Re: YAML Tiny for M::B?
David Golden <[email protected]>
| Newsgroups | gmane.comp.lang.perl.modules.module-build |
|---|---|
| Message-ID | <[email protected]> |
On Sat, Sep 5, 2009 at 12:35 PM, David E. Wheeler<[email protected]> wrote: > On Sep 5, 2009, at 4:42 AM, David Golden wrote: >> I suspect that would be more robust than YAML, since META.yml isn't >> YAML and Parse::CPAN::Meta is the read guts of YAML:Tiny anyway. > > If it works, what difference does it make? Is that not just shaving yaks? If Parse::CPAN::Meta is the target 'canonical' parser (by virtue of being in core), then it's important to produce YAML that it can read. Here is the result of an analysis I did a while ago on how different parser interpret META.yml files -- these are the "problem" ones, where parsers either failed to parse or didn't agree. Mouse over the blocks to see info text on what the columns are: http://echo.dagolden.com/~xdg/meta-yaml-testing/ So, it's yak shaving, to some extent, yes, but with "correctness" as a goal. David