[Openmcl-devel] Query about ASSERT and COMPILER-LET
Michał "phoe" Herda
phoe at disroot.org
Thu Apr 16 15:12:23 PDT 2020
I always ask Quicklisp about where it pulls its sources from.
In this case:
This means that https://gitlab.common-lisp.net/iterate/iterate is the
place we want to submit our patches to.
On 17.04.2020 00:11, Robert Goldman wrote:
> Thanks. In that case, I think it should be relatively easy to fix
> ITERATE to walk this form.
> Unfortunately, I don't know who is maintaining that package now -- the
> common-lisp.net page lists only a darcs repository, which is kind of
> like saying it's available on Betamax...
> On 16 Apr 2020, at 16:16, Ron Garret wrote:
> I’m running v1.12-dev.5-23-gb25693fa on Catalina and AFAICT the
> code for ASSERT has not changed, and still invokes COMPILER-LET.
> COMPILER-LET is an undocumented CCL extension, but AFAICT from
> looking at the source it is safe to simply treat it as if it
> expanded as follows:
> (defmacro compiler-let (bindings &body body)
> (declare (ignore bindings))
> `(progn , at body))
> On Apr 16, 2020, at 12:53 PM, Robert Goldman <rpgoldman at sift.info
> <mailto:rpgoldman at sift.info>> wrote:
>> In the SHOP3 planner, I use the ITERATE package for looping. When
>> I compile SHOP3 on my Linux test machine with 1.11, I get
>> compiler warnings, because ITERATE can't successfully code walk
>> CCL's code.
>> When I look at the warning messages, it looks to me as if
>> |ASSERT| calls are being translated into a complex s-expression
>> wrapped in |ccl:compiler-let|, which ITERATE doesn't know how to
>> However, on my laptop, which runs macOS Catalina, and CCL 1.12, I
>> do *not* get these compiler warnings.
>> So I was wondering: going forward can I rely on this being OK?
>> Has the compiler-let rewrite been removed from CCL? Or is this
>> something that is related to the OS, so that it will continue to
>> appear on Linux?
>> I would like to know because if this is something that will
>> continue to appear, I can try to add support for
>> |ccl:compiler-let| to ITERATE. But this might be quite difficult,
>> and I would prefer not to, unless it's absolutely necessary.
>> Openmcl-devel mailing list
>> Openmcl-devel at clozure.com <mailto:Openmcl-devel at clozure.com>
> Robert P. Goldman
> Research Fellow
> Smart Information Flow Technologies (d/b/a SIFT, LLC)
> 319 N. First Ave., Suite 400
> Minneapolis, MN 55401
> Voice: (612) 326-3934
> Email: rpgoldman at SIFT.net
> Openmcl-devel mailing list
> Openmcl-devel at clozure.com
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the Openmcl-devel