Hacker News Reader

Comment by dnautics | Hacker News Reader

Nope. Buggy declarative code is harder to debug than buggy imperative code.

I think the core thing is that if you have a bug in the declarative code, it usually affects everything and can be fixed in the execution engine, or you didn't actually declare what you wanted but something else.

By contrast, bugs in imperative code can manifest simply because you happened to write it in such a way that relied on some assumption that no longer holds and your imperative code does something it should not do because its preconditions are invalid.

Declarative approaches are generally more robust because they by necessity tend to be designed to handle more diverse initial states and still reach the correct result.

You still are implicitly assuming it’s easier to write bug-free imperative configuration compared to declarative. It doesn’t really matter which is harder to debug if it’s much easier to write bug-free code using declarative code. Again, some specific examples that underpin your assumption would be helpful in making your argument more convincing.

Read beefwellingtons SQL EXPLAIN PLAN, comment. I've also watched my roommate have a four hour argument with her coworkers about one line in their kubernetes deployment[0]. I've also spent days trying to fix a declarative Ansible playbook, only to table flip and rewrite it from scratch in 20 minutes (this was a MySQL install-and-configure command where MySQL version had drifted underfoot).

[0] iirc this was not a bug in their kubernetes operator (or maybe it was, I don't recall exactly) but a situation where they didn't understand why their containers were exhibiting the anti affinity properties they were expecting out of their declarations.

edit: were not exhibiting the anti-affinity. They were trying to make sure distributed postgres replica nodes weren't on the same physical hardware, for obvious reasons.

Reply on news.ycombinator.com