On 01/10/19 12:10, Patrick Georgi wrote:
Hi Kevin, Laszlo,
- Januar 2019 19:38, "Kevin O'Connor" kevin@koconnor.net schrieb:
Patrick, can you confirm that we need to update the above links?
Yes, the links changed because the entire architecture changed.
I kept pipermail as a read-only copy so old links into the archive keep working (as far as possible: pipermail links are notoriously unreliable).
Also, a few people asked if we could support the format of the older mail archives - is that possible?
It seems that mailman3 allows multiple message archive systems, with hyperkitty being the only one supported right now.
A pipermail-style backend that creates a plain html archive would certainly be a welcome addition, as Laszlo definitely isn't the only "dinosaur" (I'd propose some changes though: for example basing the URL on the hashed message-id, like hyperkitty does, instead of pipermail's plain counter, improves URL stability), but I'm not sure if my python-fu is up to the levels to quickly build something like that.
Once it exists, I'll happily integrate it, though.
Message-Id-based archive URLs are the best. GMANE used to support that, but GMANE is dead. mail-archive.com and public-inbox.org also support Message-Id-based archive URLs, and I think we could simply subscribe at least mail-archive.com's agent to the seabios list:
https://www.mail-archive.com/faq.html#newlist
Of course, it would be best if the list's own hosting provided the MID-based search / URLs too.
... Wait, am I dumb? Yeah, I'm dumb. Because, mail-archive.com is already subscribed to the seabios list:
https://www.mail-archive.com/seabios@seabios.org/ https://www.mail-archive.com/seabios@seabios.org/info.html
So, I'd suggest incorporating these links (as secondary archive URLs) into the wiki article as well.
And, for example, here's a Message-ID based URL pointing to the initial message of the current thread:
http://mid.mail-archive.com/20190108225218.GA13236@morn.lan
Thanks! Laszlo