Re: Adding dot-notation in an ISO conforming manner

On 12/09/2013 07:41 PM, Alan Baljeu wrote:
>> What you want is another way of writing.  Why not! Make a new
>> write option. You could write out {} and even , to a limited
>> degree: the arguments must not be atoms. ____________
>
> Just a thought: In general is there any downside to choosing to make
> new predicates rather than alter standard predicates?  For this case,
> keep write_canonical. In general I expect there is no downside and
> the upside is no worries about compatibility while plenty of freedom
> to disregard historical but inferior choices.

Seems I missed the mail quoted somehow. Yes, in general there is an
issue just keeping the existing predicates as is because the data
structures in SWI7 have changed. Lists are '[|]'(H,T), [] is not an
atom, there are strings, there is f() and there are dicts. If you don't
upgrade relevant system predicates to deal meaningfully with these new
things, you get a strange and inconsistent language. The only compromise
in SWI7 is to leave functor/3 and =../2 and add new predicates that
properly handle the new data.

I think it is much better to accept small issues (and quite a few people
have indicated that the issues are small) and keep the language consistent
than to get two somewhat disjoined languages with a lot of confusion.

	Cheers --- Jan