Re: [Sbcl-help] (in-package ...) in 2.3.2 (fwd)

[Andrey G. Grozin]

Hello *,

fricas cannot be compiled with the new sbcl-2.3.2. Can something be done
about it?

Andrey

---------- Forwarded message ----------
Date: Sat, 4 Mar 2023 04:40:20 -0500
From: Douglas Katzman via Sbcl-help <sbcl...@lists.sourceforge.net>
Reply-To: Douglas Katzman <do...@google.com>
To: Marco Antoniotti <marco.an...@unimib.it>
Cc: sbcl <sbcl...@lists.sourceforge.net>
Subject: Re: [Sbcl-help] (in-package ...) in 2.3.2

As an academic question, this is a valid point- changing a present symbol from internal to external should potentially be allowed. 
But in this case, Fricas violates the actual technical specification because the do-symbols loop exports symbols that were not
present in the package. Export is specified to first perform import: "If it is accessible as an internal symbol via use-package, it
is first imported into package, then exported." And I think it's pretty clear that importing is "interning"

On Sat, Mar 4, 2023 at 2:08 AM Marco Antoniotti <marco.an...@unimib.it> wrote:

Hi 

The restriction only mention interning and un uninterning.  Not exporting.  Does un/interning affect or is affected by
exporting?

MA



On Fri, 3 Mar 2023 at 20:42, Douglas Katzman via Sbcl-help <sbcl...@lists.sourceforge.net> wrote:
Fricas seems to violate http://www.lispworks.com/documentation/HyperSpec/Body/03_f.htm
In particular, line 37 of 'fricas-package.lisp' says:

(do-symbols (x "FRICAS-LISP") (export (list x)))


It is not technically permissible to export symbols in a package while iterating over symbols in that package.
I have not verified that removing this unportability will fix the problem, but I suspect it will



On Fri, Mar 3, 2023 at 1:46 PM Grégory Vanuxem <g.va...@gmail.com> wrote:
Hi,

(in-package ...) or its availability seem to have been modified in
SBCL version 2.3.2.

(in-package ...) in FriCAS, which is a modified version from what I
have read in a GitHub issue, is no longer imported. FriCAS used to
compile smoothly with SBCL, but no longer since 2.3.2.

Has 'in-package' been modified or something that leads to:
"boot:in-package is unknown" or something like that. So I guess
'in-package' is no longer imported in usual FiCAS settings.

Any hints?
__
Greg


_______________________________________________
Sbcl-help mailing list
Sbcl...@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sbcl-help

_______________________________________________
Sbcl-help mailing list
Sbcl...@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sbcl-help

--
Marco Antoniotti
https://dcb.disco.unimib.it

[Grégory Vanuxem]

I gave a workaround in FRICAS GitHub issues. But frankly it's just a "test case" not a real solution I think. 

--
You received this message because you are subscribed to the Google Groups "FriCAS - computer algebra system" group.
To unsubscribe from this group and stop receiving emails from it, send an email to fricas-devel...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/fricas-devel/73155aa8-3977-9a1-376c-b2d7bb4d1889%40starnew.inp.nsk.su.
--
You received this message because you are subscribed to the Google Groups "FriCAS - computer algebra system" group.
To unsubscribe from this group and stop receiving emails from it, send an email to fricas-devel...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/fricas-devel/73155aa8-3977-9a1-376c-b2d7bb4d1889%40starnew.inp.nsk.su.

_______________________________________________
Sbcl-help mailing list
Sbcl...@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sbcl-help

--

You received this message because you are subscribed to the Google Groups "FriCAS - computer algebra system" group.
To unsubscribe from this group and stop receiving emails from it, send an email to fricas-devel...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/fricas-devel/73155aa8-3977-9a1-376c-b2d7bb4d1889%40starnew.inp.nsk.su.