I should say I can see results of new engineering/refactoring/bug hunt efforts inside Sun/MySQL.
Over last couple of weeks I started getting a lot of messages from the bugs system about bugs I reported long ago which were deferred to be fixed later or were left in open state. Here is example of such a bug.
I really hope this effort will result in a lot of these old annoyances fixed, which really matches my vision for MySQL – we do not need more big features we need old ones to work well and be convenient.
fix the url, remove &edit=2 from it 🙂
Sergei,
Fixed link.
Thanks
At my company we have a default, we push only stable releases to PROD…
So, as you said, let’s hope that the bug handling will radically change, because if you see how this nasty bug (http://bugs.mysql.com/bug.php?id=14194) has been fixed (“This bug cannot be fixed in 5.0, as 5.0 is now stable. Will be fixed in 5.1.”), it is not acceptable. This bug is a catastrophy for many latin languages…
This bug has been corrected in 2006 but was still not commited in a stable release at the end of 2008… What a shame…
Well your request ‘we do not need more big features’ will most likely be met, now that Oracle are in charge.
Hehe,
The “big features” is probably a wrong word here. I just think recent years a lot of work was done trying to move up market for enterprise database adding views, triggers, stored procedures etc which is what MySQL “core” market did not need so much.
At the same point stability, performance, usability, operations tasks were not taken care of and these new features often were not well designed and integrated with other things.
For example there is Replication since MySQL 3.23 but there is still not out of the box ways to check master is in sync with slave or resync them (maatkit or Google patches have solution for these). Debugging and profiling stored procedures is another example.