Question

I have tables with latin1_swedish_ci column. I’ve done tests and inserted into them some texts in languages like french, greek, japanese, spanish and it’s working fine.
Data written in japanese for example (ぴゃ) are well stored and display.
I don’t use UTF-8 and don’t want to use UTF-8.

Can I let latin1_swedish_ci or should I use another type/Collation (text/latin1_swedish_ci) for storing texts which can be in many languages ?

Answer 1

Using anything but utf8 is still okay. You just have to make sure that the connection you make to your database is made in the same collation.

Also, if you store unicode characters in non-unicode fields, you will face problems when you define indices on them, as the reference table for sorting and indexing the data will not reflect the correct character set.