On Tue, Nov 28, 2000 at 02:13:47PM -0800, Guido van Rossum wrote:
229 is not sufficiently worked out and doesn't seem a priority. (This can still make it into 2.1 if the author does the work.)
I was waiting for further comments on the idea before actually doing the work. Anyone see gaping holes in the basic idea? (I *can* just charge ahead and work on it anyway...)
I'd say charge ahead, but check where you are every once in a while. Some things I'm not sure about include:
- Always shared libs. What about Unixish systems that don't have shared libs? What if you just want something to be hardcoded as statically linked, e.g. for security reasons? (On the other hand there are some reasons why defaulting to shared libs is a good idea -- e.g. it makes it easier to upgrade a single extension module in an installation, e.g. for a critical fix.)
- Doesn't distutils currently always look for an existing installed Python installation to get its Makefile info? How to force it to use the info from the build tree?
BTW off-topic: I didn't demote your PEP 222 (Web Library Enhancements) but I'm pretty neutral on this -- except that it's probably a good idea to start from scratch with a new cgi-ish module (cgi2.py?) rather than making the existing cgi.py even more complicated. (Also, it's very fragile -- you never know who uses what mis-feature.)
--Guido van Rossum (home page: http://www.python.org/~guido/)