[GHC] #1112: The testsuite setup is broken

[Lennart Augustsson]

No, it was not 6.6, it was HEAD.

So at a minimum, I think that the summary of a log from the nightly
run should be checked in together with the source of the tests.

The fact that there are 0 failures for a release is of very little
interest.  If you are going to do any GHC development you are like
to grab HEAD and not a release version.  Setting up GHC it would be
very nice to know when you've reached the "normal" state.  Running
the tests would be useful for this, but only if there is some indication
what the expected outcome of running them is.  Right now there is no
such indication.

For instance, some of my failures were due to not having installed the
mtl package.

	-- Lennart


On Jan 22, 2007, at 05:57 , GHC wrote:

> #1112: The testsuite setup is broken
> ---------------------- 
> +-----------------------------------------------------
>  Reporter:  guest     |          Owner:
>      Type:  bug       |         Status:  closed
>  Priority:  normal    |      Milestone:
> Component:  Compiler  |        Version:  6.6
>  Severity:  normal    |     Resolution:  wontfix
>  Keywords:            |     Difficulty:  Unknown
>  Testcase:            |   Architecture:  Unknown
>        Os:  Unknown   |
> ---------------------- 
> +-----------------------------------------------------
> Changes (by simonmar):
>
>   * resolution:  => wontfix
>   * status:  new => closed
>
> Comment:
>
>  266 unexpected failures is certainly high.  Your bug report is  
> marked as
>  version 6.6: is it really a stock 6.6, or the 6.6 branch, or even  
> HEAD?
>  We only guarantee zero failures at a released version.  Between  
> releases,
>  test failures indicate bugs that remain to be fixed, or testsuite  
> sample
>  output that needs to be updated.
>
>  FYI, there were 165 unexpected failures on the HEAD on x86 last  
> night, and
>  27 on the 6.6 branch.
>
> -- 
> Ticket URL: <http://hackage.haskell.org/trac/ghc/ticket/1112>
> GHC <http://www.haskell.org/ghc/>
> The Glasgow Haskell  
> Compiler_______________________________________________
> Glasgow-haskell-bugs mailing list
> Glasgow-haskell-bugs at haskell.org
> http://www.haskell.org/mailman/listinfo/glasgow-haskell-bugs