From zellerin at gmail.com  Thu May  5 09:53:28 2005
From: zellerin at gmail.com (Tomas Zellerin)
Date: Thu, 5 May 2005 11:53:28 +0200
Subject: [cl-prevalence-devel] cl-prevalence on clisp
In-Reply-To: <8deee56e72c669bfcd1f218b1ffd7c06@common-lisp.net>
References: <c0d688ae05042601594219287@mail.gmail.com>
	<8deee56e72c669bfcd1f218b1ffd7c06@common-lisp.net>
Message-ID: <c0d688ae05050502536de7eba3@mail.gmail.com>

Okay,

I enclose a patch that does following to port cl-prevalence to clisp:
- solves problem of probe-file being an error when called on directory
(clisp speciality)
- defines serializable-slots for standard classes and structures
- defines needed process and socket functions
With the patch, I was able to succesfully run all tests in demo1 and
demo2. However, I do *not* suggest including this patch to
distribution as whole - there are too many things that are not clean
there (including internal clos:: function in serializable-slots). Take
it more as a suggestion of what would be needed to port to clisp, and
include what you see fit.

The patch is against cvsroot version from yesterday, and works on
linux, clisp 2.33.2. Newer clisp may have correct mop functions, but I
do not have it compiled atm. You need bindings module (linux:) and
socket functions in clisp; the former can be modified to ffi.

Tomas Zellerin

P.S.: You know you have two #+sbcl branches in serializable-slots?

On 4/26/05, Sven Van Caekenberghe <scaekenberghe at common-lisp.net> wrote:
> On 26 Apr 2005, at 10:59, Tomas Zellerin wrote:
> 
> > Hello,
> >
> > I tried yesterday to run cl-prevalence on recent clisp. I found that
> > following does not work:
> > - serialization of classes, the reason being that serializable-slots
> > for standard objects has no #+clisp section. I was able to overcome
> > this particular problem, can make and send patch, if clisp is also
> > considered as target for this package
> > - serialization of pathnames. I could make plausibly looking
> > serialization code, but did not get how to handle de-serialization.
> >
> > Is the latter problem of clisp only? Is it supposed to work?
> >
> > With regards,
> >
> > Tomas Zellerin
> 
> I would certainly welcome new ports, so please do send any patches you
> have. If I have time, I will test on CLISP. However, since CLISP has no
> multi[processing|threading], part of cl-prevalence will not work
> (luckily this additional/optional stuff).
> 
> Adding serialization code for very specific CL type would also be
> appreciated ;-) Be sure to add some unit tests as well, thx!
> 
> Sven
> 
> --
> Sven Van Caekenberghe - mailto:sven at beta9.be
> Beta Nine - software engineering - http://www.beta9.be
> .Mac - svc at mac.com - http://homepage.mac.com/svc
> 
> "Lisp isn't a language, it's a building material." - Alan Kay
> 
>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: cl-prevalence.patch-clisp
Type: application/octet-stream
Size: 12133 bytes
Desc: not available
URL: <https://mailman.common-lisp.net/pipermail/cl-prevalence-devel/attachments/20050505/dfd41dd8/attachment.obj>