What happens when a destructor throws (sandordargo.com)

by jandeboevrie 21 comments 24 points
Read article View on HN

21 comments

[−] NooneAtAll3 39d ago
I... I did not expect that noexcept(false) works on destructors
[−] deterministic 37d ago
Asking that kind of minutia detailed question in an interview is seriously bad practice.

It is easy to come up with questions that the author would fail in an interview. Simply because there are thousands of minutia details that most senior developers wouldn't know about.

What the interviewer should focus on is the ability to solve real world problems. For example by walking the interviewee through a real world business scenario and letting the interviewee describe how he/she would solve the problem.

[−] avalys 39d ago
Don’t know who this guy is, but I’m glad I never interviewed with him. This is language-version-specific behavioral minutiae that anyone can look up in 5 minutes in the rare case it matters, and is otherwise irrelevant to engineering software at a senior level.

This article is a junior engineer’s idea of what a senior engineer should know.

[−] p1necone 39d ago
This feels like an overly negative comment. language specific minutiae is interesting to a lot of developers, and this kind of stuff is exactly what you'd ask if someone claimed to be an experienced C++ developer. You're not going to decide not to hire them based on them not knowing this specific thing, but if you ask them 5 different questions about specific behaviour/edge cases/whatever and they don't know any of them it's probably a bad sign.

(Although "this is bad practice, I've never done it, I didn't care to look up details" would be a perfectly fine answer to me if I was the interviewer)

[−] menaerus 38d ago
I am an experienced C++ developer, I know what happens in this particular case, but this type of minutiae are only interesting to the developers who have never had an actually hard problem to solve so it's a red flag to me as well. 10 years ago I would have thought differently but today I do not. High performance teams do not care about this stuff.
[−] BobbyTables2 38d ago
Damn. Just the new C++ syntax for this stuff makes it seem like a foreign language.

It’s no longer the C++ from 20 years ago.

Raising exceptions in a destructor sounds even more fun than a “return” statement inside a Python “finally” block of a method.

The footgun store will never go out of business!!!

[−] zarzavat 38d ago
It's as close as you can get to useless knowledge. It's like asking a pilot "exactly how will the aircraft break apart if you nose dive it at high speed into the ground?"
[−] throwaway27448 38d ago
This analogy makes a lot of sense until you need to deal with an exception emanating from a destructor... then it looks a lot more like "what's the proper way to hold a chainsaw"
[−] bsder 38d ago

> This is language-version-specific behavioral minutiae that anyone can look up in 5 minutes in the rare case it matters, and is otherwise irrelevant to engineering software at a senior level.

The fact that C++ programming books have entire sections about destructors (see: Effective C++) shows that this is very much not irrelevant minutiae. C++ forces you to deal with this kind of detail all the time.

Now, we can have a much more interesting discussion about whether C++ is a disaster of a language precisely because you are forced to deal with this kind of minutiae by hand. We could also have an interesting discussion about whether RAII is the "object oriented" of our time. We could even have an interesting discussion as to why so many companies ban constructors/destructors in their C++ programming guidelines.

However, irrelevant minutiae C++ destructors are not.

[−] 112233 38d ago
If you noticed, the article did not actually tell you what happens when destructor throws. It was only about double-exception case and throwing in nothrow() function (both perfectly valid things to know when jobbing).

What state are members left in when destructor throws? If exception happens in virtual base? If member destructor throws, what other class members have they destructor executed? will delete[] be called?

The author possibly does not care to know themselves! As you say, totally irrelevant to any normal programming. Unless you are writing clang or gdb

[−] rhoopr 39d ago
“The camera is the least important element in photography.”
[−] nvch 39d ago
Well, now those who will go to look it up in 5 minutes may end up reading this guy’s article.
[−] zarzavat 38d ago
There's nothing essential to know about what happens when a destructor throws other than it's bad and don't do it. Especially if the effect is to terminate.

I hate this style of interviewing which is more about the interviewer feeling good about themselves rather than doing their job and assessing candidates. Just wastes everyone's time all around.

[−] supliminal 39d ago
No mention of vtables or thunks? We decided to proceed with another candidate.