[gentoo-dev] Portage in Prolog: how about logic programming in Python?

[gentoo-dev] Portage in Prolog: how about logic programming in Python?

[Jean Jordaan]

Hi all

I saw drobbins's mention of Prolog as a possibility for future
portage, because of the need for logic programming. As a Python
lover, I have to point at the Logilab.org - Python Logic SIG,
implementing logic programming in Python.
   http://www.logilab.org/projects/python-logic

 From the site:

"""
The goal of the "unofficial" Logic-SIG is to provide logic programming and 
constraint-propagation features in Python. If you are interested in 
participating, please join the mailing list.

1. Available software

     * Logilab's constraint package is a full-python constraint satisfaction 
solver similar that what Oz/Mozart implements.
     * PyLog from Christophe Delord is a first order logic library in Python
"""

-- 
Jean Jordaan
http://www.upfrontsystems.co.za


--
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] Portage in Prolog: how about logic programming in Python?

[Pieter Van den Abeele]

[-- Attachment #1: Type: text/plain, Size: 2141 bytes --]

On 20 Nov 2003, at 09:50, Jean Jordaan wrote:

> Hi all
>
> I saw drobbins's mention of Prolog as a possibility for future
> portage, because of the need for logic programming. As a Python
> lover, I have to point at the Logilab.org - Python Logic SIG,
> implementing logic programming in Python.
>   http://www.logilab.org/projects/python-logic

Hi,

I'm writing portage core in (iso-compliant) prolog as part of my thesis 
at the free university of brussels, I'll make my papers with 
implementation available to gentoo.
I'm not familiar with the libraries you suggested, but Pylog seems very 
interesting, because at first sight it is able to interpret 
iso-compliant prolog. However I saw the following warning:

WARNING: The cut doesn't work well in the current PyLog version 
(backtracking over a cut doesn't undo the unifications done before the 
cut). This problem seems hard to fix with the current structure of 
PyLog and won't be tackled soon.

since cut is a rather crucial part of prolog, (and surely affects 
performance), I doubt that Pylog implements enough of the iso standard 
(efficiently?) to run a real prolog program. Same goes for the 
constraints package, which seems to provide only limited support. We 
also need to investigate whether bugs are fixed rappidly enough...

I was under the impression that prolog has very few dependencies and is 
able to interact with most languages (including python), so I guess the 
need to really include it in python or some other language is rather 
low, but I'm not excluding anything :-)

Regards,

Pieter

> From the site:
>
> """
> The goal of the "unofficial" Logic-SIG is to provide logic programming 
> and constraint-propagation features in Python. If you are interested 
> in participating, please join the mailing list.
>
> 1. Available software
>
>     * Logilab's constraint package is a full-python constraint 
> satisfaction solver similar that what Oz/Mozart implements.
>     * PyLog from Christophe Delord is a first order logic library in 
> Python
> """
>
> -- 
> Jean Jordaan
> http://www.upfrontsystems.co.za
>
>
> --
> gentoo-dev@gentoo.org mailing list
>

[-- Attachment #2: Type: text/enriched, Size: 2189 bytes --]

On 20 Nov 2003, at 09:50, Jean Jordaan wrote:


<excerpt>Hi all


I saw drobbins's mention of Prolog as a possibility for future

portage, because of the need for logic programming. As a Python

lover, I have to point at the Logilab.org - Python Logic SIG,

implementing logic programming in Python.

  http://www.logilab.org/projects/python-logic

</excerpt>

Hi,


I'm writing portage core in (iso-compliant) prolog as part of my
thesis at the free university of brussels, I'll make my papers with
implementation available to gentoo. 

I'm not familiar with the libraries you suggested, but Pylog seems
very interesting, because at first sight it is able to interpret
iso-compliant prolog. However I saw the following warning:

<fontfamily><param>Arial</param>

WARNING: The cut doesn't work well in the current PyLog version
(backtracking over a cut doesn't undo the unifications done before the
cut). This problem seems hard to fix with the current structure of
PyLog and won't be tackled soon. 

</fontfamily>

since cut is a rather crucial part of prolog, (and surely affects
performance), I doubt that Pylog implements enough of the iso standard
(efficiently?) to run a real prolog program. Same goes for the
constraints package, which seems to provide only limited support. We
also need to investigate whether bugs are fixed rappidly enough...  


I was under the impression that prolog has very few dependencies and
is able to interact with most languages (including python), so I guess
the need to really include it in python or some other language is
rather low, but I'm not excluding anything :-) 


Regards,


Pieter


<excerpt>From the site:


"""

The goal of the "unofficial" Logic-SIG is to provide logic programming
and constraint-propagation features in Python. If you are interested
in participating, please join the mailing list.


1. Available software


    * Logilab's constraint package is a full-python constraint
satisfaction solver similar that what Oz/Mozart implements.

    * PyLog from Christophe Delord is a first order logic library in
Python

"""


-- 

Jean Jordaan

http://www.upfrontsystems.co.za



--

gentoo-dev@gentoo.org mailing list


</excerpt>

Re: [gentoo-dev] Portage in Prolog: how about logic programming in Python?

[Jean Jordaan]

'lo Pieter

Thanks for taking a look!

> However I saw the following warning:
> 
> WARNING: The cut doesn't work well in the current PyLog version 

Ah, but as usual with these open source things the next line is
"Any collaboration is welcome ;-)" .. perhaps your prolog zen is
deep enough to disolve the difficulty the PyLog devs saw :]

> I was under the impression that prolog has very few dependencies 
> and is able to interact with most languages (including python), 

My concern as a relative outsider is just that I love to be able
to browse the source of the glue that holds my system together,
namely portage, with some chance of understanding it .. which is
way higher with Python than with Prolog .. that said, PyLog parses
Prolog .. so much for that idea I guess.

Regards,
-- 
Jean Jordaan
http://www.upfrontsystems.co.za


--
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] Portage in Prolog: how about logic programming in Python?

[Paul de Vrieze]

[-- Attachment #1: signed data --]
[-- Type: text/plain, Size: 1134 bytes --]

On Thursday 20 November 2003 17:02, Jean Jordaan wrote:
> 'lo Pieter
>
> Thanks for taking a look!
>
> > However I saw the following warning:
> >
> > WARNING: The cut doesn't work well in the current PyLog version
>
> Ah, but as usual with these open source things the next line is
> "Any collaboration is welcome ;-)" .. perhaps your prolog zen is
> deep enough to disolve the difficulty the PyLog devs saw :]
>
> > I was under the impression that prolog has very few dependencies
> > and is able to interact with most languages (including python),
>
> My concern as a relative outsider is just that I love to be able
> to browse the source of the glue that holds my system together,
> namely portage, with some chance of understanding it .. which is
> way higher with Python than with Prolog .. that said, PyLog parses
> Prolog .. so much for that idea I guess.

What is the point in interpreting one language (prolog) in another interpreted 
language (pyton)? It can't be anything but inefficient.

Paul

-- 
Paul de Vrieze
Gentoo Developer
Mail: pauldv@gentoo.org
Homepage: http://www.devrieze.net

[-- Attachment #2: signature --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

Re: [gentoo-dev] Portage in Prolog: how about logic programming in Python?

[Aron Griffis]

[-- Attachment #1: Type: text/plain, Size: 578 bytes --]

Paul de Vrieze wrote:	[Fri Nov 21 2003, 03:37:16PM EST]
> What is the point in interpreting one language (prolog) in another interpreted 
> language (pyton)? It can't be anything but inefficient.

I can't claim to be either a Prolog or Python expert, but I'd imagine
the point is better integration between the languages.  Implementing
Prolog in Python would probably allow for very easy integration between
modules.

Aron

-- 
Aron Griffis
Gentoo Linux Developer (alpha / ia64 / ruby / vim)
Key fingerprint = E3B6 8734 C2D6 B5E5 AE76  FB3A 26B1 C5E3 2010 4EB0


[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]