gcl-devel
[Top][All Lists]
Advanced

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

Re: [Gcl-devel] Re: trouble with cl-launch


From: Faré
Subject: Re: [Gcl-devel] Re: trouble with cl-launch
Date: Fri, 13 Oct 2006 02:21:24 -0400

On 12/10/06, Camm Maguire <address@hidden> wrote:
(let* ((compiler::*compiler-default-type* (pathname ""))
       (compiler::*lsp-ext* "")) (compile-file "/tmp/y"))
works in 2.7 at least.
And not in 2.6. Thanks a lot! That's exactly what I was looking for.
Maybe it deserves being documented better?

(si::save-system "foo")
Great! I've implemented system dumping for cl-launch, however, I have
the following problems:
* cl-launch calls gcl with option -batch, and this causes gcl to NOT
re-read system::*tmp-dir* from (si::getenv "TMP"). I had to explicitly
redefine it, but you probably know a better fix (especially if other
variables need to be reinitialized), and the real fix might be on your
side.
* gcl 2.7 seems to have a buggy make-pathname, that tries to merge a
relative :directory with the directory of the default, which causes
clc cached objects to be in the wrong place
 (make-pathname :directory '(:relative "foo" "bar") :defaults
(make-pathname :directory '(:absolute "baz" "quux")))
Should return #P"foo/bar/" but instead returns
#P"/baz/quux/foo/bar/"

That's all for now. I'm sure I'll find more bugs along the way. My
benchmark will be to successfully run Exscribe through cl-launch.

[ François-René ÐVB Rideau | Reflection&Cybernethics | http://fare.tunes.org ]
"How many libertarians does it take to change a lightbulb?"
A- "None, the market will take care of it."
B- "Every one of them and non-libs too, because we all *are* the market."
C- "*I*'ll do it, for a dollar."




reply via email to

[Prev in Thread] Current Thread [Next in Thread]