Re: [Chicken-users] functions for hash tables

[Larry White]

A third ideal to aspire to is ease of porting between scheme
implementations.  I would like to be able to develop with MzScheme and
have the option of deploying with chicken.

I know that's kind of a pipe dream, but there's not much point in
making it harder.  Some existing MzScheme software (like the swindle
clos implementation) would be wonderful to have in chicken.

I agree that the scheme naming is more "scheme-like".  The changes to
tiny-clos setters for example are better when viewed in isolation, but
they do make porting more diffiicult.


On Wed, 26 Jan 2005 15:44:44 -0500, Ed Watkeys <address@hidden> wrote:
> 
> On Jan 26, 2005, at 2:41 PM, address@hidden wrote:
> 
> > Hi all.
> >
> >
> > Here is a useful comparison of hash tables in different Scheme
> > implementations (great work, Shiro):
> 
> I think there are two ideals to aspire to:
> 
> (1) Does Chicken let people accurately guess procedure names given an
> understanding of basic concepts and terms?
> 
> (2) Does Chicken let people apply their knowledge of (a) other Schemes,
> (b) other languages, (c) operating systems, err, UNIX?
> 
> Scheme has list-ref, therefore hash-table-ref, not hash-table-get, is
> the way to go according to (1). One could argue that given (2),
> hash-table-get is the way to go, if most other Scheme implementations
> use hash-table-get. Insanity awaits those who follow (2) at the expense
> of (1). Terminological precision is an imperative if we hope to avoid
> the PERLification* of Chicken or Scheme in general.
> 
> Ed
> 
> * The standard libraries of Python and probably every other free
> high-level language in use today are pretty much as bad as PERL's in
> terms of poor adherence to naming standards, but I give credit to PERL
> (and CPAN) for being the first to achieve this state.
> 
> _______________________________________________
> Chicken-users mailing list
> address@hidden
> http://lists.nongnu.org/mailman/listinfo/chicken-users
>