3 ms·
Software methodologies that don't actually account for and work with the way humans actually behave (instead of some idealized notion of how they should) end up
by DannyBee 7y ago
Software methodologies that don't actually account for and work with the way humans actually behave (instead of some idealized notion of how they should) end up like this - tons of essays on how it's not really the methodology's fault, it's the implementations, etc.
It is the methodology's fault, by definition - they exist to be implemented, and if the in practice implementations don't work as intended, then something is wrong in the method, level of proscription, etc.
- skybrian 7y agoThat seems a bit strong. Coming up with a methodology that no business can screw up, no matter how loosely they follow it, seems like it would be very difficult? Is that a realistic standard of success? Maybe we could make an analogy to cookbook recipes. Some recipes require cooking skill and some practice to do them successfully. But if the recipe is good, you should be able to train at least some cooks to have consistent success with it. But I would guess that such a methodology wouldn't actually generalize to lots of different kinds of software. It might work for repeatedly building similar things using similar technology in a restricted domain.
- cyborgx7 7y agoExactly my thought as well. Office politics will always be a thing, and your methodology better acomodate to it.