On 7/11/2014 14:03, Damien Kick wrote:
>> […] I had been assuming that in this case (Connection per thread)
>> there was/is no need for mysqlpp::Connection::thread_start(); i.e.,
>> the manual states that "you don’t need to call it when you use the
>> simpleConnection-per-thread strategy, because this function is
>> implicitly called the first time you create a Connection in a thread."
>> However, I had found <http://lists.mysql.com/bugs/15740> which
>> mentioned "My apologies, I haven't tested yet, but I am pretty sure it
>> is because I forgot to use mysql_thread_init()/mysql_thread_end()," so
>> I was not sure if I was understanding correctly.
> Yeah, and adding a call to mysqlpp::Connection::thread_start() even
> though it seems to contradict my current understanding of the tutorial
> doesn't help, either. So … I don't know <shrug/>
That assertion in the user manual is based on this part of the MySQL C
API reference manual:
So, it doesn't surprise me that adding those calls did not help.
By the way, the user manual sources are also in the MySQL++ source code
repository. Once you get this sorted out, if you can think of a clearer
presentation, patches to the docs are even more likely to be accepted
than patches to the code.
(We don't have to worry about ABI and API compatibility in the docs,
leaving correctness as the main reason a patch will be rejected.)
See doc/userman/README.txt for info on building the docs. It gives the
Fink option for getting a DocBook toolchain, but the tools are also
available via "brew install docbook-xsl", and presumably via MacPorts, too.