ANN: YAPDL

[Alex Martelli]

Istvan Albert <ialbert at mailblocks.com> wrote:

> Alex Martelli wrote:
> 
> > Python BUT would want to see points A, B and C changed: make your OWN
> > language, derived from Python but with all the changes you wish.  As
> 
> > If you want a new experimental language that's sort like Python, spin
off your own --
> > DON'T post here criticizing Python or trying to change it.  
> 
> This sounds very much like an excuse for not listening
> to criticism.

It would be technically easy, say for me, to avoid listening to the
unending stream of criticism and demands for change that has been in
this forum and its predecessors "since forever" -- killfiles work well
for that.  I still do see some problems in leaving those diatribes
systematically unanswered, but, ah well, if somebody can't be bothered
to do a tiny bit of homework, a little research on googlegroups say, and
see how his criticisms and demands have been made and answered a
thousand times in the past, I guess they deserve what they get (or
don't).   But as far as actual requests for help and guidance this group
can be quite a bit more helpful -- I have no problem answering the same
beginner question for the 100th time -- if I don't have to start
killfiling systematically to avoid wasting my time in the 100th flamewar
about braces versus indentation, and the like.

> It is also a somewhat of a conceited statement
> since not everyone has either the time or the expertise
> needed to start their own language. What then?
> Should those that can't make a better python just shut
> up? Is their view always wrong?

Not always, but frequently enough that cluttering this forum with those
unending discussions substantially reduces its usefulness for its best
purposes.  The new languages are of course going to be much more open to
any suggestion for change and innovation than a language like Python, so
the new languages' discussion forums are going to be much more suitable
than c.l.py for all of that jazz.  If something gets proven in the field
in another, experimental language, that is close in other ways to
Python, then obviously that's a much better argument for it than any
amount of diatribe and verbiage, which is generally all the critics in
question appear to be up to offering.

So, it would be of advantage to all if the critics went criticizing
elsewhere.  Anywhere but here and python-dev, in fact.


> This sort of (IMO flawed) reasoning: 'if you don't like it make
> your own' is often heard in the open source world. Arguably  it is the
> cause of so many redundant and unfinished pieces of software
> that replicate 90% of each other's functionality yet lack the
> maturity that would make them useful

I entirely disagree with you: I see it as the cause of the existence of
Python (Guido didn't waste time on diatribes on Perl forums -- and Perl
itself, Larry ddn't waste time on diatribes on Awk forums...!), Gnome
(there was a lot of diatribe on KDE forums but eventually the Gnome
people did the right thing), and most other worthy technical innovations
that didn't happen to be the historically first attempt in their fields.

But here we're talking about examples where restarting from scratch was
clearly warranted because the desired differences were so huge.  More
significant and relevant to the main topic are cases like OpenBSD, which
doesn't at all "lack maturity" and indeed IS very useful -- I personally
prefer it to FreeBSD for many security-crucial applications.  If Theo
had spent his time ranting and raving rather than doing, we would not
have OpenBSD and I, for one, think that would be very sad.

If the would-be innovators don't want to "stand on the shoulders of
giants" by reusing that 90% of the other technology that they DO like,
when that's feasible (and opensource makes it feasible), they're
obviously pretty poor in judgment and technical savvy, infected with a
bad case of NIH syndrome or something of that ilk.

If they DO reuse, it's always possible (in an opensource perspective) to
selectively fold some changes back -- those which have proven their
worth and are judged compatible with the original 'trunk'.  Much, much
better to experiment with all of this _outside_ of the mainstream
development of a language with millions of users and tens of millions of
existing lines of code still in production and needing maintenance!

 
> The last thing the world needs is another python-like language
> with say braces or without decorators...

Absolutely not: it needs yet another diatribe on this forum about say
braces (decorators are a quite different case as they're still new and
almost experimental) MUCH LESS than it needs anything else.


Alex