[Seaside-dev] Failing tests - WAPlatformTest>>#testReadWriteStream*

[Julian Fitzell]

As I said today, if we don't have any senders of them and they break,
I'm happy to remove them. We should, though, make sure to document the
methods on the portability page as methods not to use and maybe add a
test using #position/#position: for storing/restoring the position,
since we do use that?

I'm pretty certain I added the tests I did because I came across a
sender somewhere, but maybe they've been removed or maybe I'm
mistaken. I do see a sender of #position in
WAHttpVersionTest>>testStreamHeaderLineParsing. Aha! I wrote that...
that's probably when I added the test for #position:. :) I also found
senders of #atEnd at the same time, which is why I added those.

So I just replaced "position: 12" with "next: 12"... (committing that now)

I'll let John comment on the tests he added...

Julian

On Tue, Sep 1, 2009 at 10:17 PM, James Foster<Smalltalk at jgfoster.net> wrote:
> Julian & John,
>
> We have had three failing tests in Pharo for some time and I believe that
> they were introduced by the following:
>
> Seaside-Tests-Platform-jf.83
> Add tests for stream methods #position and #position: (defined by ANSI) and
> #upToEnd (not defined by ANSI).
>
> Seaside-Tests-Platform-jok.84
> Add platform tests for the stream answered by WAPlatform current
> readWriteStream.
>
> Some time ago we discussed the possibility of avoiding #position and
> #position: because GemStone has a (pre-ANSI) one-based implementation of
> streams (the position is the index of the next element rather than the count
> consumed). I believe that other protocol ought to be adequate for most uses
> (#atEnd, #skip:, #peek, etc). If we find we need position-based access that
> relies on the actual value, then we should go to a seaside-based name for
> the method.
>
> Note that the only Seaside use of #position and #position: that I could find
> is safe (JSJsonParser>>#match:).
>
> Could we remove the two erroring tests?
>
> One test is failing, WAPlatformTest>>testReadWriteStreamNextPutAll. Could
> you confirm that the test accurately describes the expected ANSI behavior,
> and if so flag it as an expected failure (or submit a fix to Pharo)?
>
> James
> _______________________________________________
> seaside-dev mailing list
> seaside-dev at lists.squeakfoundation.org
> http://lists.squeakfoundation.org/mailman/listinfo/seaside-dev
>