CPAN Testers is only made possible with the support of our sponsors.
For more information on sponsoring, please visit the I CPAN Testers website.

Upgrade Notice

The CPAN Testers Blog site has been upgraded since you last accessed the site. Please press the F5 key or CTRL-R to refresh your browser cache to use the latest javascript and CSS files.

News & Views

Posted by Barbie
on 29th December 2009

At the London Perl Workshop, I spoke with Leo Lapworth, who mentioned that he'd shown one of his colleagues the CPAN Testers Reports site. His colleague highlighted that the distribution pages were slightly upside down, as what he wanted to see first was the summary PASS matrix, as this gave him a better perspective of how successful the distribution was.

He makes a good point, and I began by simply bringing the summary matrix to the top of the page. However, the matrix just contained the counts of successes, which while encouraging, doesn't differentiate between versions and perhaps most importantly contained all the perl versions that had been tested, including all the patched versions. As such the matrix has now been reworked to only include non-patched versions of perl and to highlight the last version containing a pass. This hopefully gives a better first impression of a particular distribution.

For each distribution, there is a link to further matrices on the CPAN Testers PASS Matrix site, which now includes on the associated distribution page the full perl version matrices for both the latest version with a PASS report, and the count of PASS reports (as it was originally on the Reports distribution page). Examples of the new matrices can be seen here and here.

Further rework of the matrices on the site are planned at some point, but work on the other sites is planned for the new year, as well progress on CPAN Testers 2.0.

Posted by Barbie
on 18th December 2009

In recent times we have seen a dramatic increase in the number of testers, smokers and reports. So much so that we are seeing over 400,000 reports each month. This in turn is putting a strain on the Perl NOC, specially the email and NNTP parts of the system.

Initially David Golden has put out a request to CPAN Testers, and posted about it too, to throttle their testing for the next few months, and ease the pressure on the Perl NOC.

As CPAN Testers has now become such a strain on the Perl NOC, they have now given us a deadline of March 1st 2010 to move to CPAN Testers 2.0, which is the HTTP submission form of mails, rather than the SMTP form that it is currently. There have been several steps forward over the last 2 years, particularly by David Golden and Ricardo Signes with their work on the Metabase, and David has actually got the system to work in its basic form, but there is still work to be done.

Now that we have a deadline, the task of completing the relavent parts is now upon us. David is planning to put together a plan of the outstanding work to be done, and get a number of developers involved to help move it forwards quickly to have it all up and running by March 1st 2010. If you're interested in helping out, please join the cpan-testers-discuss mailing list, and let us know.

Posted by Barbie
on 10th December 2009

At the weekend, the Westminister University hosted the London Perl Workshop. The event was a great success, and many thanks to Mark Keating, his crew and all the sponsors for making it happen.

Chris Williams, our most prolific CPAN tester, presented a talk about CPAN Testing, which seemed to go down very well. Interestingly Tomas Doran made the observation that when anyone has problems with Catalyst distributions, the developers usually encourage people to submit a CPAN Testers report, as it includes all the information they need, rather than trying to explain to the person how to get it.

Chris' talk Rough Guide to CPAN Testing is now online.

In November we reached the 6 million reports submitted mark. It's quite staggering how many reports are being submitted these days. It's now roughly 1 million reports every 3 months! So expect a 10 million reports post some time in August 2010 :)

Now that we are producing so many reports, while there is a desire to get more reports from less tested operating systems, Tim Bunce recently highlighted his interest in getting reports that included a diverse set of Perl configuration flags, in particular regarding how Perl was compiled (with and without threads, etc). At the moment the CPAN Testers Statistics database doesn't include that information, but the Metabase that is behind CPAN Testers 2.0 will. In addition the Metabase will be able to be queried to glean the reports that contain a specific set of flags, etc. At the moment there are quite a few different setups testing on the top few operating systems being tested. While some authors see these as just repeated results, in some cases they provide slight differences in the test results. This is particularly what Tim was interested in for Devel-NYTProf. Hopefully we'll be closer to getting more of that information more readily available soon. In the meantime, if you do want to get involved with CPAN Testers, and only have a traditional operating system available, take a look at some of the reports posted by current testers for the same platform, and see what different setups you could provide.

In the CPAN Testers namespace, CPAN has seen a new upload, CPAN-Testers-Data-Addresses. This release will be the new way for me to manage the tester address mappings. To begin with the testing is being run stand-alone, but it will be shortly be integrated to the CPAN Testers Statistics website. From there it will also be integrated into the new site that is hopefully being launched early next year, which will allow testers to register their testing addresses (among other things). More uploads to the CPAN Testers namespace are being worked on, in particular ones to provide a more programmatic access to the CPAN Testers APIs. More news on those hopefully next month.

This weekend sees the annual London Perl Workshop. Featured in the schedule is Chris 'BinGOs' Williams' talk "Rough Guide to CPAN Testing". If you are a CPAN Tester and are planning to attend the event, please come along and say hello :)

Last month we had a total of 164 tester addresses submitting reports. The mappings this month included 17 total addresses mapped, of which 7 were for newly identified testers. A bit of a low mapping month, mostly due to my attention being elsewhere. With the new mapping system hopefully this will become a little more streamlined for next year.

Until next time, happy Christmas testing :)

<< January 2010 (5) November 2009 (5) >>