[rt-users] [RT] Message Threading

[Michael Witten]

On Mon, May 16, 2011 at 12:45, 20/20 Lab <lab at pacbell.net> wrote:
> On 05/16/2011 9:43 AM, Michael Witten wrote:
>>
>> Unfortunately, email systems have evolved in such a way that the standard
>> headers `References' and `In-Reply-To' are horribly [ab]used; the result
>> is that interfaces for navigating threads invariably mangle threading
>> (in my opinion, the main problem is caused by the fact that nobody cares
>> what `In-Reply-To' says and everybody makes naive assumptions about what
>> `References' supplies).
>>
>> In any case, I dare you to go here:
>>
>>   http://www.nntp.perl.org/group/perl.perl5.porters/2011/05/msg172431.html
>>
>> and play around with `Thread Previous' and `Thread Next', and then provide
>> an explanation for WTF is going on (note dates too).
>>
>> Now, admittedly, this isn't necessarily a problem with perlbug or
>> rt.perl.org's software, but I would add that the request tracker
>> exacerbates this situation by flinging copies of emails all over the
>> place with various Message-Ids and semi-spoofed `From' headers.
>>
>> Surely something can be done to make working via email more seamless;
>> the request tracker should just be a third-party recording the
>> discussion (and only spoofing messages sent explicitly via its
>> interface).
>>
>> Sincerely,
>> Michael Witten
>>
> Looks like whatever software is threading the messages is not compensating
> for time zones and/or is running on utc.
>
> If you wanted a guess.
>

Well, it's worse than that: You'll notice that following `Thread
Previous' lands you on the top message from which no other message in
the thread is accessible; the link in my original email to the
rt-users list was found by running through the URL message ID numbers
by hand because it is otherwise completely impossible to retrieve.

Also, the latest message from Thomas Sibley, which can be seen here:

  http://rt.perl.org/rt3//Public/Bug/Display.html?id=90632

does not appear to have even been sent out to the perl5-porters list
(at least yet).

However, as Jesse has pointed out, this is perhaps a bit of a
digression from the affairs of the rt-users list.