[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Scheme-reports] DISCUSSION/VOTE: The character tower

On Tue, 2014-05-06 at 18:48 -0400, John Cowan wrote:

> You may not like the specification, but it requires the value #f
> in this case, at least if the procedure call completes without an error.

> Same story:  #f is required.

Well, with all due respect, that's breakage.  That is, IMO, 
a misconception about what strings are and an elevation 
of representation details over semantics.  

That will have programs making distinctions among canonically
identical strings, with the result that there'll be constant
bugs when searching and pattern matching return false negative 
results.  And this is *REQUIRED* behavior, not just permitted,
so this is a bug that language implementers aren't even allowed
to fix?!

This goes well beyond "not like the specification" - this 
is into the realm of "apparently do not comprehend the basic
values informing the specification."

I think that forbidding the bug fix is incomprehensible, 
and must reveal some agenda or value completely alien to 
my whole way of thinking. Therefore I don't believe that 
I am capable of meaningfully contributing to a discussion 
of these strings.  

Apparently there is no common ground in values to work 
with.  With this requirement, it becomes clear that I 
fundamentally don't understand what the committee would 
like these strings to be or do, so no contribution I can 
make would be directed toward helping to achieve whatever
that goal might be. 


Scheme-reports mailing list