Re: [ANNOUNCE] Test::Builder/More/Simple 0.72
On Sep 22, 2007, at 9:06 AM, Salve J Nilsen wrote: - Set up some kind of syndication feed [RSS, Atom] [somewhere sensible] where one can read which distributions currently have a "pre-release" status. The feed might be called "NEED TESTING" or something like that. Make this feed visible on central Perl-related websites (Perlmonks, use.perl.org, etc.) and make it easy for Perl Monger groups to include this info on their websites. - Add some way for authors to state what the purpose of a module release is (e.g. "stable", "security update", "pre-release", "dev- release") and make this available in the distribution META.yml file. Well, one option might be something like: http://www.cpanforum.com/tags/name/helpwanted Gabor, would it be easy to add an Atom/RSS feed for a particular tag? Chris
Re: [ANNOUNCE] Test::Builder/More/Simple 0.72
On 9/25/07, Chris Dolan <[EMAIL PROTECTED]> wrote: > Well, one option might be something like: >http://www.cpanforum.com/tags/name/helpwanted > Gabor, would it be easy to add an Atom/RSS feed for a particular tag? Please no! Let's not spread module metadata around any more than we have to. Extend META.yml to include the same kind of information that used to be managed via the modules list. If someone wants to write a website to index and syndicate *that*, fine, but don't make authors go to multiple places to create and manage their metadata. David
Re: [ANNOUNCE] Test::Builder/More/Simple 0.72
On Mon, Sep 24, 2007 at 08:54:12AM +0200, Andreas J. Koenig wrote: > > On Thu, 20 Sep 2007 03:00:59 -0700, Michael G Schwern <[EMAIL > > PROTECTED]> said: > > > Most CPAN smoke testers wouldn't have caught it because even though they > often > > run alphas they usually don't install them. So the interactions with > > dependencies would be lost. > > Not true. My smokes would have caught it. And I'm very grateful. Certainly helped with DBI testing. Tim.
Re: [ANNOUNCE] Test::Builder/More/Simple 0.72
> On Thu, 20 Sep 2007 03:00:59 -0700, Michael G Schwern <[EMAIL PROTECTED]> > said: > Most CPAN smoke testers wouldn't have caught it because even though they often > run alphas they usually don't install them. So the interactions with > dependencies would be lost. Not true. My smokes would have caught it. -- andreas
Re: [ANNOUNCE] Test::Builder/More/Simple 0.72
> On Thu, 20 Sep 2007 03:04:14 -0700, Michael G Schwern <[EMAIL PROTECTED]> > said: > Well, the repository trunk is always kept at passing so if folks want to smoke > with that it's safe. Smoking repositories is not comparable with smoking release candidates. The number of possible collisions when smoking repositories against each other is close to infinity. Dev releases is the way to go. -- andreas
Re: [ANNOUNCE] Test::Builder/More/Simple 0.72
On 9/23/07, Michael G Schwern <[EMAIL PROTECTED]> wrote: > would have the absolute latest release, even devs. You could set your CPAN.pm > to pull from one or the other. If it goes this way, I'd at least want a command or something that lets me act against the alpha list as a one-off without having to edit my config. cpan> alpha look Test::Harness cpan> alpha report Test::Harness Let the debate on proper naming begin... David
Re: [ANNOUNCE] Test::Builder/More/Simple 0.72
Salve J Nilsen wrote: > - Set up some kind of syndication feed [RSS, Atom] [somewhere sensible] > where one can read which distributions currently have a "pre-release" > status. The feed might be called "NEED TESTING" or something like that. > Make this feed visible on central Perl-related websites (Perlmonks, > use.perl.org, etc.) and make it easy for Perl Monger groups to include > this info on their websites. > > - Add some way for authors to state what the purpose of a module release > is (e.g. "stable", "security update", "pre-release", "dev-release") and > make this available in the distribution META.yml file. I've been thinking about just that. The CPAN shell could give a user friendly description of how important an update is to the user. It would also get rid of the the nasty 1.23_45 alpha version convention. One of the biggest barriers is its difficult and annoying to get the CPAN shells to get alpha releases. That's because there's only one index tracking only the latest stable release. You could instead have two parallel CPAN indexes. One would be just the latest stables, what 02modules is now. One would have the absolute latest release, even devs. You could set your CPAN.pm to pull from one or the other. -- Whip me, beat me, make my code compatible with VMS!
Re: [ANNOUNCE] Test::Builder/More/Simple 0.72
Just a few thoughts Jonathan Rockway wrote: chromatic wrote: I have my doubts that most dev releases get *any* attention. I think the problem is that 99% of Perl users don't read mailing lists. They download stuff from search.cpan (maybe), and then forget about it until something breaks. I think this is an interesting problem. :) I'm guessing we have to somehow improve the channels where people DO read about the modules. I think this means mainly through search.cpan.org and through the different distribution tools (CPAN(PLUS)?|DPKG|RPM|etc). Most people are obviously interested in stable software, but if they can get the option to answer to "There is a pre-release of [module] on CPAN. Would you like to test it? [Y/n]" ..at the point they're about to download it, they might just answer "yes" to the question. If it's trivial to submit a test report, then I'd guess a few more than 1% of the users would "help" a little more. Maybe if we break stuff more often people will pay more attention? ;) Let's think "community management" - The search.cpan.org website could show the different kinds of releases more clearly. Not just a "This release field", but also a "Latest developer release" if there has been one. - Also, make it more clear what's it means to download a "developer release" or "pre-release", including making it trivial/obvious to see that feedback is welcome, where/how to do it. - Set up some kind of syndication feed [RSS, Atom] [somewhere sensible] where one can read which distributions currently have a "pre-release" status. The feed might be called "NEED TESTING" or something like that. Make this feed visible on central Perl-related websites (Perlmonks, use.perl.org, etc.) and make it easy for Perl Monger groups to include this info on their websites. - Add some way for authors to state what the purpose of a module release is (e.g. "stable", "security update", "pre-release", "dev-release") and make this available in the distribution META.yml file. There. I'll go hide beneath my rock again. - Salve
Re: [ANNOUNCE] Test::Builder/More/Simple 0.72
chromatic wrote: > I have my doubts that most dev releases get *any* attention. I have had the same experience with Catalyst. We asked users for months to test the dev releases of the new Authentication architecture to make sure it was 100% backwards compatible. No reports of anything. When the stable version was released, we had people whining about breakage the very same day. *sigh* I think the problem is that 99% of Perl users don't read mailing lists. They download stuff from search.cpan (maybe), and then forget about it until something breaks. Maybe if we break stuff more often people will pay more attention? ;) Regards, Jonathan Rockway signature.asc Description: OpenPGP digital signature
Re: [ANNOUNCE] Test::Builder/More/Simple 0.72
Ovid wrote: > For something as high profile as Test::Simple, it would get much more > exposure than most modules. Plus, there wasn't much in the changelog > which suggested this was a rush OMG Must Deliver Now sort of change (it > had been six months since the last release). I get a bug up my ass, spend a few hours going through the tracker, fix some bugs and release. Otherwise it's > And it would be nice to have this run against a substantial portion of > a minicpan install since breaking the toolchain is a very, very bad > thing to do (as Schwern as *repeatedly* pounded into my head :) Well, the repository trunk is always kept at passing so if folks want to smoke with that it's safe. -- The interface should be as clean as newly fallen snow and its behavior as explicit as Japanese eel porn.
Re: [ANNOUNCE] Test::Builder/More/Simple 0.72
Ovid wrote: > --- Michael G Schwern <[EMAIL PROTECTED]> wrote: > >> Because of this I'm TEMPORARILY rolling the fix back at least a week >> to give >> CPAN authors a little breathing room to make their fixes. The fix is >> usually to plan in a BEGIN block. > > This has happened a couple of times now. Why don't you use release > devel versions so that these issues become less serious in the future? I have, when it's something that looks important. It's scattershot whether it helps. Remember the big OMGWTFUBROKECPAN with Test::Builder::Tester? That was in two alpha releases for almost three months. Still caught everyone unawares. Also I figured this was a pretty obscure problem and just a bug fix. Most CPAN smoke testers wouldn't have caught it because even though they often run alphas they usually don't install them. So the interactions with dependencies would be lost. -- I have a date with some giant cartoon robots and booze.
Re: [ANNOUNCE] Test::Builder/More/Simple 0.72
On Thursday 20 September 2007 00:15:31 Ovid wrote: > > In your experience, do devel versions get the kind of testing that > > would catch these issues sufficiently? > > > > In my experience, they don't. > > For something as high profile as Test::Simple, it would get much more > exposure than most modules. I dunno. Something as high profile as a bleadperl release (or a maintperl RC) doesn't get much testing. I have my doubts that most dev releases get *any* attention. -- c
Re: [ANNOUNCE] Test::Builder/More/Simple 0.72
--- chromatic <[EMAIL PROTECTED]> wrote: > On Wednesday 19 September 2007 22:46:19 Ovid wrote: > > > This has happened a couple of times now. Why don't you use release > > devel versions so that these issues become less serious in the > future? > > In your experience, do devel versions get the kind of testing that > would catch these issues sufficiently? > > In my experience, they don't. For something as high profile as Test::Simple, it would get much more exposure than most modules. Plus, there wasn't much in the changelog which suggested this was a rush OMG Must Deliver Now sort of change (it had been six months since the last release). And it would be nice to have this run against a substantial portion of a minicpan install since breaking the toolchain is a very, very bad thing to do (as Schwern as *repeatedly* pounded into my head :) Cheers, Ovid -- Buy the book - http://www.oreilly.com/catalog/perlhks/ Perl and CGI - http://users.easystreet.com/ovid/cgi_course/ Personal blog - http://publius-ovidius.livejournal.com/ Tech blog - http://use.perl.org/~Ovid/journal/
Re: [ANNOUNCE] Test::Builder/More/Simple 0.72
On Wednesday 19 September 2007 22:46:19 Ovid wrote: > This has happened a couple of times now. Why don't you use release > devel versions so that these issues become less serious in the future? In your experience, do devel versions get the kind of testing that would catch these issues sufficiently? In my experience, they don't. -- c
Re: [ANNOUNCE] Test::Builder/More/Simple 0.72
--- Michael G Schwern <[EMAIL PROTECTED]> wrote: > Because of this I'm TEMPORARILY rolling the fix back at least a week > to give > CPAN authors a little breathing room to make their fixes. The fix is > usually to plan in a BEGIN block. This has happened a couple of times now. Why don't you use release devel versions so that these issues become less serious in the future? Cheers, Ovid -- Buy the book - http://www.oreilly.com/catalog/perlhks/ Perl and CGI - http://users.easystreet.com/ovid/cgi_course/ Personal blog - http://publius-ovidius.livejournal.com/ Tech blog - http://use.perl.org/~Ovid/journal/