This is the mail archive of the mauve-discuss@sourceware.cygnus.com mailing list for the Mauve project. See the Mauve home page for more information.


[Date Prev][Date Next][Thread Prev][Thread Next]
[Date Index] [Subject Index] [Author Index] [Thread Index]

Re: Character test (incomplete)



> 
> >>>>> "Godmar" == Godmar Back <gback@cs.utah.edu> writes:
> 
> Godmar> Are classpath & libjava implementing their own conversion &
> Godmar> output routines for floating point numbers?
> 
> libjava does.  As I recall, we took our implementation from newlib.

What is newlib, if I may ask?  In what language is it implemented?

> 
> Godmar> About the beans stuff: does this really test java.beans or
> Godmar> does it test certain properties of the specific classes (like
> Godmar> java.lang.Object) that are used as test cases?  If the latter,
> Godmar> placing the test in java.beans might be misleading.
> 
> I added these tests, but I don't have a clue about what they are
> doing.  I just took them from Classpath to see how hard it would be to
> convert their tests to this framework.  If these tests are broken, we
> should fix or remove them.

I didn't say or mean that they're broken, I'm just saying I also
don't know what they're doing.

> 
> Godmar> Oh, and Kaffe fails those parts of the Unicode stuff which
> Godmar> requires unicode tables; this is not yet implemented.  This
> Godmar> seems to be pretty separable, I might use steal classpath's
> Godmar> implementation for my personal use should that ever become a
> Godmar> necessity for me.
> 
> At some point we need to figure out a set of tags that can be used to
> describe "interesting" subsets of the full class library
> functionality.
> 
> For instance, we're going to want a way to avoid running any test that
> uses floating point (because we'll be able to build an FP-less
> library).
> 
> Likewise for threads.
> 
> Maybe Unicode support is a candidate for this.
> 

I agree, this seems to be very useful.
Would probably be useful for TVT too, as they have instances where Kaffe
is configured filesystem-less, for instance.

> 
> This does mean that we'll have to "artificially" split tests along
> lines according to these subsets, and not always along the most
> natural functional lines.  I don't think this is a real problem
> though.
> 

This again raise the question of how a given test is identified.
Right now, it's by invocation number, which, it seems, would make
such splitting hard to do.

See also my comments about the Thread.interrupt tests.

	- Godmar