4 ms·

There is a quote by Don Norman on the matter that I like very much (from "The Design of Everyday Things"): > Requirements made in the abstract are invariably w

There is a quote by Don Norman on the matter that I like very much (from "The Design of Everyday Things"): > Requirements made in the abstract are invariably wrong. Requirements produced by asking people what they need are invariably wrong. Requirements are developed by watching people in their natural environment.

Spartan-S63 6y ago
Hm... sounds a lot like the scientific process and basic research :)
bob1029 6y ago
In my experience, requirements are never fully developed until after many iterations have been undertaken. You can never perfectly plan how an application or any other arbitrary convenience should be in isolation of the user. You can certainly get very close, but there is something about actually using a thing that cannot ever be fully simulated in an authentic way. The way a knob feels to a person is not a thing you can ever truly encapsulate and store in a database field. Constant feedback and iterating on that feedback is how you arrive at ideal outcomes. The art of developing software or other high-quality user experiences is a very iterative, chicken-egg-style problem. It is a human problem. Being prepared to throw it all away and start over each time you iterate is one of the most powerful things you could ever hope to do. Just keep a copy of what you tried before as a stick to lean on. Each time you iterate, that stick should be getting stronger. Eventually you wind up with a bunch of sticks and you can start to chop them up, reorganize them, and build a proper structure.
mc32 6y ago
Also it misses the point. During the USSR’s and China’s attempts at grand projects often times they knew about mechanized machinery, but that solution wasn’t always applicable (too costly, diverted too many resources, or they had ample idle hands that weren’t easily deployed to something more productive).

Spartan-S63 6y ago

Hm... sounds a lot like the scientific process and basic research :)

bob1029 6y ago

In my experience, requirements are never fully developed until after many iterations have been undertaken. You can never perfectly plan how an application or any other arbitrary convenience should be in isolation of the user. You can certainly get very close, but there is something about actually using a thing that cannot ever be fully simulated in an authentic way. The way a knob feels to a person is not a thing you can ever truly encapsulate and store in a database field. Constant feedback and iterating on that feedback is how you arrive at ideal outcomes. The art of developing software or other high-quality user experiences is a very iterative, chicken-egg-style problem. It is a human problem. Being prepared to throw it all away and start over each time you iterate is one of the most powerful things you could ever hope to do. Just keep a copy of what you tried before as a stick to lean on. Each time you iterate, that stick should be getting stronger. Eventually you wind up with a bunch of sticks and you can start to chop them up, reorganize them, and build a proper structure.

mc32 6y ago

Also it misses the point. During the USSR’s and China’s attempts at grand projects often times they knew about mechanized machinery, but that solution wasn’t always applicable (too costly, diverted too many resources, or they had ample idle hands that weren’t easily deployed to something more productive).