On Dec 30, 2003, at 8:21 AM, Geoffrey Young wrote:
but the connection cache mechanism has always seemed rather funky -
when we
were writing about Apache::DBI for the cookbook, I remember some issues
between Win32 and linux that randy and I were having when it came to
emulating the connect-string
Paul Makepeace wrote:
> Hi,
>
> Apache::DBI doesn't seem to be caching connections here causing a test
> failure. I'm using Debian/unstable, Debian's perl 5.8.2, mysql 4.0.16,
> DBD::mysql 2.9003, and CPAN's DBI 1.39. (DBI warned about perl having
> threading enabled, FWIW.)
>
> Any ideas what
Geoffrey Young said:
> this is what leads me
> to believe that Apache::DBI (at least as it existed then) relied on
> non-random hash behavior.
It sorts the keys, so it should have no problems with the new randomized
order.
DBI->connect_cached() is fine, except that it doesn't do the automatic
roll
> Well, what else can you recommend instead of Apache::DBI ?
> Suppose DBI->connect_cached ?
oh, sorry - I didn't mean it like that :) what I meant was that I'm not
really running production sites anymore, so I'm more and more out of touch
with Apache::DBI and the specifics of its implementation/
Hello Geoffrey,
Geoffrey Young wrote:
I've been wondering if the random-hash ordering foo in recent perls would
cause Apache::DBI to break - IIRC it was somewhat dependent on the order of
the connect() arguments. I've been meaning to dig into it just out of
curiosity (I don't really use Apache::
Paul Makepeace wrote:
> Hi,
>
> Apache::DBI doesn't seem to be caching connections here causing a test
failure. I'm using Debian/unstable, Debian's perl 5.8.2, mysql 4.0.16,
DBD::mysql 2.9003, and CPAN's DBI 1.39. (DBI warned about perl having
threading enabled, FWIW.)
>
> Any ideas what could be
