| ▲ | A sufficiently comprehensive spec is not (necessarily) code(buttondown.com) | |
| 6 points by BerislavLopac 2 days ago | 3 comments | ||
| ▲ | uptodatenews a day ago | parent | next [-] | |
This is exactly why I built these two things https://github.com/RCSnyder/tlaplus-process-studio This is to extract the bare minimum state machine out of users is tla+ https://github.com/RCSnyder/lights-out-swe And pop that into input/docs/* with any other user requirements And add your tech stack in preferences.md And say "build me the thing described in input" And it goes for it | ||
| ▲ | squirrellous a day ago | parent | prev | next [-] | |
I’ve always thought about this as more of a meme than a serious point. Trivially, aren’t most of the network protocol RFCs “sufficiently comprehensive spec that isn’t code”? | ||
| ▲ | bediger4000 2 days ago | parent | prev [-] | |
Wayne is usually lots better than this. He basically defined the problem away, while missing the point of the comic. Maybe defined specifically "code" is not the tightly worded spec, but almost universally "code" is the result of the process of clarifying the vaguely worded spec. That's the point of the comic, after all. We all recognize the shorthand of "code" in the comic for the process we've all been through so many times. | ||