squeak gmane archive broken

Hi !

I'm using intensively the squeak gmane archive through my mail reader,
and noticed that all the threads are now broken, the result is
unfortunately not readable. Is it possible to fix it ? Or maybe the
problem was already mentionned ? I guess that it is due to the recent
addition of older archived mails !

Cheers,
Samir

Samir Saidani <[hidden email]> writes:

> I'm using intensively the squeak gmane archive through my mail reader,
> and noticed that all the threads are now broken, the result is
> unfortunately not readable. Is it possible to fix it ? Or maybe the
> problem was already mentionned ? I guess that it is due to the recent
> addition of older archived mails !

Perhaps the archive had broken Message-IDs or something?

--
(domestic pets only, the antidote for overdose, milk.)
  [hidden email] * Lars Magne Ingebrigtsen

Hello Lars,

I'm not sure because the threads was ok before inclusion of the old
mails. During this operation, I've noticed that the number of mails
into the nntp+news.gmane.org:gmane.comp.lang.smalltalk.squeak.general
was growing, and each time I browsed the archive through gnus, threads
was broken. So I thought that this problem would only occur during the
extension of the archive, but nothing has changed.

Threads are ok when you browse it from the official squeak-dev mailing
list (here http://lists.squeakfoundation.org/pipermail/squeak-dev/). I
noticed the same phenomena with the squeakland archive, threads are
flattened too.

So I think that the recent operation - inclusion of the old mails -
breaks maybe the Message-ID of the list.

Check the page 2 of
http://news.gmane.org/gmane.comp.lang.smalltalk.squeak.general, there
is a lot of messages entitled "REMINDER : Squeak Chat ..." which
didn't appear before the inclusion, and don't appear in the squeak-dev
mailing-list. Compare with this thread :
http://lists.squeakfoundation.org/pipermail/squeak-dev/2005-September/thread.html

Ok, see what I've noticed right now:

The mail from "Alan Lovejoy Re: About squeak image compatibility" has
a message-ID : Message-ID:
<21459.4053723433$[hidden email]>, from gmane, and before,
all threads are broken. And the situation seems to be ok starting with
a mail from "Daniel Vaisencher" (the next one on the thread) with a
message-ID : Message-ID:
<43C8F02F.3010107__16977.8764983223$1137074823$gmane$[hidden email]>
(not a news.gmane.org one)

I tried to browse the 5000 latest messages of squeak-dev, and I
noticed that the broken threads problem starts with this message:

From: "Ned Konz" <[hidden email]>
Newsgroups: gmane.comp.lang.smalltalk.squeak.general
Subject: [ENH] DupKeys-nk
Date: Thu, 11 Jul 2002 07:48:58 -0800
Lines: 69
Approved: [hidden email]
Message-ID: <30139.7983603694$[hidden email]>

(before the Message-ID is not a news.gmane.org one and threads are ok)

and finishes with this message

From: "Alan Lovejoy" <[hidden email]>
Newsgroups: gmane.comp.lang.smalltalk.squeak.general
Subject: RE: About squeak image compatibility (3.6/7/8)
Date: Mon, 9 Jan 2006 01:47:19 -0800
Lines: 13
Approved: [hidden email]
Message-ID: <21459.4053723433$[hidden email]>

(after the Message-ID is not a news.gmane.org one at all, and threads
are ok)

It seems that this change is involved in the problem. Hope this
indication would help you to figure out what is the problem !

Regards,
Samir

Lars Magne Ingebrigtsen <[hidden email]> writes:

Samir Saidani <[hidden email]> writes:

> It seems that this change is involved in the problem. Hope this
> indication would help you to figure out what is the problem !

It seems likely that the messages already were in a different group
(or didn't have Message-IDs, or they were syntactically invalid), so
the mail-to-news script assigned new Message-IDs.  Which breaks
threading.

--
(domestic pets only, the antidote for overdose, milk.)
  [hidden email] * Lars Magne Ingebrigtsen

Ok, do you mean that this problem can not be fixed ? Or it could but
it would be hard to fix manually each broken message-ID ?

Lars Magne Ingebrigtsen <[hidden email]> writes:

Samir Saidani <[hidden email]> writes:

> Ok, do you mean that this problem can not be fixed ?

Nope.  It'll stay broken.

--
(domestic pets only, the antidote for overdose, milk.)
  [hidden email] * Lars Magne Ingebrigtsen

Gmane FAQ:
Why are all the articles out of order after doing an import?
They aren't on the news server. On the web server they'll remain out of
order until weaverd has rethreaded them, which'll happen six months later.
Is it the problem you are referring to?

Petr

Samir Saidani wrote:

Not exactly, (I think !), it seems that they are out of order on the
news server as well, and that these messages are intrinsically broken
(message-ID broken).

Petr <[hidden email]> writes:

I see nothing wrong with the messages as I provided them to Lars.  Feel
free to check it out yourself at

http://discuss.squeakfoundation.org/squeak-dev.mbox.gz

(Warning to anyone considering downloading this file - I don't recommend
it as it is approximately 109MB in size and doesn't contain anything you
can't already find at

http://lists.squeakfoundation.org/pipermail/squeak-dev/

and is in fact a few days behind at this point.)

This file is just a copy of the mbox file created and used by Mailman
and I can't see why it would have caused a problem.

Ken

On Tue, 2006-01-17 at 00:18 +0100, Samir Saidani wrote:

Ok, after checking the squeak-dev.mbox.gz file,

Message-ID for one problematic mail (From [hidden email] Mon Jan
09 07:55:26 2006) is:

Message-ID: <op.s23pmazvzb486t@wynorski>

which is maybe not conventional ?

Another one (From [hidden email] Mon Jan 09 06:00:00 2006)
concerning Squeak Chat has no Message-ID field at all (same for the
others mails about Squeak Chat), so I guess that this kind of message
are broken (maybe due to the automatic script which does not send a
standard mail ??).

Cheers,
Samir

Ken Causey <[hidden email]> writes:

On Fri, 20 Jan 2006 05:48:20 -0800, Samir Saidani <[hidden email]>  
wrote:

>
> Ok, after checking the squeak-dev.mbox.gz file,
>
> Message-ID for one problematic mail (From [hidden email] Mon Jan
> 09 07:55:26 2006) is:
>
> Message-ID: <op.s23pmazvzb486t@wynorski>
>
> which is maybe not conventional ?

What's wrong with it?