Collation utf8_bin is not portable and breaks @Column annotation about shortid-doctrine HOT 8 CLOSED

I guess that we can remove collation, or at least make it optional.

from shortid-doctrine.

Do you think related PR can solve?

from shortid-doctrine.

It solves for mine environment, to be 100% sure, one should verify also with MySQL.
On postgres it generates valid code and respects collation setting.
👍

from shortid-doctrine.

Fixed and released in v0.6.0

from shortid-doctrine.

@garak please reopen. why the forced collation at all? isn't this a simple string value to store? please make it configurable or extend your type from string type. is there a special reason for using utf8_bin? this breaks the usage for shortids on sqlite.

from shortid-doctrine.

Sqlite should be unaffected, since last change restricted utf8_bin to mysql.
I can't recall the reason for such choice, I guess it was performance-related.

from shortid-doctrine.

oh, after again having a look at the code, i see you are right (i have overlooked the !). my problem was that shortid-bundle installed 0.5.0 and is not "compatible" with 0.6.0 but i simply dropped the bundle and the doctrine type and made my own doctrine type.

but still i think the code is risky and can have unexpected behaviors. as you still hardcode the value for mysql which overrides the annotation. even if it has performance reasons, shouldn't it be configurable and be up to the user if enabled or not? default can of course be on.

from shortid-doctrine.

I agree that a configurable option is better than an hard coded one. Feel free to open a PR

from shortid-doctrine.