some comments for Python 3000

[lobozc at my-deja.com]

In article <8n3ppd$rv0$1 at slb7.atl.mindspring.net>,
  "Arthur Siegel" <ajs at ix.netcom.com> wrote:
> >I think Python is already quite good at that (win32all for MS,
> >JPython for Java, CPython for Linux). But, of course, >improvements
can
> (and will) be made here.
>

I wasn't implying that Python is bad :-) - otherwise I wouldn't bother
to write that post.

In fact I believe Python as a core language is good is several
categories (apart from the ones you mention):
  - it allows clean expressiveness because of lists as first-class
objects and passing functions as parameters
  - it allows a set of high expressiveness constructs with its list
operations
  - it has clean and uncluttered syntax
  - yet it allows 'hacking' by doing some awful things to classes at
runtime :-)

Coming objectivisation of strings and list comprehensions are further
steps on that way.

Still, I'm afraid that a) this may be not enough to get Python
popularity it deserves; b) considered improvements should be in the
area of the language more than it's library

Examples here would be a) implementation of generators (like in icon)
[they can be imitated by classes but generators are simply more concise
for small things]; b) consideration of some form of goal-directed
evaluation [icon, again, saves plenty of coding]. Smaller examples
could be extension of the functionality of lambda functions [rewrite
model for them?] as well as consideration for some higher model of
string handling (icon again - they are talking about that in Perl
group) [which could be combined with lists and goal directed eval]

All of the above rather as examples of possible development direction
than proposed features.

> Comments I will take as an opportrunity to ask after the
> realities surrounding JPython's development.
>
> Sourceforge now hosts two sites - one with a familiar
> and impressive list of developers (affiliated with PythonLabs) and 0
> activity, and that of Finn Bock - who is quietly incorporating fixes,
adding
> compatibility features, answering users questions and generally
keeping
> JPython alive.
>
> Shouldn't the community be discussing the situation and its
> implications?
>
yes, I second that
in more general terms it is also a problem how to keep python on the
three platforms (unix, ms, java) together as much as possible.


Sent via Deja.com http://www.deja.com/
Before you buy.