[Bug 650283] Review Request: rubygem-rspec-core - Rspec-2 runner and formatters

[bugzilla at redhat.com]

Please do not reply directly to this email. All additional
comments should be made in the comments box of this bug.


https://bugzilla.redhat.com/show_bug.cgi?id=650283

--- Comment #4 from Mamoru Tasaka <mtasaka at ioa.s.u-tokyo.ac.jp> 2011-02-21 12:38:17 EST ---
(In reply to comment #3)
> At first, could you please explain how are you going to build all these
> packages? I assume that in first iteration, you are going to change
> need_bootstrap to 0 and in the second iteration, you are going to remove this
> constant and conditions completely, is that right?

- As you see "Depends on" field on this bug, with version-compatible
  rubygem-rspec-{expectations,mocks} installed as BR, bootstrapping
  is not needed.
  (So the reason that need_bootstrap craft is added to -core spec file
   is only that I want to make all rubygem-rspec-foo related spec file
   alike).

  ! Note that for rubygem-rspec-{expectations,mocks}, bootstrapping
    is always needed if we want to enable tests, everytime we upgrade
    the version of rspec related gems.

> * Bindir
>   - Please consider usage of '--bindir .%{_bindir} \' in the gem install
> command.
>     This would allow you to remove all the cruft beginning by line 86.
- Will do.

> * Autospec
>   - Since autospec command is deprecated and moreover completely useless, I
> would
>     suggest its remove.
>   - Moreover, autospec-2 is not aligned with MiniTest naming, where testrb2 is
>     used to avoid conflicts.
- Will rename to autospec2 (I won't remove this script completely "by myself"
  - as the upstream will anyway remove this in the later version)

> * Test suite
>   - The test suite is really tricky :/ I have tried to run all the specs and
> there
>     are failing not only for autospec part, also for formatters. I have created
>     several upstream bugs to cover this:
>     https://github.com/rspec/rspec-core/issues/318
>     https://github.com/rspec/rspec-core/issues/319
>   - Cucumber specs has several failures as well
>     https://github.com/rspec/rspec-core/issues/320
>   - It is not hopefully showstopper
>   - The selected subset works just fine
- Well, I am using
  $ ruby -rubygems -Ilib/ ...
  note that "-Ilib/" is used here, which may explain the difference of test
  results between you and me.

> * Documentation
>   - Files in -doc subpackage are not marked as documentation. That is reported
>     by rpmlint. They can by appropriately queried later:
>     $ rpm -qp -d noarch/rubygem-rspec-core-doc-2.5.1-1.fc14.noarch.rpm
> 
>   - It is messaged by rpmlint
- It is intentional. I always say that "%doc attribute in -doc subpackage
  is redundant" because
  - The name of rpm already says that the package is for documentation
  - If --excludedocs is specified with $ rpm -ivh (or similar effect is
    set in rpm config file by default), files marked as %doc won't 
    be installed,  although the admin is just about to install -doc 
    package to see document files, which is perhaps not expected.

-- 
Configure bugmail: https://bugzilla.redhat.com/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.