> >With this system you can declare a structure like
> >
> > struct('foo', (('a', 'char'), ('b', 'double')))
> >
> > <more snip!>
Mark Hammond replied:
> This sounds very nice.
I agree, it's a great idea. Unfortunately Donald tried to sell it to
me as a complete overhaul of Python's type system, with his own
incompatible naming and building conventions, etc. It took me some
time to figure out that underneath all that there was actually very
useful functionality. Donald keeps telling me that it's the
functionality he's after, not the implementation. Maybe this interest
in his code will convince him to sumbit a more "Python-friendly"
implementation?
> I am toying with the concept of being able
> to call an OS API (Under MSWindows) from Python, without C support
>
> (simiar to the way SmallTalk and Visual Basic can - ie, name the
> function, and just call it, with the params magically working themselves
> out).
I did something like that for systems supporting SunOS ? SVR4 dynamic
loading. Unfortunately without the power of a type-describing system
like Donald's it's kind of hard to support functions that take
structure arguments...
> So count me as one more Python user supportive of this :-)
Hear, hear!
--Guido van Rossum, CWI, Amsterdam <mailto:Guido.van.Rossum@cwi.nl>
<http://www.cwi.nl/cwi/people/Guido.van.Rossum.html>