System administrator, part-time Perl hacker, full-time POE [perl.org] evangelist. One day he will be made to pay for his crimes.
He has some modules on CPAN [cpan.org]. They may or may not be useful
First off a developer release of POE::Component::SmokeBox should be hitting CPAN mirrors.
POE::Component::SmokeBox is a unified POE based CPAN smoke testing framework. It will eventually replace POE::Component::CPANPLUS::YACSmoke, POE::Component::CPAN::YACSmoke and POE::Component::CPAN::Reporter.
Production smoker-wise, work is ongoing to update configurations to use the latest CPANPLUS::YACSmoke and Test::Reporter. The long running smokers on Darwin and NetBSD 3.1 have already been updated, the Vmware based smokers are being updated as boxen come up in the rotation.
As mentioned during my talk at YAPC::EU in Copenhagen, I'm still working on code to perform auto-rolling of smoker configurations. I have worked through some proof-of-concept scripts that seem to do the job, but the end goal ( in my mind ) is to get this auto-rolling working with what will eventually become smokebox so that the whole process setting up the smoker and actually smoking CPAN modules can be fully automatic.
Some of the stages the auto-rolling will deal with are (assuming a CPANPLUS based smoker):
The trick now is to wrap up all this into a module.
Enjoying myself at YAPC::EU in the lovely city of Copenhagen
Gave my first proper YAPC talk about CPAN Testing, the slides for which are here
I've uploaded a few pictures I've taken to my Picasa album
Must sleep
YAPCs are always a good opportunity to meet other perl users. As the organisers of this year's YAPC::EU in Copenhagen have been good enough to allow me to do a talk on CPAN Testing, Barbie and myself will be arranging a CPAN Testers Birds of Feather (BOF) session.
Barbie's original announcement:
This BOF is for anyone who is or wants to be a CPAN Tester. A chance to air any thoughts for improving CPAN testing and highlight any common pitfalls experienced by current testers. Also a chance for testers and authors to meet face to face, if they haven't already done so. Any contributions to improve the http://cpantest.grango.org/ are also welcome.
I would also like to take this opportunity to get group photo of all the CPAN Testers attending. Last year we had http://perl.grango.org/images/dscf9239.jpg, so please come along and get your face on the front page of the http://perl.grango.org/ site
This will very likely be at the end of the first day of the conference, ie. the Wednesday.
If you are attending YAPC::EU and are interested in CPAN Testing, please add your name to the wiki page
Maybe not a definitive list, but this is what I have gleamed:
Kindly pointed out by Dave Golden:
Added by Jos Boumans:
Added by Simon Bertrang:
Set up a new smokebox this weekend, Dragonfly BSD 2.0.
Tidy.
I uploaded CPANPLUS::YACSmoke recently, along with POE::Component::CPANPLUS::YACSmoke, to CPAN.
So what is the fuss and how is it so different from CPAN::YACSmoke ?
Well, it is a port/reimplementation of CPAN::YACSmoke to use a CPANPLUS::Dist distribution class called CPANPLUS::Dist::YACSmoke. This backend deals with all the stuff that CPAN::YACSmoke used to deal with, such as munging test reports, deciding whether to send reports if we have sent them already, etc, etc.
And it also handles checking whether we have reported a PASS grade already for a prereq and if so it skips running the test-suite for that prereq again.
This functionality has been in CPAN.pm for a little while, and I felt it should be easy to implement with CPANPLUS.
That said, I did have lots of advice and patience from Jos Boumans to assist me. Cheers!
Seems good in initial testing. I'm starting to migrate my smokers over to using it and so far the speed of smoke testing has increased considerably.
Hurrah.
Niamh Olivia Williams was born on the 25th June, 4 days ahead of schedule. >:)
Sleep is a distant memory.
Also, this made me choke myself laughing.
DrHyde pointed out on #cpantesters-discuss channel that there has been an announcement by Robert Spiers regarding the cpan-testers mailing list.
For those who receive their recent uploads notifications via email, this may be a PITA.
Some alternative mechanisms already exist:
Those are the ones I can think of off the top of my head.
Earlier this week, Dave Cantrell sent the one millionth CPAN Test Report.