<<< Date Index >>>     <<< Thread Index >>>

Re: recent header-cache changes?



Hi,

after I rebuilt mutt from CVS, header caching didn't work anymore as it used
to. It turned out that I had to set the variable imap_cachedir for which I
could not find any documentation in the manual. On mutt-user, Kyle Wheeler
pointed me to some information in init.h, but I still have questions that
maybe Brendan can answer: 

 * How do imap_cachedir and header_cache interact? IIUC, caching of messages
   in an IMAP mailbox is a new feature that is supposed to speed up, e.g., the
   reading of messages. Why is opening an IMAP mailbox (with hcache enabled)
   significantly slower than before unless I set imap_cachedir?

 * Does mutt ever purge messages from the cache? (E.g., if I delete or move a
   message.) 

Regards

Christoph

----- Forwarded message from Kyle Wheeler <kyle-mutt@xxxxxxxxxxxxxx> -----

Date: Wed, 8 Feb 2006 22:30:07 -0500
From: Kyle Wheeler <kyle-mutt@xxxxxxxxxxxxxx>
Subject: Re: recent header-cache changes?
To: mutt-users@xxxxxxxx

On Wednesday, February  8 at 10:51 AM, quoth Christoph Ludwig:
>>The imap_cachedir setting *is* documented in the manual:
>>
>>   imap_cachedir
>>       Type: path
>>       Default: ""
>>
>>       Set this to a directory and mutt will cache copies of messages
>>       from your IMAP servers here. You are free to remove entries at
>>       any time if space becomes an issue.
>>
>>I suspect you have a somewhat more complex problem. Try compiling from a 
>>fresh CVS pull.
>
>Thanks for the information, but I still cannot find it in the manual:

Sorry - I always forget that mutt’s manual and mutt’s man page are to 
very different things. That text above was from the man page, which is 
built from the contents of init.h

>And I am still wondering how 'imap_cachedir' and 'header_cache' 
>interact - why did opening an imap mailbox repeatedly take much longer 
>compared to a mutt build from last fall until I set imap_cachedir? Or 
>was this a side effect of Brendan's changes to the imap code that is 
>now masked by the additional caching?

That I don’t know - might have to ask Brendan himself. I’m a fan of as 
much caching as I can get (though to be honest, I’m uncertain why I use 
both of them… imap_cachedir would seem to be enough to make header_cache 
irrelevant).

~Kyle

----- End forwarded message -----

-- 
FH Worms - University of Applied Sciences
Fachbereich Informatik / Telekommunikation
Erenburgerstr. 19, 67549 Worms, Germany