Re: [Evolution] Expunging not working on INBOX (IMAP).

["Zot O'Connor" <zot zotconsulting com>]

On Thu, 2002-03-07 at 16:00, Jeffrey Stedfast wrote:
> On Thu, 2002-03-07 at 18:40, Zot O'Connor wrote:
> [snip]
> > So now I wonder if the message range is "too big" for cyrus, I mean it
> > is 1760 chars big.  Or is it evolution not waiting long enough?
>
> There is no waiting period, it either does it or it doesn't. Your first
> guess is probably correct though.

I am now trying to remember if it worked for me or not.  I will retest.

> This is already fixed in CVS by splitting the command into multiple
> commands.
>
> > At a minimum it seems like evolution needs to look for the BAD response
> > and then either send a full expunge, or send partial lists.
>
> Already does in CVS.
>
> [snip]
>
> Basically, if the server says it supports UIDPLUS and doesn't support
> UID EXPUNGE, then it's broken. You CANNOT claim to support UIDPLUS if
> you don't support UID EXPUNGE.

As per usual the RFC is not explicit in this, but it seems the onus is
on the client (as it should be IMHO):

http://asg.web.cmu.edu/rfc/rfc2359.html#Sec-3

    Clients that wish to use the new command in UIDPLUS must of course
    first test for the presence of the extension by issuing a CAPABILITY
    command. Each of the features in UIDPLUS are optimizations; clients
    can provide the same functionality, albeit more slowly, by using
    commands in the base protocol. With each feature, this document
    recommends a fallback approach to take when the UIDPLUS extension is
    not supported by the server.

So the last line makes a "recommendation" which is odd.  It seems it
should make a "statement."  Dang wishy washy RFCs.




> Jeff
>
> -- 
> Jeffrey Stedfast
> Evolution Hacker - Ximian, Inc.
> fejj ximian com  - www.ximian.com
>
>
> _______________________________________________
> evolution maillist  -  evolution ximian com
> http://lists.ximian.com/mailman/listinfo/evolution
-- 
Zot O'Connor

http://www.ZotConsulting.com
http://www.WhiteKnightHackers.com