On 27/06/2013 22:22, Tim Bunce wrote:
On Thu, Jun 27, 2013 at 01:17:32PM -0700, Robert Dodier wrote:
Hi, I think I have run into this bug;
https://rt.cpan.org/Public/Bug/Display.html?id=71555
I am working on a 64-bit Windows Server 2008 system
and 32-bit Strawberry Perl (32-bit since it appears that
there is a module I need which doesn't work with 64-bit).
perl/vendor/lib/DBD/mysql.pm says $VERSION = '4.022' in it.
Does anyone know of a patch or workaround for bug #71555?
I was amazed and dissapointed that such a serious apparent bug
has gone without even a comment for over 18 months.
Then I looked at the RT queue for DBD::mysql
https://rt.cpan.org/Public/Dist/Display.html?Name=DBD-mysql
and felt sad at the apparent neglect. 127 bugs (7 critical)
of which over half are still 'New'.
and of those I see:
o a load of supplied patches and many have been applied by MICHIELB to
github
o quite a number of duplicates
o a few that look like they may be misunderstandings or errors on the
part of the reporter
o 4 items marked as wishlist/enhancements - there are obviously more not
marked as such
o some which seem to be fixed but not marked as such e.g. rt60085
o 70+ rts over 2 year old
Of the criticals some look like they might be fixed but the reporter has
gone away e.g., the 6 year old 19091.
The 2 year old critical 68374 looks to contain a fix by the reporter.
68169 is marked
CHORNY [...] cpan.org - Fixed in 4.020 added. 72501 is --testuser switch
to Makefile.PL does not work on windows (not what I'd classify as
critical). 76835 contains a pull request with fix. So by my reckoning
that only leaves 2 critical issues that need investigating and the
others are fixed or not critical.
It looks to me like a little house keeping and a new release would
seriously dent the queue.
As for the really old reports that situation existed with DBD::Orace a
few years ago. John, Yanick and I had a concerted effort to clear up the
old ones but in some cases they could not be reproduced and the reporter
a) was not contactable b) wasn't using DBD::Oracle any more. In these
cases I didn't see any point in keeping them open and a small few were
written off. However, it was amazing what a concerted effort managed in
a few weeks.
On the other hand, I see the code is on github (along with 13 other bug
reports) at https://github.com/CaptTofu/DBD-mysql and Patrick seems
active in recent months.
I'd like to know why there hasn't been more activity on this and the
other critical bugs.
Do the DBD::mysql maintainers need more help?
If you're a DBD::mysql user and care about the future of the code,
please help out.
Tim.
I've no idea if the DBD::mysql maintainers need more help. When I was
actively using DBD::mysql (a lot of years ago) I supplied patches to
Patrick and they were applied. Not everyone is capable of supplying
patches. I'm guessing DBD::mysql is probably the most used DBD and as
mysql is so easily installed by most package managers I guessed a lot
more "inexperienced" users are using DBD::mysql and that might be
leading to a load of wrong rts but the last one rejected was 3 years ago.
Three years ago (I think it was just after a LPW), I suggested having a
DBD team where DBD maintainers (and anyone else who was capable) could
pitch in to help keep the rt queues down - there wasn't much interest. I
think I repeated it again about 1 year ago with a similar result
(although that may have only been on irc). Now most DBDs are in git it
should be a lot easier. I also see around 25 people on #dbi regularly
these days whereas 2 years ago it was probably around 5, if that. I get
a feeling there is a lot of potential there if it can be stirred up a
bit. I wish I could have attended the last QA hackathon as I would have
been happy to work on rt queues for any of the DBDs I thought I could
have helped with. I wonder if we could organise a DBI/DBD hackathon even
if it was one done remotely.
Anyway, here is my offer to DBD maintainers and in particular DBD::mysql
maintainers. If you want some help with rt queues in particular whether
that be simply some house keeping on the queue itself (I'd need
privilege to do that bit - cpan id MJEVANS) or bug finding/fixing I'm
prepared to help out where I can - just ask. For reasons I'm not going
to go into here I will have more spare time over the next month and the
DBD::ODBC queue is almost non existent and DBD::Oracle queue is almost
under control so get it while you can.
Martin
--
Martin J. Evans
Wetherby, UK
