martin-evans.me.uk Always finish the day on a success

I currently maintain DBD:ODBC and needed to obtain a lob (large object) in chunks. It doesn't really matter why but in general the lob is very large and I can process it in chunks in my Perl code so it makes sense to read it in chunks instead of read it all in one go (thus requiring loads of memory). DBD::ODBC does not have this functionality currently.

In the last few days on and off I've been trying to locate a problem with a fairly complicated Perl daemon using POE which was taking a huge amount of time to return a large file via LWP::UserAgent. It's taken a while since in reality we were not doing a simple HTTP GET and the circumstances under which the problem occurs are difficult to set up. However, I've at last tracked it down to taint mode.

Not for the first time a Perl POE daemon in our development system seemed to be taking a long time on certain requests. It is a large amount of code and although we had our suspicions we've learnt the lesson now - Devel::NYTProf is so simple to use we had no hesitation in running it on this daemon. Instead of: you simply run: then a nytprof.out file is created and then you can generate the HTML report and view it with:

A great deal of issues arise in the dbi-users mailing list and to me personally concerning ODBC connection issues so I thought I'd describe the ODBC connection process here (with a slight Perl angle since I look after DBD::ODBC). The information here should be of general interest to anyone using ODBC.

I have uploaded the official 1.24 release of DBD::ODBC to CPAN. Many thanks to everyone who has helped with this whether it was patches or testing. Here are the changes since 1.23: =head2 Changes in DBD::ODBC 1.24 May 14, 2010 Minor change in Makefile.PL to only use NO_META if ExtUtils::MakeMaker is at least at version 6.10. Reported by Chunmei Wu. Minor change to test rt_50852 which had wrong skip count. =head2 Changes in DBD::ODBC 1.23_5 May 6, 2010 Added advice from Jan Dubois (ActiveState) on building DBD::ODBC for ActivePerl (see README.windows).

I the last few days I released another development release of DBD::ODBC 1.23_5. I've been waiting for smoke testers to attack this release and then I will do an official 1.24 release. Since very few smokers actually run the test code I'm going to look at Test::Database for the next release - if you are a smoker and have Test::Database setup for DBD::ODBC I'd love to hear from you (really). 1.23_5 contains the following changes since 1.23: =head1 CHANGES =head2 Changes in DBD::ODBC 1.23_5 May 6, 2010 Added advice from Jan Dubois (ActiveState) on building DBD::ODBC for

I've maintained DBD::ODBC for some years now but one thing that constantly causes me problems is finding enough testers to be sure a development release is good enough to be released officially. CPAN Testers do a great job (and I joined the clan for a while until the throttling issues last Xmas and now I wait for CPAN Testers 2) but very few (if any) have a data source set up to a real database. Mostly the reports I get are PASS but when I look at them there was no actually data source set up so no real database tests actually ran.

I have just uploaded a new development release of DBD::ODBC 1.23_4 to CPAN. This is very much an experimental release but unless I hear any negative reports I will release this as 1.24. If you use DBD::ODBC you SHOULD test this release as it contains some significant changes to speed up prepared statements and as DBD::ODBC has to support a large number of ODBC drivers I have not been able to test the dozens of possibilities myself. If you use the MS SQL Server ODBC Driver or Native Client this change will definitely impact you and for other drivers this should reduce calls to SQLBindParameter.

In the last few days whilst hanging out on the #dbi IRC channel I saw messages from someone using/developing OTRS (Open source Ticket Request System) with DBD::ODBC.

I know that not everyone is in to IRC and I for one have only flirted with it over the years but for the last 6 months or so I've frequented the #dbi channel and there are a number of knowledgeable and helpful people on it but there is a low representation from DBD authors/maintainers.