Why does one need a Makefile.PL to evaluate the distribution? One should be able to run it just fine with 'prove -lr t', 'perl -Ilib t/foo.t' etc.
However, that README doesn't sound right -- that sounds tailored towards developers (which really belongs in something like CONTRIBUTING), whereas a README should be targeted towards *users* -- containing instructions that users need to install and run the distribution. (This is the type of content what the core [Readme] plugin generates.) On Fri, Dec 06, 2013 at 07:01:32PM -0500, James E Keenan wrote: > On 12/4/13 3:12 AM, Francesc Guasch wrote: > >On Tue, Dec 03, 2013 at 07:09:16PM -0500, James E Keenan wrote: > > > > >>2. Is the source code publicly available somewhere, e.g., github.com? > > > >I don't know what I was waiting for. Here it is now: > > > > https://github.com/frankiejol/Test-SQL-Data > > Okay. Now, although I will probably catch hell from the Toolchain > Gang for saying this, the fact that your distribution is structured > *only* with dzil poses a barrier to my evaluating it. > > Your README says that to install one should call 'dzil test', then > 'dzil install'. However, for 'dzil test' to complete I have to > install someone's (not your) authordeps bundle: > > $ dzil authordeps | cpanm > ... > --> Working on Dist::Zilla::Plugin::VersionFromModule > Fetching > http://search.cpan.org/CPAN/authors/id/C/CJ/CJM/Dist-Zilla-Plugins-CJM-4.20.tar.gz > ... OK > > That's more work than I want to do to evaluate a module. If you > supplied a Makefile.PL, that would give me an incentive to take a > second look. > > Thank you very much. > Jim Keenan -- "Perfection (in design) is achieved not when there is is nothing more to add, but rather when there is nothing more to take away." - Antoine de Saint-Exupery . . . . . Karen Etheridge, ka...@etheridge.ca GCS C+++$ USL+++$ P+++$ w--- M++
