gcl-devel
[Top][All Lists]
Advanced

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

Re: [Gcl-devel] conditions/clos/gcl unified build patch and instru


From: Vadim V. Zhytnikov
Subject: Re: [Gcl-devel] conditions/clos/gcl unified build patch and instructions.
Date: Sun, 09 Jun 2002 09:25:37 +0300
User-agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; ru-RU; rv:1.0.0) Gecko/20020526

Camm Maguire пишет:
Thanks Vadim!

so I suppose

(make-package "COMMON-LISP" :use "LISP")

will not then allow us to import/export changes from LISP into
COMMON-LISP?

Take care,



With

(make-package "COMMON-LISP" :use "LISP")

COMMON-LISP automatically inherits all LISP's external symbols
including ones which can be added later. This is not quite
good since COMMON-LISP contents is fixed one and forever.
So I prefer to make

(make-package "COMMON-LISP" :use '())

and later explicitly import in COMMON-LISP only symbols we
need.

Finally I understood my problems with exporting/importing of NIL
symbol. The whole trouble was purely due to my limited
experience CL packages (previously I worked with Standard Liap
which has no packages at all). The problem is that
import/export instruction in CL has two a bit different forms

(import 's ...)

for single symbol and

(import '(s1 s2 s3 ...) ...)

for list of symbols. Certainly in such situation

(import 'nil ...)

is ambiguous and different CL implementations treats
such instruction in different fashion. Some, including
GCL, assume that 'nil here is the empty list and just
do nothing.  Others, like clisp, treat this like
a single nil symbol and successfully export/import it.
In any case

(import '(nil) ...)

works well in all implementations.

After I realized all this I was able to make preliminary
COMMON-LISP and COMMON-LISP-USER packages and even launched
ansi-test with some result better than immediate
error.  More details later.

Best wishes,

Vadim





reply via email to

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