Re: [Liberty-eiffel] bug #44278

[Raphael Mack]

Am Samstag, den 02.01.2016, 20:16 +0100 schrieb Raphael Mack:
> In
> this case the class ARRAY could be compilable in the adler-version, as
> it is perfectly legal to create the THING_B instance with the
> dynamically bound feature make in slice. This might also invalidate my
> statement above... For me it looks acceptable to compile the adler code
> of ARRAY.slice iff all subclasses of ARRAY have this feature
> (considering renamings and redefinitions) in the list of creation
> procedures.
>
> For this case I think we have a compiler bug.

I should try before confirming a bug ;-) I think it works - even with
the adler-version of ARRAY.slice in case a (potentially
renamed/redefined) version of ARRAY.make is a creation procedure of the
subclasses...

Could you take a look at the test case I just committed:
http://git.savannah.gnu.org/cgit/liberty-eiffel.git/commit/?id=98b71e8ea229e9d19f56d6faf7ba0aec8c632f9f

and check whether I messed something up? I think this is the case you,
Laurie described in your last email, right?

Regards,
Rapha