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

Re: A liitle note on the side

This page is part of the web mail archives of SRFI 55 from before July 7th, 2015. The new archives for SRFI 55 contain all messages, not just those from before July 7th, 2015.



campbell@xxxxxxxxxxxxxxxxxxxxxxxxxxx wrote:
> If 'the typing argument' isn't the main issue, _what_is_?

That's my question too.

It doesn't standardize existing practice; the name, syntax, and
semantics are different from similar features in existing Schemes. It
ignores the standard feature identifiers introduced by SRFI-0 and
SRFI-7, so it isn't even compatible systems that provide those.

Furthermore, I think it's unlikely that Scheme vendors will support
SRFI-55 well enough to make it common practice in the future. Contrary
to Felix's claims, it takes more than a simple macro to implement this
even in Schemes that could support it. For example, there is no simple
way to load SRFI-1 in PLT Scheme; the language imposes special
requirements because SRFI-1 overrides standard Scheme procedures. They
differ depending on whether you want to load SRFI-1 at top-level or in
module scope.

This SRFI is gratuitously incompatible with previous specifications, and
it's much harder to implement than the author has claimed. It duplicates
features that Scheme implementations already provides, but with an
incompatible syntax and, in some cases, very difficult semantics. I see
it as no improvement over existing practice, and in some cases a
significant step backwards.
-- 
Bradd W. Szonye
http://www.szonye.com/bradd