Christopher H. Laco wrote: > Adriano Ferreira wrote: >> On 8/17/07, Christopher H. Laco <[EMAIL PROTECTED]> wrote: >>> chromatic wrote: >>>> On Thursday 16 August 2007 20:14:31 Eric Wilhelm wrote: >>>> >>>>> I am certain that more than one 'extra tests' directory is needed, thus >>>>> the thought to make them into subdirectories (objections?) >>>>> >>>>> (They cannot live under 't/' due to compatibility issues.) >>>> Which compatibility issues? I've used t/author/ for quite a while without >>>> any >>>> compatibility issues. >>>> >>>> -- c >>>> >>>> >>> Right, accept when your t/author tests are tests for you, the author and >>> my t/author tests are for my Author class/classes/pages/etc. >> Testing some Author stuff would be rarer than having author tests. So >> maybe we could standardize on something like "t/author" and when other >> value is desirable, a key/value pair may be specified in META.yml (and >> in Makefile.PL/Build.PL). >> >> # META.yml >> --- >> >> # the default location for author tests (not needed) >> author_tests: t/author >> >> # when something different is wanted >> author_tests: t/developer >> >> > > That convinces me even more that me other email was spot on imho. > Looking for a directory and assuming what's in it are an authors' tests > is pure guesswork. It's rude that you assume whatever directory I have > isn't used for other purposes and will ignore those tests on make test. > > *.t files need a way to declare themselves as an authors' test. Let the > runner/harness inspect/ask each *.t file if it's an author test or not, > and run it or not...regardless of what directory it's in. >
To whit: Who knows better whether foo.t is an author test or not [in the absence of the actual human author]... The harness, or foo.t? The harness asking the t files will always be more accurate than the harness just find t files somewhere. But I digress...
signature.asc
Description: OpenPGP digital signature
