latin1_swedish_ci collation ?

[Eminos]

Hello,

I noticed in PMA that "latin1_swedish_ci" collation is set for all new databases as default (including the ispCP database). Why is that? And what is collation anyway? Should I change this default to something more convenient and neutral like UTF-8, and how do I do that?

/E

[joximu]

latin1_swedish_ci is the default from MySQL - they are from sweden :-)

Most important is the first part latin1 - it's the charset. the second is the sort order, which is in case of swedish not really great (ü is sorted near y which is not the case in the german language...).
utf-8 would be better. We already did a step in this direction with Changeset r993 but there are some more steps todo...

/J

[Eminos]

Alright ! Fucken swedes, I know they had something to do with this (I'm a swede, hehe)

So this is default, and not that importent then? Or is it good and easy to change this defult to UTF-8 for at least all new databases?

/E

[joximu]

If you use characters which are not in ascii127 then you may better convert all to utf8... not needed for the lazy english speaking admins...

/J

[Eminos]

Alright, but my question about how I make the change so that all new databases are UTF-8 per default is unanswered.
Do I have to edit some config file for mysql, or does the Changeset r993 make this change?

/E

[joximu]

Changeset r993 does it.
at least php uses utf8 then - the engine part (perl) maybe doesn't...

[Eminos]

Alright, I changed the stuff manually for changeset r993, but all new databases are still "latin1_swedish_ci". Maybe I shouldn't care so much becouse this is default.

/E

[joximu]

Seems so - we're still working on a full UTF8 adaption...