http://www.virCIO.Com
716 Congress, Suite 200
Austin, TX  78701		+1 512 374 0500

  World War III:  The Wrong-Doers Vs. the Evil-Doers.