<br>If I were Arthur and trying to ensure my PyGeo source was "camera ready" from a code teaching point of view, I'd probably open source it on sourceforge, making the checkin/checkout process easy, and recruit apprentice PyGeo coders into the fold. They'd learn the ropes and propose enhancements or refactorings. There'd be more than one pair of eyes, more than one brain, working on the code. Design patterns such as we've been discussing, would get hashed out among insiders, less so in a generic public forum like edu-sig.
<br><br>As long as it's just Arthur writing everything, and soliciting advice from people who haven't made a concerted study of his work, I don't have any confidance that the code will be camera ready as some showcase for beautiful code. If this were Gerald de Jong, or Guido, or Tim Peters we were talking about, I might have a different assessment. But Arthur didn't grow up in computer world, is a johnny come lately, bearing much cultural baggage. He needs professional help, in my estimation.
<br><br>In my own case, I've posted working code, e.g. for a presentation manager using PyGame, that's obviously idiosyncratic, not peer-reviewed let alone peer-enhanced. I've been up front with my corresponding expectation: this source code won't be very useful to others -- except maybe in the case of the OSCON 2005 source wherein I buried some notes of historical interest, but that's a different notion of utility.