GC In Python: YAS (Yet Another Summary)

Andrew Dalke dalke at bioreason.com
Fri Jun 25 06:55:06 CEST 1999


Stuart Yeates <syeates at cs.waikato.ac.nz> said:
> Of course, if anything does pointer XORing or other crimes,
> tools such as these are unavaliable.

Here's one which isn't a crime but would still have problems.
One of the C libraries we use also supports FORTRAN.  To do this,
it internally maps all pointers to integer handles, and it is
these integers which are used in the client apps.

On the other hand, maybe supporting FORTRAN is a crime, but
that's the perfered language of the vendor's customer base;
mostly chemists.

So if Python switches to these sorts of GC, I'll have to stick
with an older (pre-GC) version.

						Andrew Dalke
						dalke at bioreason.com




More information about the Python-list mailing list