[tibix]

I don't understand all this hatred against web2py. For some months I've been working on it and I didn't found it any harder, slower, nor unsafer than other frameworks. Can somebody please finally give me reasons not to use them? Instead of a simple "Don't use it. You'll have major headaches in the future", like in http://us.pycon.org/2011/schedule/presentations/152/

[ubernostrum]

The biggest problem is that web2py openly encourages learning practices which do not transfer over into general Python programming and, in fact, are the literal opposite of how actual Python programmers work. Things like magic imports seem great at first, of course, but in the long run A) get you lost quickly if you ever have to write real Python and B) have the same problem any bit of hidden magic does: sooner or later something's going to go wrong and you're going to have one hell of a time figuring out which bit of implicit "helpful" code is causing it.

[pbreit]

Has anyone actually run in to this problem or is it speculative?

[viraptor]

It's like running with the scissors. I'm not sure if anyone ever hurt themselves that exact way, but any grown up person knows it's a bad idea and you should not try it (or a similar action). If the general idea can be explained and makes sense, do you really want to look for specific examples where doing the opposite failed?

You could probably check the number of questions about php's autoloader on stackoverflow to get an idea why magic is not good.