[Ironruby-core] Code Review: rubyspec4

Jim Deville jdeville at microsoft.com
Mon Sep 29 13:54:34 EDT 2008


Right now, the only reason I keep our fork of MSpec is to not force everyone to change again. We are running MSpec completely unmodified. I actually run it out of the official repo. We have some changes to Rubyspec that I haven't pushed to the official repo due to lack of time. After we push those we may want to follow your idea. Some of our changes are to guards like you mention. Right now, most of the changes I make to RubySpecs are made to the official version then pulled into our repo, so I'm trying to head in that direction already.

As for tags, I've already asked Brian (and he has agreed) to tag releases, so we can have people sync to tags already.

The tags are a powerful concept, and eventually it'd be nice to tag failing tests with a unique tag, so you could run just the tests that demonstrate a bug. It's fully possible, we just aren't set up for it yet. I'd like to get the Rakefile cleaned up and more flexible to allow for that.


JD


> -----Original Message-----
> From: ironruby-core-bounces at rubyforge.org [mailto:ironruby-core-
> bounces at rubyforge.org] On Behalf Of Peter Bacon Darwin
> Sent: Monday, September 29, 2008 10:47 AM
> To: ironruby-core at rubyforge.org
> Subject: Re: [Ironruby-core] Code Review: rubyspec4
>
> It is possible that the many files that are only different in newlines
> is
> actually my fault.  I think that Jim integrated a load of my changes to
> the
> rubyspecs last month but I was having newbie issues with Git and my
> setup
> changed a load of the line endings.  I have now got over this issue
> after
> some mucking about with git config.
>
> I have subsequently had these changes incorporated into the official
> rubyspec repository on github.  If the only files are ones that I broke
> then
> you can just bin those commits and use the versions from the official
> github
> repository instead.
>
> Is there much benefit in keeping a separate fork of mspec and rubyspec
> now
> anyway?  Ironruby now seems to be able to run mspec unmodified and the
> guys
> who manage these projects seem happy to give responsible people commit
> rights to the official repositories.
> You could apply tags to versions of rubyspec or mspec to baseline
> versions
> that are good for ironruby.
> In addition there are a number of guards available in mspec that allow
> rubyspecs to be filtered based on ruby implementation and version.
> This is
> how the jruby and rubinius projects use rubyspec.
>
> In terms of tracking regression tests for a specific ironruby bug in
> rubyforge.  How about you just list the descriptions of the examples
> (i.e.
> "it" blocks) that fail for the bug.  It would be fairly easy to check
> that
> these now run ok.  In many ways the tagged rubyspec examples are a
> direct
> representation of the bugs that are still to be solved.
>
> Pete
>
> -----Original Message-----
> From: ironruby-core-bounces at rubyforge.org
> [mailto:ironruby-core-bounces at rubyforge.org] On Behalf Of Srivatsn
> Narayanan
> Sent: Monday,29 September 29, 2008 18:07
> To: ironruby-core at rubyforge.org
> Subject: Re: [Ironruby-core] Code Review: rubyspec4
>
> Using the tf diff tool, I see that a lot of files are identical except
> for
> changes in newline characters (maybe \n changed to \r\n)? From my
> random
> sampling I hardly found any files that have actually changed. Maybe it
> would
> help to have a guideline about the preferred newline character.
>
> Also, regarding the baselining, how are we planning to track bugs and
> their
> related regression tests? If we are closing a bug on rubyforge and want
> to
> make sure that regression tests exist for that scenario, it would be
> good to
> have a link between the disabled test and the bug itself. In the
> IronPython
> testcode this is done by adding a disabled decorator to the test with
> the
> bug id. Here maybe a tag can have one more field to indicate the bug
> id?
> This is a change to the mspec runner itself and I'm not asking for it
> to be
> done with this shelveset but it's something to be discussed about.
>
> -----Original Message-----
> From: ironruby-core-bounces at rubyforge.org
> [mailto:ironruby-core-bounces at rubyforge.org] On Behalf Of Peter Bacon
> Darwin
> Sent: Monday, September 29, 2008 9:48 AM
> To: ironruby-core at rubyforge.org
> Subject: Re: [Ironruby-core] Code Review: rubyspec4
>
> The diff has only 94 bytes
>
> -----Original Message-----
> From: ironruby-core-bounces at rubyforge.org
> [mailto:ironruby-core-bounces at rubyforge.org] On Behalf Of Jim Deville
> Sent: Monday,29 September 29, 2008 17:35
> To: IronRuby External Code Reviewers; Srivatsn Narayanan
> Cc: ironruby-core at rubyforge.org
> Subject: [Ironruby-core] Code Review: rubyspec4
>
> This is a large diff due to updating Rubyspec, MSpec and Ironruby-tags.
>
> tfpt review "/shelveset:rubyspec4;REDMOND\jdeville"
> Comment  :
>   Re-syncing MERLIN_EXTERNAL mspec to the head of MSpec to pick up new
> tests. Re-baselining to get new tests included.
>
>
>
> _______________________________________________
> Ironruby-core mailing list
> Ironruby-core at rubyforge.org
> http://rubyforge.org/mailman/listinfo/ironruby-core
>
> _______________________________________________
> Ironruby-core mailing list
> Ironruby-core at rubyforge.org
> http://rubyforge.org/mailman/listinfo/ironruby-core
>
> _______________________________________________
> Ironruby-core mailing list
> Ironruby-core at rubyforge.org
> http://rubyforge.org/mailman/listinfo/ironruby-core



More information about the Ironruby-core mailing list