Ksenia Marasanova a écrit : > 2005/9/8, Sokolov Yura <[EMAIL PROTECTED]>: > >>Django Model is wonderfull. But SQLObject more flexible (and powerfull, >>as i think, and has already more db interfaces). >>But Django Model is tied with Django, and using Django with another OO >>mapping is not comfortable. >>Why do not working together? I can't understand. > > (snip) > > BTW, while SQLObject is very advanced, there are/were some other ORM > mappers in python: > http://www.thinkware.se/cgi-bin/thinki.cgi/ObjectRelationalMappersForPython > While not all of them share the same philosophy with SQLObject, some > do. But they are not merging together either. Just like web frameworks > :) > > I guess it's just the joy of creating something that fits you mind and > solves all you problems, instead of picking up something that does it > partially and was created by another person with different views and > philosophy. >
Also, there's something like darwinism at play here. Yes, there are a lot of concurrent ORM/Templating/Web Publishing/GUI/Whatnot projects around, but I guess only the best of them will survive - eventually 'absorbing' what's good in the others. I also noticed something like a 'converging evolution' scheme in Python these days. It strikes me that more and more frameworks seems to be based on concepts like intelligent properties assembled in schemas (Zope, Plone, Django, BasicProperties, SQLObjects...), interface adaptation (PEAK, Zope3, ...) and 'configurable polymorphism' (PEAK, Zope3, ...). Strange enough, I was developping a similar solution for a higher-level LDAP api when I noticed this... It seems to me that it has to do with recent evolutions of Python (decriptors, decorators, metaclasses made simple etc) beginning to be widely adopted and taken advantage of by Python programmers. Anyone else on this ? -- http://mail.python.org/mailman/listinfo/python-list
