That basically is the same as saying "next time we will write correct code" in your postmortem, which I don't think is very useful. It's much more useful to say "our code is not structured in a way that makes testing easy" and "our smoke tests should cover the thing that broke." That gives you something actionable to work on that will actually prevent this from happening in the future. Otherwise, you'll end up writing essentially the same postmortem over and over again, each time saying "we will write correct code."

[–] BB_C@lemm.ee 0 points 1 year ago* (last edited 1 year ago)

False dichotomy much!

See this postmortem from Cloudflare as an example.

Under "What went wrong", point 1 and 3:

1. An engineer wrote a regular expression that could easily backtrack enormously.

3. The regular expression engine being used didn’t have complexity guarantees.

And on what needed to done, point 4

4. Switching to either the re2 or Rust regex engine which both have run-time guarantees.

See! Plenty of procedural talk in that postmortem. Plenty of corporate talk too. But you have to mention that a bad backtracking regex was used. And you have to mention that using regexes with no complexity guarantees was glaringly wrong. To not have done so would have been silly. To not even come close to mentioning those things beyond the specific error in that specific regex, and you wouldn't have been taken seriously.