[Mailman-Users] Introduction, and another archive question

[Mark Sapiro]

Jan Cohen wrote:
>
>Here's how our current system works. Everything starts with
>the Mailman listserv. All member messages addressed to discuss at ixda.org go
>through Mailman. It mails them back out to the entire membership and at the
>same time creates an HTML archive of the message. That archive is here:
>
>
>http://tinyurl.com/3b9zm8
>
>
>( http://tinyurl.com/33ek3s is the link what's actually presented on the web)
>
>
>
>
>When Mailman creates the individual HTML page for a particular message, it
>assigns a sequential unique id and embeds it in the filename for the created
>page. About a year ago we started scraping those HTML archives and putting the
>contents of messages in a database and using them to drive the website at
>ixda.org. The unique ids we use for our posts are based on the unique ids Mailman
>generates.
>
>
>
>The problem is, the unique ids don't exist until after Mailman sends out the
>e-mail to the list, so it's not clear how to embed them in the e-mail
>beforehand.


Even Mailman doesn't know them when it sends the message to the list.


>On the web we have an advantage. We don't need to embed the actual
>post id, which is unknowable in advance; we only have to embed the id of the
>initial post in the thread, which we already know at the time of composition.
>We append that link onto the body of the message before we send the e-mail from
>the website to Mailman.
>
>
>
>Members who send e-mail directly to discuss@ get their e-mails redistributed
>through Mailman directly without an opportunity to add any link to the e-mail.
>Even if we had a way to know the post id to embed, it's not clear whether it's
>possible to create a dynamic footer of any kind in Mailman.


You can create 'dynamic' footers, but without modifying the Decorate.py
handler, you are limited as to what dynamic content can be in them.

In the current Mailman archiving system and architecture, it's just not
possible for the mail delivery process to know the id of the message
in the archive.

This issue has been discussed at length on the
Mailman-Developers at python.org list. See the threads with subject
"Improving the archives" at
<http://mail.python.org/pipermail/mailman-developers/> for July
through November, 2007.

There will be improvement in the Mailman 3.0 time frame.

-- 
Mark Sapiro <mark at msapiro.net>        The highway is for gamblers,
San Francisco Bay Area, California    better use your sense - B. Dylan