No macros in Python

[Laura Creighton]

> On Mon, 16 Dec 2002 14:56:37 +0100, Laura Creighton <lac at strakt.com>
> wrote:
> 
> >Does it go down any better this way?  'The language designer does not 
> >know better than I do, but he does know better than Laura Creighton, 
> >aged 18, arrogant as Satan and whose code I am likely to maintain?'
> 
> But how would the 18 year old Laura ever have learned, except by
> trying stuff out? ^.^ (I was arrogant as Satan at age 18 too, but I'd
> never have learned anything if I hadn't had the opportunity to try
> things my way and find out what worked and what didn't.)

Indeed.  Notice that I think every programmer should have a LISP system.
I just don't think it should be invoked by typing:

% python

> 
> >I don't have any lying around, but the problem was that every floor invented
> >its own particular syntax for doing things like Discrete Fast Fourier 
> >Transforms.
> 
> I still don't get this - how did it differ from every floor writing
> their own FFT function with its own calling conventions etc in
> Fortran?

It wouldn't have, except that nobody ever believed that they could
invent the complete-and-utter-Holy-Grail-of-Programming-Languages
out of hacking FORTRAN.  Even we weren't that crazy.

> 
> >We made the code resemble the hardware,
> >a lot, because we thought the hardware was cool. 
> 
> In what way? Were there any efficiency gains from doing that?

To use the hardware, yes, you saved hours, weeks with some calculations.

To program so that you have to have read the hardware specification,
understood it, and remembered it clearly if you wanted the best performance?
Adding layer and layer of special case functionality and refusing to
make a slower (if more intelligible) system?  No, we were idiots that way.
Very, very, very, smart idiots who wrote very fast code.  But we had the
'fork the project' problem in spades.  If one way was faster for your
typical usage and another was for mine, we would each write our own
macros and optimize to suit us.

> 
> >By the time a department 
> >of Physicists gets done with any language, in the 1970s at any rate, it all 
> >looks like FORTRAN anyway.
> 
> I'd've thought they'd have no trouble reading it then - don't all
> physicists learn Fortran as their first programming language? :)

I hope not any more ...

Fortran is rather more varied than you would think; and besides it doesn't
matter.  Every physicist had the goal to take the ugliness out of the
language and keep the speed.  Damned shame we could never reach consensus
over 'what is ugly'?

> >Lispers from the AI community who shared an
> >office with me held their noses ....
> 
> When I used to write Lisp for AI, it tended to look like C. I'm sure
> real Lispers would hold their noses at that too :)
> 
> -- 
> "Mercy to the guilty is treachery to the innocent."

Yours?  I like  'All that is needed for evil to prosper is for men of
good will to do nothing.' better ...

> Remove killer rodent from address to reply.
> http://www.esatclear.ie/~rwallace
> -- 
> http://mail.python.org/mailman/listinfo/python-list

Laura Creighton