On 11/01/2011 11:49 PM, Glyph wrote: > of "real" reactors. New contributors sometimes say that they don't feel > like the code is "really" being tested unless they're testing against > the real implementation of something, and it may not be immediately > obvious that testing against a not-real implementation is desirable in > many cases. Much of it can just be copy/pasted from this thread :).
Without commenting on everything else in the thread - this realisation was something I came to late, and I feel it would have helped enormously if it had come earlier. So, +1 on convincing people to go for in-memory / simulated transport/clock/etc. But I'll note that most people I've seen crib from existing test code during the early phases of TDD, tutorials or not - and if most test code in Twisted uses the real reactor... _______________________________________________ Twisted-Python mailing list Twisted-Python@twistedmatrix.com http://twistedmatrix.com/cgi-bin/mailman/listinfo/twisted-python
