[IronPython] Re: CPython's test_builtin results summary

Michael Spencer mahs at telcopartners.com
Mon Apr 4 18:52:06 PDT 2005


Martin Maly wrote:
> Hi Michael,
> 
...
> 
> I can see your emails just came through... Since you have spent the time
> and modified the tests so that they can readily run on IronPython,
> filing the actual bugs is not necessary because we can simply run your
> modified test and fix whatever the test indicates failure on. It will
> save you the time needed to file the bugs and it will save us the time
> having to deal with the bugs one by one rather than just run long repro
> and proceed as tests fail.
> 
Thanks, that's what I hoped you would say ;-)

FWIW, I started out by trying to run some existing algorithmic modules (i.e., 
not dependent on external modules), and quickly ran into various problems with 
builtins.  That was one of my reasons for taking the tack I did.


> At this early stage there are still many useful programs that people can
> build with IronPython 0.7. These are newly written programs using .NET
> libraries rather than existing Python programs. Right now we're most
> interested in capturing the problems from these early users both to
> understand issues with .NET integration and to fix blocking issues as
> quickly as possible so they can get on with writing code. This is a new
> set of issues where we need to build experience and can't rely on the
> existing test suites to cover us as we push towards 1.0.

OK

Michael










More information about the users-ironpython.com mailing list