Hacker News Reader

Comment by za3faran | Hacker News Reader

How does it address it? By making it painstakingly verbose (not to mention error prone) to deal with errors?

Not referring to you personally, but I've heard that sentiment several times now, and I have not seen anything to back it up (as with several other golang claims).

View on HN

given a function fn that can fail, it will return a result and an error e.g.

    result, error = fn(...)

calling this function should yield to the caller two possibilities, somehow: a success value _or_ a failure error

the important thing is that in both cases, the control flow is visible in the source code as written

    result, error = fn(...)
    if there was an error, ...
    if it was successful, ...

when an expression fails, you want to see the consequence in-line

the success path and the failure path are equally important

Yeah, but you're still relying on the programmer to do it correctly which is nothing but false hope.

no, it isn't

Nothing is preventing the code from returning both at the same time. I've seen code (including in the standard library) that returns both an error and a return value. In a language with disjoint unions, such cases would be encoded properly.

convention prevents it

and in the case where returning both is OK, then documentation makes that clear

this is not difficult

Convention in no way prevents anything. Convention is simply that. People are free to not follow convention when nothing is enforcing it. You frequently see juniors, who may be brand new to the language, making mistakes with conventions. If I'm supposed to depend on the vagaries of some accepted standard that is only documented in text then it is less than useless in the real world

if we say a language "addresses" a given concern, is it necessary that this is accomplished in the compiler, and that the rules for that concern, whatever they are, are enforced at compile-time?

(spoiler: no)

spoiler: yes

If the language "addresses" it by convention then it is not addressing it at a language level at all

"the language level" is not only what is defined and enforced by the compiler

but i'm sure i won't convince you of anything here, so good luck to you

We all know that documentation always 100% matches the behavior of the code :-)

with robust and potentially high volume code, the most important feature is good behavior in failure domains. disk full, do you abort or continue once the cron job frees some space; cant alloc memory, do you abort or return a static 503 page? bad contents in some file, do you exit or log the error and carry on? does a bad pyc file generate a good error message or crash python. this robustness is the famed second 90% of the project. normal go looks a cerain way when it is handling these errors.

Nothing you wrote is specific to golang's error handling though, and in actuality, ends up being more brittle because it is possible to miss handling such errors. At least an exception would bubble up instead of keeping the program running in an undefined state.

it is my very clear experience that rust programs are more brittle than go programs, precisely because rust makes it possible (even encourages) error "bubbling" via `?`

in practice, go code bases that are subject to even minimal code review have basically no ignored errors

yeah my linters won't allow it. and bubbling up is exactly what you dont want for robust code. the UX is different for a failure on rereading a config file when you already have a good config vs. on startup.

go code tends to be robust because the authors of the code and the community are the sort that worry about each err value, like Linux and like Python and like C itself.

Reply on news.ycombinator.com