Re: Perl Magpie, a next-generation CPAN Testers reporting platform

On Fri, Mar 28, 2025, at 11:20, James E Keenan wrote:
> On 3/28/25 10:03, Ruth Holloway wrote:
> > I. Rationale
> > [snip]
> > V1.0: The basics -- Hope to have this completed and deployed prior to 
> > PTS May 1:
> >    * Accept a properly-formatted test result from cpanm-reporter, using 
> > the Metabase transport pointing at its address. (DONE)
> 
> In our current setup, when a test result arrives at Metabase, does a 
> report generated by cpanm-reporter look any different from one generated 
> by cpan-reporter or cpanplus-reporter?
> 
> If not, then shouldn't we be able to send results generated by 
> cpan-reporter *already* be acceptable by Magpie?

I'm going to hedge here, and say "*probably."  *Both Task::CPAN::Reporter and App::cpanminus::reporter use Test::Reporter::Transport::Metabase under the hood, so it should work.  The 1.x string after the basics will be time for us to be *certain* that's the case, and make sure that all extant toolkits work.  My expectation is that Task::CPAN::Reporter won't need anything, just the same config changes that a cpanm-reporrter user would make.

> I ask because, while I used cpanm-reporter within my "CPAN River 3000" 
> project in the late 2010s, I've been using Task::CPAN::Reporter in 
> recent years and would prefer not to have to change.

Absolutely understood; part of our philosophy from the get-go on this was to pick *one* toolkit, get that working, and then make sure the other paths work--in your case, that should just be a matter of assurance testing.  There may be other kits where something more will be needed, obviously.

> Thanks for this well thought-out proposal.

You're very welcome!  
--Ruth