[Numpy-discussion] global overloading of 1+1 -> MyClass(1, 1)

Andrew Dalke dalke@dalkescientific....
Mon Aug 18 18:57:20 CDT 2008


On Aug 19, 2008, at 1:06 AM, Christian Heimes wrote:

> [long posting]
>
> Oh h... what have I done ... *g*

*shrug* I write long emails.  I've been told that
by several people.  It's probably a bad thing.

> The ideas needs a good PEP. You are definitely up to something. You  
> also
> came up with a list of possible issues and corner cases. Are you
> interested in pursuing the proposal? *wink*

No.  I'm doing this to head off endless discussion about
some proposed hypothetical future Python.  If someone wants
to experiment with new operators, new ways of handling
int/float/whatever constructions, new lexical forms, or
whatever then IMNSHO the best thing is to implement it and
see if it's actually useful.

In most cases there are seriously hard questions that
are hand-waved away, because it's more fun to talk about
language design than to implement one and because
it's hard to tweak CPython to try out new things.

Hence I wrote python4ply and an extensive tutorial
specifically so people could easily jump into the
implementation details, find the sticky points, then
either quickly reject a bad idea or iterate to get a
good solution.  Many of these changes can be done
with less than a couple of days of work.

Then to show that there are sticky points, I listed a few.

BTW, it's *fun* to modify an existing language and
afterwards it you know a secret - that programming
languages are just flimsy facades held together by
a shared hallucination.  Like in a dream, change
things too much or leave gaps and people notice the
illogic, wake up, and look elsewhere for refuge from
harsh digital machine reality.

Hmmm, it really is too late for me.

				Andrew
				dalke@dalkescientific.com




More information about the Numpy-discussion mailing list