Importing a maillist archive?

Have you considered trying to import one of our maillist archives to see if that actually builds a usable knowledgebase?

It seems like the Bugs or Announce MailLists are obvious test candidates… since they are less active and smaller. Announce might be best since it is also a Good Candidate for limiting Topic creation to admin only users.

I briefly investigated this. We can download the mailing list archives, but the sender details are obfuscated. In fact, SourceForge no longer gives us access to the user management section of the MailMan administrator interface, so we can no longer see a list of subscriber email addresses.

However, the main reason is that we don’t really have time. If the trial is successful then we may revisit this idea.

Much of the mailing list archives wouldn’t be of much interest anyway. Perhaps we could just manually copy interesting articles into a FAQ category?

In some ways, the Manually copying interesting articles into a FAQ is an unachievable goal.

Discourse allows Threads/Topics to be deleted/moved as they are processed into a FAQ or WIKI. Without some way to mark a Thread as transcribed (or obsoleted), the archive remains an overwhelming monolithic objective. It may take whittling forever to build from an import but at least it is possible.

Yes, and if the trial is a success we may well choose to import the archives.

1 Like

Some of the recent history from gramps-devel could be interesting - but it all come down to how much effort is involved.

The search of the current mailling list on sourceforge is next to useless. I just tried searching for discussion on python update in 5.2 and had to resort to manually serching each months archive :frowning:


What tool is used for the original Gramps Maillists & its archives?

A thread popped up in the monthly Discourse summary for a group migrating 21 million vBulletin postings to Discourse about preserving links to other postings within their archived site.

It might be that an appropriate tool will appear in the future if SourceForge used a proprietary maillist database.

Also, do we have the maillists archived? If so, could the curator add a provenance note to the discussion or “See Also” of the wiki Contact page so those archives will be harder to lose. We lost a chunk out of the MantisBT database during a migration… so hopefully someone came up with a plan to avoid a similar recurrence?