[plt-dev] Test failure

From: Robby Findler (robby at eecs.northwestern.edu)
Date: Thu Oct 29 23:12:51 EDT 2009

FWIW, I committed a change to shared-test.ss that was just wrong. I
missed shared.ss. Instead I should have disabled drdr for
shared-test.ss. (I can't fix this now; only able to read email, not be
online properly).

Robby

On Thursday, October 29, 2009, Eli Barzilay  wrote:
> On Oct 29, Matthew Flatt wrote:
>> Some change to the test suite has made "quiet.ss" stop printing
>> sections after `names'. That makes it more difficult to see where
>> the failure happens.
>
> That wasn't the problem -- it's `load-in-sandbox' that stopped showing
> the section names, the `names' section is just the last one that is
> not done in a sandbox.  Fixing it is easy, "testing.ss" has this line
> in the sandbox setup:
>
>     (e `(define real-error-port (quote ,real-error-port)))
>
> and a similar line should be added for `fake-error-port'.
>
> However, looking back at this change, I don't see what it's supposed
> to achieve.  The `real-error-port' thing is used in three places:
>
> * In "quiet.ss", it is used to display any real errors; this wasn't
>   changed in Jay's commit.
>
> * In "testing.ss" it is used (through `eprintf*') to print section
>   headers.
>
> * It is also used by `report-errs', when invoked by "quiet.ss", so the
>   error reports will actually show up (since all other reports not
>   shown when running the tests through "quiet.ss").
>
> I don't see the point in making the last two go to stdout rather than
> stderr.
>
> --
>           ((lambda (x) (x x)) (lambda (x) (x x)))          Eli Barzilay:
>                     http://barzilay.org/                   Maze is Life!
> _________________________________________________
>   For list-related administrative tasks:
>   http://list.cs.brown.edu/mailman/listinfo/plt-dev
>


Posted on the dev mailing list.