Re: [gobo-eiffel-develop] getest fails when .epr exists but EIFGEN not

[Franck A. <fr...@ne...>]

Eric Bezault:
> But obsolete features and non-standard features are two different
> things. I think that it would be misusing 'obsolete' to use it
> to mark non-obsolete non-standard features.

That's right. (If a vendor adopted the idea that extensions to 
kernel classes are not desirable, maybe most of their non 
standard features would indeed be obsolete, but that could 
still leave a few features which are non-standard but not 
obsolete.)

> Or would it be better to have unicode suppport directly in
> the standard classes (and hence avoid the need for UC_STRING)?

Of course that'd be better, my question was only about what 
is preferable pending such direct support.

--
fr...@ne...