4 ms·
Yep, we build a foundation with plywood and styrofoam and then someone comes and asks you to build a skyscraper on top of it.
by brigadier132 3y ago
Yep, we build a foundation with plywood and styrofoam and then someone comes and asks you to build a skyscraper on top of it.
- trwaw 3y agoWe build a train and then someone asks to turn it into a skyscraper because skyscrapers are the in thing now. It already looks like one on its side, how hard can it be to turn it right side up?
- smaudet 3y agoWhile I do think proof constructed code could be superior, I am not convinced for two reasons: 1. The problem expressed here has to do with a particular, well understood problem. It is not subject to the whims or constraint needs of the "customer". Proof construction is great when the problem is properly understood, I don't know how well it handles constantly changing requirements. (This encompasses business and interaction constraints too - perhaps you can construct a provable correct UI, but will it be engaging or usable?) 2. Test driven development can guide high quality code construction, arguably at lower construction cost. That is the reason I'd argue this has not taken off in comparison to traditional "test" based design. You can argue it produces lower quality code, and you'd probably be right. But in comparison to construction of complex, rich design, I suspect that it produces faster results at lower cost. The provable solution will win out...eventually. Depends on whether it can make significant marker penetration once it is complete. Nevertheless, yes of course we should be building foundations witb concrete not plywood and Styrofoam, so this is a good development. I do think a proper engineering approach with test based verification can build a "somewhat concrete" foundation, at least.