[12:23:44] <phedenskog>	 greg-g dpifke: great news :)
[18:43:05] <AaronSchulz>	 Krinkle: https://gerrit.wikimedia.org/r/c/mediawiki/core/+/681170
[20:15:36] <AaronSchulz>	 Krinkle: https://gerrit.wikimedia.org/r/c/mediawiki/core/+/666780/9 can be reviewed
[20:16:59] <Krinkle>	 AaronSchulz: what is the 0000000000000000000000 default for? (as opposed to NULL)
[20:17:26] <Krinkle>	 I'll land it together with the commit that uses it but I can re-review it, it's a fairly straight forward schema change indeed. 
[20:17:34] <AaronSchulz>	 Krinkle: it's easier to do the inequality updates
[20:19:04] <Krinkle>	 AaronSchulz: Hm.. you mean like `modtoken < $val` ?
[20:19:18] <AaronSchulz>	 yes
[20:19:37] <Krinkle>	 is NULL not considered < any thing?
[20:19:50] <Krinkle>	 will mysql use default for existing rows?
[20:20:01] <Krinkle>	 like a magic inheritance sort of thing?
[20:21:48] <Krinkle>	 ah, right, it always returns false in comparisons
[20:25:02] <AaronSchulz>	 yep
[20:25:18] <Krinkle>	 and so for existing rows, truncate, migrate, or magic?
[20:25:49] <AaronSchulz>	 only x2 needs the change, the rest can be left alone (via feature flag)
[20:26:30] <Krinkle>	 yeah, but i don't want to leave a landmine for other users. what should people do when considering this option.
[20:26:58] <AaronSchulz>	 it's not used by default
[20:31:28] <AaronSchulz>	 https://dev.mysql.com/doc/refman/8.0/en/innodb-online-ddl-operations.html huh, might be easy to add to existing tables though
[20:33:28] <Krinkle>	 yeah, it's not on by default, but we presumably do want it to be useful to at least some third parties (most likely Fandom at the very least).
[20:33:50] <Krinkle>	 If they have mainstash db ahead of time, it'd be great if it was obvious what to do.
[20:34:16] <Krinkle>	 If applying the schema patch before enabling the option suffices (the default assumpotion for developers, I think), then all is good.
[20:34:20] <Krinkle>	 but I wasn't sure that's the case.
[20:44:13] <AaronSchulz>	 Krinkle: that's how it works
[20:51:46] <Krinkle>	 AaronSchulz: so mysql accepts a schema change to add a non-null column with a default value to an non-empty existing table, and that default will effectively propagate as far as select queries are concerned?
[20:52:39] <AaronSchulz>	 yes
[20:55:34] <Krinkle>	 ok :)
[21:26:04] * AaronSchulz amends a few bits
[21:35:21] * AaronSchulz forgot some exptime logic