[Mailman-Developers] Update 3.0 Important Implementation Change

[Pranjal Yadav]

Hello
I'm writing this email to notify one major change in the implementation
which I learned last weekend, as Steve mentioned earlier, it would be
pretty hard for users to type in bulky Subjects/Tos in the message which is
a primary need for me to assess messages as dlist messages or not with a
follow up thread info to be extracted. To get rid of that I used the thread
id (which is generated for a particular thread depending upon the newness
of the thread) and wrapped it around with list name and web host and added
that to Reply-To so that a user can easily choose that option and the bulky
content is autogenerated. However while coming up with this solution I
missed an important detail i.e. how the lmtp runner would respond to such a
'To'.  So I modified the split_recipient method in lmtp after discussing
few things with barry and currently I can accept
listaname+HASH-request at domain,  where HASH is my thread id which is
typically a sha1 hexdigest and the request part is as before. Later I will
be adding tests to check the HASH part with relevant test cases. So sending
a pull request will take a little longer than expected since I need to re
do my rules with some other changes. Although its a little bump and will
not hinder my work for long, I think its a crucial one too and needs your
attention before going ahead. Please feel free to reply when ever you find
time.

I understand that its not easy to workout or check things which are working
and which aren't since there is no documentation yet, I request you to be
patient and spend a little time with my shell logs from previous blog
posts, I have made it similar to the mailman documentation and anybody who
had earlier gone through the documentation can easily catch up.

I also take this opportunity to thank Stephen, Terri, Florian, Barry &
Abhilash for being a constant support as I cleared the mid term evaluation
recently. I am looking forward for some code reviews soon!