Re: Re: SmartEiffel shortcomings?

In <1117126412.30533.28.camel@eden>, Roger Browne <[email protected]> typed:
> Mike Meyer wrote:
> > For other projects, tackling the bugs database is the
> > *recommended* way to start contributing to the project.
> That's not going to happen on any project where the test suite is kept
> secret.

That's what I was doing. It's not really that big a deal. If a test
suite is publicly available, it won't include the cases reported in
the bug - otherwise the bug would have been caught by the test
suite. So you have to write test code to verify the bug, and that
you've fixed it. Having a test suite available makes that a bit
easier, in that there's a framework you can plug things into. On the
other hand, having a test suite available also means you have to
include the new test in the patch.

This was what drove me to start developing a test suite for the SE
libraries. The near-complete lack of response caused me to abandon
that. If an open Eiffel compiler project does come about, I'll
certainly donate what I've done to the project, and possibly start
working on it again.

	<mike
-- 
Mike Meyer <[email protected]>		http://www.mired.org/consulting.html
Independent Network/Unix/Perforce consultant, email for more information.