From metawilm at gmail.com  Fri Jun 13 09:37:20 2008
From: metawilm at gmail.com (Willem Broekema)
Date: Fri, 13 Jun 2008 11:37:20 +0200
Subject: [clpython-devel] Python just got a bit less dynamic
Message-ID: <f6bc9b490806130237r4f90403ud193bccf8d0bef1c@mail.gmail.com>

Guido just limited the dynamicness of Python in a very sensible way,
by declaring that the attributes of objects can only be regular
strings. For users this does not change much, but for implementors
this has significant benefits, so I like this limitation a lot.

The implementation of attributes in CLPython and other implementations
can now be cleaned up and made more efficient by removing hairy corner
cases. CLPython can now always represent attributes by Lisp symbols.
Also as a result, the five known failures in the CLPython test suite,
which tested some hairy corner cases, are now defined to not be
failures at all.

The decisions:

 - Using non-string keys is undefined behavior;
   non-empty strings that are not valid as identifier according to the
grammar are allowed:
   http://mail.python.org/pipermail/python-dev/2008-June/080309.html

 - String subclasses may be normalized to regular strings:
   http://mail.python.org/pipermail/python-dev/2008-June/080312.html

 - The restriction applies to both class dicts and instance dicts:
   http://mail.python.org/pipermail/python-dev/2008-June/080313.html

 - Frameworks currently using non-string attributes should change
their behavior;
   http://mail.python.org/pipermail/python-dev/2008-June/080319.html

- Willem


From jdn at math.carleton.ca  Sat Jun 14 21:13:58 2008
From: jdn at math.carleton.ca (Jason Nielsen)
Date: Sat, 14 Jun 2008 17:13:58 -0400 (EDT)
Subject: [clpython-devel] New errors in recent CVS
Message-ID: <Pine.LNX.4.64.0806141651560.27172@buddy>

Hello,

I just pulled the recent changes from cvs and get the following warning 
when building:

The constant CLPYTHON.PARSER::+ALL-VALUES-NAMES+ is being redefined
(from
(CLPYTHON.PARSER:+NORMAL-VALUE+ CLPYTHON.PARSER:+AUGASSIGN-VALUE+
  CLPYTHON.PARSER:+NO-VALUE+)
to
(CLPYTHON.PARSER:+NORMAL-VALUE+ CLPYTHON.PARSER:+AUGASSIGN-VALUE+
  CLPYTHON.PARSER:+NO-VALUE+))

which I accept and the build finishes.  When I run the test suite I get 
the following:

**********************************
End CLPython test
Errors detected in this test: 104 UNEXPECTED: 100
Successes this test:609
;
; compilation unit finished
;   caught 114 fatal ERROR conditions
;   caught 4 ERROR conditions
;   caught 3 WARNING conditions
;   caught 22 STYLE-WARNING conditions
;   printed 26 notes
NIL

There are a whole bunch of error messages when running the tests of the 
form:

Reason: detected an unexpected error of type `#<SB-PCL::CONDITION-CLASS 
SB-C:COMPILER-ERROR>':

which are the likely cause of the 100 extra failures.  All of this is on a 
Ubuntu 7.10 box with sbcl 1.0.17... I built the latest sbcl along with the 
previous tarball I had and the test suite ran as expected so the problem 
is in recent updates.

By the way I've been following Euro 2008 closely (I'm a huge 
soccer/football fan which isn't quite as common here in Canada) and the 
Dutch team are looking awesome!  Nothing like kicking the asses of the 
reigning world champs and runners-up to start off the tournament ;-).

Jason


From metawilm at gmail.com  Sat Jun 14 22:07:11 2008
From: metawilm at gmail.com (Willem Broekema)
Date: Sun, 15 Jun 2008 00:07:11 +0200
Subject: [clpython-devel] New errors in recent CVS
In-Reply-To: <Pine.LNX.4.64.0806141651560.27172@buddy>
References: <Pine.LNX.4.64.0806141651560.27172@buddy>
Message-ID: <f6bc9b490806141507o2a7a4ac3p2f2ad8d120e28ab2@mail.gmail.com>

Hi Jason,

On Sat, Jun 14, 2008 at 11:13 PM, Jason Nielsen <jdn at math.carleton.ca> wrote:
> End CLPython test
> Errors detected in this test: 104 UNEXPECTED: 100
> Successes this test:609

Oops... obviously I did not test the latest changes on SBCL. I just
committed fixes for SBCL, so it should run fine for you in an hour.

I am making changes to the compiler, in order to implement yield
expression soon - that is the most important missing language feature.

> By the way I've been following Euro 2008 closely (I'm a huge soccer/football
> fan which isn't quite as common here in Canada) and the Dutch team are
> looking awesome!  Nothing like kicking the asses of the reigning world
> champs and runners-up to start off the tournament ;-).

Yeah, that's going incredibly well! I hope the team continuous playing
like it did in those two matches. Btw here's a great picture of the
orange crowd: http://www.flickr.com/photos/irouge/2568725720/

Cheers,
- Willem


From jdn at math.carleton.ca  Sat Jun 14 23:17:14 2008
From: jdn at math.carleton.ca (Jason Nielsen)
Date: Sat, 14 Jun 2008 19:17:14 -0400 (EDT)
Subject: [clpython-devel] New errors in recent CVS
In-Reply-To: <f6bc9b490806141507o2a7a4ac3p2f2ad8d120e28ab2@mail.gmail.com>
References: <Pine.LNX.4.64.0806141651560.27172@buddy>
	<f6bc9b490806141507o2a7a4ac3p2f2ad8d120e28ab2@mail.gmail.com>
Message-ID: <Pine.LNX.4.64.0806141818030.27172@buddy>

On Sun, 15 Jun 2008, Willem Broekema wrote:

> Oops... obviously I did not test the latest changes on SBCL. I just
> committed fixes for SBCL, so it should run fine for you in an hour.

No worries, I'll take care of testing sbcl for you ;-)!  All fixed and 
back in business, thanks!

> I am making changes to the compiler, in order to implement yield
> expression soon - that is the most important missing language feature.

Great!  Any plans for python 3k?  The one PEP that looks very common lisp 
friendly is 3107 (function annotations) which should make clpython scream! 
I'm pretty excited about clpython as it looks like a tool I can use more 
generally in both my research and teaching.  I'm sick of doing mixed 
language programming in Fortran, C, R, Matlab... etc!  For various 
reasons, common lisp, which I appreciate greatly, isn't a language I can 
use much for either outside of my own personal tinkering... which sucks! 
Actually, I got interested in common lisp to help me generate bindings 
since it is generally just a bunch of tedious boilerplate and lisp was the 
perfect tool ... sadly I got hooked ;-)!  As an example of my plight I 
dare you for instance to try and teach numerical analysis to math majors 
using common lisp in the labs ... most of the students have only very 
rudimentary programming skills and what they do possess is usually some 
java or C++.. barf (the dare is a death wish ;-)!).

> Yeah, that's going incredibly well! I hope the team continuous playing
> like it did in those two matches. Btw here's a great picture of the
> orange crowd: http://www.flickr.com/photos/irouge/2568725720/

Indeed, if they keep playing such a fluid and dangerous counterattack they 
won't be easy to beat.  The sleeper team I'd keep my eye out for is 
Croatia (although I guess not as much of a sleeper anymore)... I've been 
impressed by their play.  May the orange be with you ;-)!

Jason