Re: extending redefinition

[acken@informatik.uni-kl.de]

> From: Guy Laden <guy@math.tau.ac.il>
> Regarding "extended" and "standard" modes for the compiler, if
> extended means Oberon/F compatibility and standard means ETH
> compatibility then fwiw I'd vote for it.

> From:	hartmut Goebel <hartmut@oberon.nbg.sub.org>
> Even when I'm new on the list: are you writing a compiler or designing a
> language? I thought the first, so the answer is obvious.

The non-standard compilation mode would
1) prevent WITH from changing the formal parameter list of a procedure
2) allow to compare a procedure variable with a procedure value
3) implement a slightly less restrictive rule for type-bound procedure
   redefinition

For one thing I wouldn't call the above points "designing a new
language".  They only remove unnecessary quirks in the language
resp. ETH compiler design.  Point (1) isn't even part of the language
report, it's a failure of the ETH compilers.

On the other hand the changes aren't intended to deliver Oberon/F
compability, although they might close the gap somewhat (or fully?  I
don't know O/F good enough to decide this).

-- mva