Last modified: 2012-04-12 19:30:47 UTC
Right now, comments, commit summaries and everything fun is all stored with a latin1 charset in reviewdb. This is kind of annoying, especially when you're talking about fixing non-English bugs. Hopefully it's just a matter of adjusting the CHARSET on each of the tables, but we'd like to test first of course.
See comment 1 in https://gerrit.wikimedia.org/r/#change,3505 for a comment using unicode.
Setting the charset to utf-8 can cause some trouble with key length according to upstream author : http://groups.google.com/group/repo-discuss/msg/b9584ce01b4e4812 though he is speaking about utf-24 (or is it utf-32?). Anyway, we might only want to the utf-8 to be on some specific fields and tables. Recommendation is to use either the embed H2 database or postgreSQL.
*** Bug 35536 has been marked as a duplicate of this bug. ***
Chad thinks he can adjust the collation without an issue. He'll test his theory.
Duping to the other bug, it's got more information. *** This bug has been marked as a duplicate of bug 35626 ***