[20:10:16] [discord] XML dump is currently at 11.7GiB after 1h19m. xz has it down to 53.1MiB, though, so there's hope I won't have to stream the backup over the network to my workstation. I suspect a 10GiB+ XML file will fail on upload to the import request page, though. [20:10:17] [discord] [20:10:18] [discord] https://meta.miraheze.org/wiki/Special:RequestWikiQueue/26377 [20:10:21] [url] Wiki requests queue - Miraheze Meta | meta.miraheze.org [21:44:16] [discord] XML dump completed. 17.3GiB uncompressed, which is less than I expected. Compressed to 104.0MiB via xz and a custom xz filter chain, though, which is good. [21:48:33] [discord] :DoneMH: **Approved**, thank you for choosing Miraheze! [21:49:42] [discord] If you want, you can send us the compressed xz file at sre@miraheze.org and you can file a [[Phabricator]] task so that we can track the import [21:49:42] https://meta.miraheze.org/wiki/Phabricator [21:49:45] [discord] [21:49:46] [url] Phabricator - Miraheze Meta | meta.miraheze.org [21:50:04] [discord] Will do. First, time for dinner... [21:50:21] [discord] Sounds good, we'll be expecting your email [21:51:52] [discord] Since this dump has PII in it (peoples' names, email addresses, passwords (hashed, I hope)), I'll need to encrypt it first. Preferably pubkey, rather than password. Have a GPG public key I can encrypt it to? [21:53:00] [discord] XML dumps don't have any user account info, only info on their pages, revisions, and so on [22:05:08] [discord] Uhf. So I lose edit attribution for the users? That could be a problem. [22:06:58] [discord] Oh, it looks like author is in the schema. Not sure how author/user mapping is done, though. [22:09:20] [discord] No, edits will appear as "imported>[Username]" and link to their user account. We can run a script to manually assign imported edits to users/remove that "imported>" prefix too after the import is done [22:10:10] [discord] Sounds fine, then. How are the entries under the User: namespace handled? [22:11:39] [discord] User: pages are treated like normal pages by MediaWiki so they won't be touched or automatically moved by MediaWiki, etc [22:13:24] [discord] Ok. I'll leave general guidance to my users to their old user page to their new, if for some reason they weren't able to get their old username. (I.e. I wasn't; "Short Circuit" was too close to the existing "Short c1rcuit" or some such. O.o) [22:16:21] [discord] In cases like those when AntiSpoof is tripped, you can ask a Steward to override it and create an account for the user in question [22:17:00] [discord] Good to know. Of course there will be situations where there are genuine collisions. [22:19:49] [discord] Sometimes, we may be able to usurp a username if the user who uses the old username is inactive [22:24:12] [discord] Should I file a separate Phabricator task for enabling Semantic MediaWiki, or is it OK to bundle the two into one entry? [22:25:54] [discord] Eh. I'll file that separate after the import is done. Cleaner. [22:30:12] [discord] SMW is enabled by Stewards so I'll do that as soon as I see a MediaWiki Engineer around so that they can run a special script that's needed when you enable SMW [22:30:47] [discord] Ack. I read that a Steward was needed, wasn't sure if I should call that out in this phab task or a separate one. [22:31:52] [discord] The venue for contacting us is the [[Stewards' noticeboard]] in case you ever need something in the future but you'll usually find at least one of us on Discord [22:31:52] https://meta.miraheze.org/wiki/Stewards%27_noticeboard [22:31:53] [discord] [22:31:55] [url] Stewards' noticeboard - Miraheze Meta | meta.miraheze.org