This is the mail archive of the gcc-patches@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: typedef-named class member functions can be public


On Jul  5, 2002, Mark Mitchell <mark@codesourcery.com> wrote:

> I don't think we need this level of Chinese wall-ness.

Sorry, I got advice from Red Hat lawyers to proceed like this.  I
cannot, and will not derive testcases from proprietary testsuites I've
looked at.  I can't take this kind of risk.

Fortunately, the community (or at least one person from the community,
Brian R. Gaeke) has been contributing tests for postings like this.

> You can come up with a test case easily enough that doesn't copy the
> proprietary testsuite.

This is not enough for a clean-room development, and from my
understanding anything else is subject to copyright laws.

> If we do what you're suggesting, then we're all going to have to be
> writing each other's testcases.

Which happens to be exactly the right way to do testing, if you ask
me.  If the person who writes the feature also writes the test,
s/he'll probably miss a number of cases s/he hadn't though of.  Which
is precisely why testing teams should be separate from development
teams.

> In any case, the patch isn't correct, if I understand it correctly.
> You need to back and fix the linkage of the functions when the typedef
> is processed -- not later when the functions are defined.

Hmm...  Sounds reasonable, I'll try to do that.  Thanks,

-- 
Alexandre Oliva   Enjoy Guarana', see http://www.ic.unicamp.br/~oliva/
Red Hat GCC Developer                  aoliva@{cygnus.com, redhat.com}
CS PhD student at IC-Unicamp        oliva@{lsd.ic.unicamp.br, gnu.org}
Free Software Evangelist                Professional serial bug killer


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]