On Saturday, 2002-05-18 at 22:43:31 +0800, Autrijus Tang wrote: > On Sat, May 18, 2002 at 01:06:06PM +0200, Lupe Christoph wrote: > > On Saturday, 2002-05-18 at 08:06:42 +0100, Matt Sergeant wrote: > > > -----BEGIN PGP SIGNED MESSAGE----- > > > Hash: SHA1 > > > Perl 5.7.3 is known to be broken. Please use the current bleadperl instead. > > Please do not use Beta versions of perl for CPAN testing. And especially > > not bleadperl, as it is Alpha code.
> I agree that automatic cpansmoke should better run under perl 5.6.1 > or 5.005_03. However, as there is a 'perl executable configuration' > directory in http://testers.cpan.org/, wouldn't it also benefit > authors to know that their works are broken by 5.7.3, hence having > a chance to correct it before 5.8 arrives? You are assuming that what 5.7.3 breaks, 5.8.0 will break, too. I sure don't hope so. I don't have the time to see how many modules don't work with 5.7.3 or bleadperl that do work with 5.6.1. Have any guess? > > If you must test with 5.7.x or bleadperl, please do not send > > the results to the database. > 5.007_03 222 > 5.007_02 23 > 5.007_01 466 That's what I mean. Now how representative are those results? > Since the database query interface can't tell bleadperl from pointperl, > I think not testing under bleadperl is valid. But 5.7.3 does provide > many useful data points, in my very humble opinion... bleadperl looks like a patched 5.7.3 in Config.pm. And it is. Until 5.8.0 gets released. > Would it be better if we can post a policy on http://testers.cpan.org/? That would sure be better. There has been talk about the cpantest policy since cpantesters was incarnated. But like the old saying goes - put three cpantesters in a room and you have seven policies... Lupe Christoph -- | [EMAIL PROTECTED] | http://free.prohosting.com/~lupe | | I have challenged the entire ISO-9000 quality assurance team to a | | Bat-Leth contest on the holodeck. They will not concern us again. | | http://public.logica.com/~stepneys/joke/klingon.htm |
