There is no documentation

On Tue, 04 Apr 2084, by @lucasdicioccio, 111 words, 0 code snippets, 0 links, 0images.

This article is still considered unfinished and content may change significantly.

Engineers like to throw dirt on others.

The problem is that it is difficult to know upfront what is worth documenting.

Edge cases become know after the fact.

Engineers are rather rationale and well-intended, they budget their time.

So what to do better?

Probably use more formal methods when specifying.

I recently worked on lead-time computations (e.g., a pull-request process).

Tons of thing need specificaion

  • what timing do we attribute

Tons of thing can go wrong:

  • un-ordered event: an issue

Each of these questions need an answer. Unfortunately each of these answers need some analysis and have some tradeoffs. Even more complicated: there can be some non-composable global tradeoff (i.e., all questions need to be answered as a whole for consistency reasons or not).