[prev in list] [next in list] [prev in thread] [next in thread] 

List:       c-client
Subject:    Re: Announcing: UW IMAP toolkit 2000 Release Candidate 7
From:       Mark Crispin <MRC () cac ! washington ! edu>
Date:       2000-10-18 15:35:58
[Download RAW message or body]

On Wed, 18 Oct 2000 15:50:10 +0100 (BST), David Lee wrote:
> 1. Are there any further events planned for the next week or so?  (Such
> as RC8 or the official release?)

There will probably be an RC8, because it looks likely that there will be
another Pine prerelease before the release of Pine 4.30.

> 2. Should we move imap-utils at the same time?

Sounds like a good idea.

> we still have imapd, popd and tmail which (I'm fairly sure) are also
> calling initgroups().  My guess is that this is unnecessary, and I'm
> thinking of producing a patch so that "initgroups()" can be bypassed
> (using some sort of "#define" mechanism at build time).

If you depend upon group access working (e.g. for shared mailboxes), then you
need the initgroups().  Otherwise, you can turn it off as you suggest, but
please remember that you broke group access (so don't send me a bug report
about it later!).

Sooner or later, there will be mechanisms to make group access more useful
with IMAP.  You'll have to decide whether to do without the facility, or solve
the problem in some other way.

I'm a little bit surprised to hear that anyone is able to use NIS+ and get
reasonable server performance.

> Another point: why is imap-utils separate?  Given that it is so intimately
> bound to the main imap software, why not just bundle it?

I've thought about it, particularly for tmail and dmail.  The problem is that
a great deal of effort had to go into aggressive portability in the main IMAP
toolkit.  Somewhat less portability effort went into the imap-utils, and that
needs to be fixed before they could be bundled.

They're UNIX-only; I would not feel happy with bundling unless (at least) I
made them work on Win2K.  It'd be easier to fix some, harder to fix others.

Then there's all the old UNIX ports.  There are still systems out there that
don't have ANSI C compilers.  I didn't write some of the imap-utils, and those
programs don't follow the rules necessary to feed through the unansi program
properly.  Fortunately, this problem is rapidly heading into extinction.

It probably will happen, eventually; but it can't happen yet.

> Thanks for your continued provision of the great software.

Thank you for the nice words.

[prev in list] [next in list] [prev in thread] [next in thread]