GNU bug report logs - #58472
[PATCH] Make `message-unique-id' less prone to collisions

Previous Next

Full log

Message #76 received at 58472 <at> debbugs.gnu.org (full text, mbox):

From: Eli Zaretskii <eliz <at> gnu.org>
To: Lars Ingebrigtsen <larsi <at> gnus.org>
Cc: matt <at> rfc20.org, eggert <at> cs.ucla.edu, stefankangas <at> gmail.com,
 58472 <at> debbugs.gnu.org
Subject: Re: bug#58472: [PATCH] Make `message-unique-id' less prone to
 collisions
Date: Mon, 17 Oct 2022 11:34:44 +0300

> From: Lars Ingebrigtsen <larsi <at> gnus.org>
> Cc: Paul Eggert <eggert <at> cs.ucla.edu>,  matt <at> rfc20.org,
>   stefankangas <at> gmail.com,  58472 <at> debbugs.gnu.org
> Date: Mon, 17 Oct 2022 10:29:07 +0200
> 
> I sort of agree with you here, but not totally -- I think a `make-nonce'
> function would be useful in general, because this is an area that's
> genuinely difficult to get right, and having a function that does this
> for you -- correctly -- is good.
> 
> But, like you, I'm not sure about the proposed changes otherwise.
> 
> And, like I've said before, there's a reason the Message-ID is on the
> format it's on now -- it has information that allows users to do work on
> it (so changing it will break some use cases), and it's short (which
> makes it efficient in many algos), and it's obviously "good enough" --
> it's been this way for decades without any problems.

Agreed.

> So I'd prefer not to change `message-make-id', but adding a `make-nonce'
> function would be nice anyway.

If we want a make-nonce function for unrelated reasons, by all means
let's discuss that -- but in a separate thread, and with the reasons
and use cases spelled out.  Doing it as a side effect of what was a
wishlist bug report for a minor feature to begin with doesn't sound
right to me.

Previous Next

GNU bug tracking system
Copyright (C) 1999 Darren O. Benham, 1997,2003 nCipher Corporation Ltd, 1994-97 Ian Jackson.