I've upgraded to v3.0.0 and very happy so far.

Thought I'd highlight the following in the change log as it caused me a lot
of pain. It's particularly nasty if you don't realise it's happened:

Changed the return type of Connection:ing() from int to bool. If you were
calling ping() in bool context or using its return value in bool context,
you will need to reverse the sense of the test because the previous return
code used zero to mean success. Now it returns true to indicate success.


One comment: it'd be nice if ConnectionPool had an upper limit on number of
connections. I have a bespoke class (was looking forward to ditching it)
that limits number of connections... if you reach it, you have to wait for a
connection to come free, this prevents a crazy amount of simultaneous
connections ever being made.


--
MySQL++ Mailing List
For list archives: http://lists.mysql.com/plusplus
To unsubscribe: http://lists.mysql.com/plusplus?unsu...ie.nctu.edu.tw

Corin Fletcher wrote:
Regrettable, but returning a falsy value to indicate success sucked.

I think to make this work, we need to make release() virtual so you can
override it in your ConnectionPool derivative. This will break the ABI,
but I'm willing to bend the rules for a short period.

I have to ask, though, why didn't you test this during the extensive
beta period? You had 10 weeks...

--
MySQL++ Mailing List
For list archives: http://lists.mysql.com/plusplus
To unsubscribe: http://lists.mysql.com/plusplus?unsu...ie.nctu.edu.tw