02:33: #(2002 6 30 2 10)

I'm a bit disappointed that, in developing Koyemsi, I'm not (yet) using the run-time mutability inherent in Python--every time I want to change a bit of code, I go to a manually-loaded emacs-session, edit the source-file, and then exit and re-launch the program. That's pretty damned lame, isn't it? Python does give the ability to exchange a functions in place, and to dynamically build function-objects. Why am I not doing that?

I think that it's because there's no easy or straight-forward way to update both the running program and the source-file. Modifying the live program and then having it dump out to the source-file is basically out of the question, right now. I suppose that the problem with that idea is that I'm still trying to maintain readable source-files, including my style, and doing this would require parsing the files and testing equivalences and junk like that.... If I was to shift all of the code into a database, : In procedure car in expression (car syntmp-block-27): : Wrong type argument in position 1: #(... 2 31) like is done with MOO, Cold, and whatever..., and probably get comfortable with the system reformatting my code, it would work OK. Hm. I wonder if comments are preserved inside pyc-files.

[Reply]