Some nice assumptions here... are you actually suggesting that: - could ever be a stable release of web2py out there that will hold the lock for any pending select ? - the book suggesting a good behaviour is absolutely wrong ?
Web2py developers and testers in general would spot that in a breeze while testing trunk, and it'll be fixed in the following stable release. There are no evil people among web2py trying to make queries go slower, block wherever they want your code, and generally having to make you code crappy workarounds. Anyway, if you are so concerned about those problems and want to mantain your own forked DAL version where you are in control of every aspect, be my guest..... but it kind of defeat the "I'll adopt a framework" paradigm. When you face something like frameworks, you have to "give some credit" to the fact that n contributors, often very well versed in at least the piece of code they submitted/tested/patched - the 90% of the times surely more informed than yourself - did a good job. --
