These are chat archives for eventum/eventum

11th
Nov 2016
Bryan Alsdorf
@balsdorf
Nov 11 2016 01:40
@tiagom62 I think the old script should run fine, you just want to run it before you start inserting UTF-8 data into your database as well
This is an old post, but it has some other ways you can convert or test: https://coderwall.com/p/gjyuwg/mysql-convert-encoding-to-utf8-without-garbled-data
Bryan Alsdorf
@balsdorf
Nov 11 2016 02:41
@glensc I couldn't recall if I should submit upgrade scripts named with the next number or something like XX_ to avoid possible duplicates / when merging. On #219 I went ahead and used the next number but let me know and I'll rename the file
tiagom62
@tiagom62
Nov 11 2016 15:22
We do not use non-english characters. Could we get away with exporting the db and modify the import script for utf8? My (limited) understanding is that non-english characters get garbled when you do something like this, but since we dont use them it shouldn't be a problem.
Elan Ruusamäe
@glensc
Nov 11 2016 15:52
well, there's inconsistency with your information, you did get json encode error, so apparently your data is not ASCII only. however you could just assume your data is latin1, which means you should try the "charset was proper" script
but the proper/improper was explained (somewhere), using latin1 charset, but data in windows-1257 is considered improper
but if your data is iso8859-1 and connection charset also iso8859-1 then your data is "proper" and mysql can change the encoding natively
so, setup test database, and just run the script and that's it
@balsdorf i think should just add travis ci check to fail if there are multiple patches with same number
tiagom62
@tiagom62
Nov 11 2016 19:15
the json encode error is because its not utf8 encoded. json_encode states all string data must be UTF-8 encoded.
i will get to testing it. thanks for the help