On Sat, 11 Dec 1999, Fredrik Lundh wrote:
Greg Stein
wrote: There are all kinds of things that we can do here. Consider mmap'ing the archive into a shared memory segment, used by all the Python processes on the system... woo! :-)
it doesn't really look like this, but I hope we're defining interfaces here, and not just "one true solution". I'd be
Oh, I was just having fun there :-). I don't see "one true solution" at all. Just some standards.
very annoyed if it turned out that we couldn't use works' archives with the new standard importer...
get_code() and its processing is not going anywhere. Some stuff will change under the covers, and we'll be using sys.path (typically) rather than chaining (although chaining will still exist!). I would think that your Importer subclass would be directly usable, but the installation could/would be a bit different. Heck, worst case, nothing is going to invalidate your archive format -- feel free to berate me if I ever break that! Cheers, -g -- Greg Stein, http://www.lyra.org/