Fwiw Dialect discovery is far and away the preferred solution for
"specifying" Dialect. That said, I am ok with splitting off Dialect(s)
specific to MariaDB.
Relatedly, I wonder if we should begin to remove the Dialects we have
marked as deprecated and whether we should consider a "legacy" repo where
we can dump these (as well as other removed features - Javassist support
e.g.). The expectation would be no continued development, but would make
them more readily available to users who still needed them as opposed to
simply dropping them completely. We could even alter the license of those
classes to be less restrictive than LGPL in terms of contributing back
On Mon, Feb 6, 2017, 6:47 AM Sanne Grinovero <sanne(a)hibernate.org> wrote:
> +1 to split them. I don't expect people to consider them "the same"
> We probably should have split them even if there had been no technical
> differences, just to make it clear which one needs to be configured.
> On 6 February 2017 at 12:37, Vlad Mihalcea <mihalcea.vlad(a)gmail.com>
>> I've had a very interesting discussion on Twitter regarding MarisDB:
>> And, since MySQL and MariaDB have gone in different directions, we might
>> want to provide MariaDB Dialects as well.
>> For instance, it's not very intuitive for a Hibernate user to figure out
>> that they need to use the MySQLInnoDb57Dialect to handle Timestamps with
>> microsecond precision which have been available since MariaDB 5.3:
>> Let me know what you think.
>> hibernate-dev mailing list
> hibernate-dev mailing list
hibernate-dev mailing list