[imp-hackers] LOOP non-compliance

[Juan Jose Garcia-Ripoll]

On Mon, Apr 9, 2012 at 12:01 AM, Stas Boukarev <stassats at gmail.com> wrote:

> > Now, seriously, do you expect to develop _in a mailing list_, the next
> LOOP
> > implementation? Perfectly compatible with the standard, the new semantics
> > is not disruptive, and with the same detail of specification?
>
So, where do you propose to do it? It's not the new implementation of
> LOOP, it's what all implementations have been doing for years, until
> somebody read the spec more carefully.
>

Reading LOOP's specification "more carefully" will not introduce the
semantics you wish. Right now the order of statements is "variable
iteration" followed by "main statements" = groups of conditions plus
imperative statements. You want to move the conditions anywhere, but are
not specifying how this should behave in relation to the order of
evaluation of the variable initialization and iteration and the existing
imperative and conditions. So it is _not_ a reinterpretation of what there
is, but a new behavior that has to be specified fully, in relation to past
behavior and all the existing clauses.

Currently there is no forum to change the standard. The closest to it is
the CDR project http://cdr.eurolisp.org/ which documents desired
extensions. It is not just chatting on a mailing list but somebody
committing to writing how some new behavior is introduced. What I do not
like about it is that it is not possible to vote, but that is the closest
to introducing a change in a rational way in which various implementations
may agree.

I do not have time to do that.

Juanjo

-- 
Instituto de Física Fundamental, CSIC
c/ Serrano, 113b, Madrid 28006 (Spain)
http://juanjose.garciaripoll.googlepages.com
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mailman.common-lisp.net/pipermail/implementation-hackers/attachments/20120409/f45161fe/attachment.html>