Re: Module::Build cannot check required version of Parse::RecDescent 1.95.1

2008-05-20 Thread Dominique Dumont
[EMAIL PROTECTED] (John Peacock) writes: > In order to map 5.006001 to v5.6.1, all numeric versions (like 1.90) > are parsed as if there were three places to the right of the decimal > point. So 1.90 is equivalent to 1.900.0 for comparison purposes, which > is larger than 1.95.1. This is confusin

Re: Module::Build cannot check required version of Parse::RecDescent 1.95.1

2008-05-20 Thread John Peacock
Dominique Dumont wrote: This is confusing. Shouldn't 1.90 be mapped to 1.090.000 ? I know it is confusing and it's not my fault; when 5.6.0 was released, the documentation specifically stated that 5.6.0 === 5.006000. Besides, you haven't though this through; numeric versions have to [ultimat

test_requires

2008-05-20 Thread David E. Wheeler
Howdy, I'm working on a series of modules for a client. We developers need a bunch of test modules, but OPS will want to just install the module without running tests. I've been putting the test module requirements into `build_requires`: configure_requires => { 'Module::Build' => '0.27

Re: test_requires

2008-05-20 Thread Randy J. Ray
So, can we finally have a `test_requires` option? If so, I'd be happy to dig up some tuits and send in a patch. This would be very helpful indeed. Extra points for "test_recommends" for modules like Test::Pod, Test::Pod::Coverage, etc. Randy --

Re: test_requires

2008-05-20 Thread Eric Wilhelm
# from David E. Wheeler # on Tuesday 20 May 2008: >So, can we finally have a `test_requires` option? If so, I'd be happy >   to dig up some tuits and send in a patch. IIRC, previous discussions of this sort of thing have tended to feature-creep toward having also test_recommends and various othe

Re: test_requires

2008-05-20 Thread Thomas Klausner
Hi! On Tue, May 20, 2008 at 10:29:53AM -0700, David E. Wheeler wrote: > I've been putting the test module requirements into `build_requires`: > ... > But this will be annoying for OPS, since it's not the build that requires > those Test:: modules; it's testing. > > So, can we finally have a `tes

Re: test_requires

2008-05-20 Thread Eric Wilhelm
# from Thomas Klausner # on Tuesday 20 May 2008: >The general opinion was that three *requires are enough, especially as >testing seems to be seen as an integral part of the build process. This is the opinion of a QA get-together, so I'll guess that it is backed by the logic of "why wouldn't you

Re: test_requires

2008-05-20 Thread David Golden
On Tue, May 20, 2008 at 2:41 PM, Eric Wilhelm <[EMAIL PROTECTED]> wrote: > # from Thomas Klausner >>The general opinion was that three *requires are enough, especially as >>testing seems to be seen as an integral part of the build process. > > This is the opinion of a QA get-together, so I'll guess

Re: test_requires

2008-05-20 Thread Randy J. Ray
The general opinion was that three *requires are enough, especially as testing seems to be seen as an integral part of the build process. Perhaps if there were a "build_recommends", it might be. As it stands, a user often won't know that optional tests are going to be skipped until they see the

Re: Module::Build cannot check required version of Parse::RecDescent 1.95.1

2008-05-20 Thread Dominique Dumont
John Peacock <[EMAIL PROTECTED]> writes: > I know it is confusing and it's not my fault; when 5.6.0 was released, > the documentation specifically stated that 5.6.0 === 5.006000. > Besides, you haven't though this through; numeric versions have to > [ultimately] be compared as numbers (because the

Re: test_requires

2008-05-20 Thread David Golden
On Tue, May 20, 2008 at 11:03 PM, Adam Kennedy <[EMAIL PROTECTED]> wrote: > Personally, I'd like to see configure_requires: finalised first before we > start dealing with test_requires, so that we have the option of solutions > that themselves leverage configure_requires: +1 Reminds me I need to

Re: test_requires

2008-05-20 Thread Eric Wilhelm
# from David Golden on Tuesday 20 May 2008: >On Tue, May 20, 2008 at 11:03 PM, Adam Kennedy >> Personally, I'd like to see configure_requires: finalised first >> before we start dealing with test_requires, so that we have the >> option of solutions that themselves leverage configure_requires: > >+1