Re: [PATCH] improve savemessage_searchforheader - take 2

[John Goerzen <jgoerzen@xxxxxxxxxxxx>]

On Tue, May 16, 2006 at 10:03:17AM +0100, Adam Spiers wrote:
> John Goerzen (jgoerzen@xxxxxxxxxxxx) wrote:
> > On Mon, May 15, 2006 at 11:47:19PM +0100, Adam Spiers wrote:
> > > 
> > > Secondly, I found the below edge case with the Groupwise IMAP server
> > > which breaks it :-(
> > 
> > Unfortunately, that behavior from the IMAP server is not just bad or
> > wrong, but COMPLETELY in violation of standards.  This is really, really
> > broken on their part.
> 
> Really?  I'm not the RFC2060 expert so I'd be very interested to hear
> why.  Intuitively it would seem reasonable to me for the RFC to allow

Take a look at RFC2060, section 6.3.11, at
http://www.faqs.org/rfcs/rfc2060.html.

The append command gives the mailbox into which the message should be
placed.  There is simply no provision for server-side moving of mail
here.  I think it is a pretty serious bug to do it there, in fact.
The RFC stresses that APPEND is not a new mail delivery method.  Any
server-side filtering should occur during initial mail delivery, not
during APPENDs.

> Good news!  A colleague suggested a workaround to me which I was
> embarassed not to have thought of: simply add a condition to the rule
> "Item Status does not include 'Read'".  Therefore the rule only

Ahh, nice.  Makes sense.

> My pleasure.
> 
> <rant relevance="little"> 
> Why oh why is GNU Arch's usability so much worse than darcs?  I
> invested countless hours of my spare time learning it, setting up the
> right archives/toolsets etc., and failed to meet the objectives which
> I achieved with darcs in minutes - namely enabling easy sharing of
> patches with you :-(  I guess this is probably why you switched too?
> </rant>

Yes, it is.

That, plus I hated typing tla show-me-the-summary-of-changes-please and other
carpal-tunnel-inducing long commands ;-)

-- John