[Jenk]

Also not OP, and also similar experience. Most of my interviews are an hour or two of conversation, sometimes with a "I've been told I have to ask.. " techy question but never involves any actual code/whiteboard. Several times the interviewer has said something to the effect of "I won't insult you by asking you to do a tech test, it's clear to me that you understand the work" or similar. One occassion (some years ago now) I've replied to a request to complete a tech test with a "Would you rather look at my open source code/contributions?" which they did and I didn't need to perform the test.

[ARandomerDude]

Not quite as experienced as you (I’m in the 10-15 year bracket).

I never ask LeetCode questions because it’s far easier on both sides to ask experience-based questions like “what do you like/dislike about <technology on your resume>?”

If I get superficial answers that tells me something. If i get well thought-out answers from people who have clearly spent time in the trenches that tells me something too.

[CWuestefeld]

I've been doing this for 30+ years, and I've found the same. It's not just whether the candidate can answer correctly, but how they answer.

Back when I was doing C/C++ interviews, I'd ask questions like:

What's the difference between single and double quotes? A meh candidate answers something about you need single quotes for just a single character. A good candidate answers about how the data type for a double-quoted constant is a char.

If I want to pass a variable into a function, and have its value changed by the function, what do I do?* A lousy candidate says "put an ampersand in front of it". A meh candidate says to pass a pointer to the variable. The best candidate will talk about the difference between call-by-value versus call-by-reference.

People really do reveal a lot about themselves not just in what they say, but the way they say it.

[MaxBarraclough]

> If I want to pass a variable into a function, and have its value changed by the function, what do I do? A lousy candidate says "put an ampersand in front of it". A meh candidate says to pass a pointer to the variable. The best candidate will talk about the difference between call-by-value versus call-by-reference.

I agree it's good for a candidate to demonstrate a solid understanding of evaluation strategies, but the way you phrased the question, the 'meh' answer seems about right. You asked what do I do? which invites a narrow answer specific to the language.

[Sohcahtoa82]

Agreed.

In fact, I had an instructor in college that would have marked you down if a quiz/test asked this question and you wrote an entire paragraph describing call-by-value versus call-by-reference.

On the final, he said "Each of these questions is answerable in 1-3 sentences. If you're writing 1-3 paragraphs, you're wasting my time and I will subtract points even if your answer is correct."

[ryandrake]

We need your professor to coach interview candidates. On the "interviewer" side of the table, I can't tell you the number of times I've asked a question that should produce a quick, simple answer, and instead got a 5-10 minute stream-of-consciousness word salad out of the candidate. Or a huge run-on sentence this-and-that-and-this-and-also-that-oh-and-this... without coming up for air. It's so common, I must assume that these kinds of replies are being taught to candidates as some sort of best practice.

[MaxBarraclough]

I suspect many interviewers reward using a question as an opportunity to showcase knowledge though, even if it means rewarding that kind of expansive verbosity. I suspect the this and also that you allude to is the candidate not wanting to be called out for failing to mention a corner case or tradeoff. Of course, conversations are rarely legalistically constrained to answering only the precise letter of a question asked; it's a matter of degree.

Forgive a painful mixing of metaphors: if the Workplace StackExchange is anything to go by, one interviewer's flying colours are often another interviewer's red flags.

[SoftTalker]

I think it's more likely a nervous candidate who may not be comfortable or experienced in conversation, as many introverted tech type people seem to be. I don't think it's being taught; it's more likely the result of the candidate not having had interview coaching at all.

[Sohcahtoa82]

> On the "interviewer" side of the table, I can't tell you the number of times I've asked a question that should produce a quick, simple answer, and instead got a 5-10 minute stream-of-consciousness word salad out of the candidate.

Yup.

There's a fine line between giving a thorough answer and just vomiting up everything you know that's slightly relevant to the original question.

I do AppSec. If I'm interviewing a candidate, and I ask them what Cross-site Scripting is, then if at some point during their answer they bring up SQL Injection, that's a red flag.

[roxgib]

This is what I've been told. The idea being that the questions are more like prompts, and you should go ahead and elaborate to show your knowledge? Of course sometimes a questions really is just a question.

[cardinalfang]

Must have been before rvalue references.

[jenscow]

Both great questions, where the answer will give you a level of knowledge, rather than a yes/no.

(personally, I would be a meh candidate until you probed further - but I've not touched C for.. hmm, I'm old)

I often get questions where I need to determine if they're looking the textbook answer, or a real-life answer. Usually I will go with "well generally, the answer is $Textbook. However, ..."

[CWuestefeld]

I just realized that my above model answer gets broken by the formatter. the first question should get answered with char-star, that is, pointer to char.

[Izkata]

If you put a backslash in front of the asterisk it should be kept as-is instead of triggering the formatting.

[ARandomerDude]

This is why I don't use C. ;)

[EmilyHATFIELD]

Why would using pointers be better than references ? At least a reference cannot be null?

[CWuestefeld]

Well, at the time I was doing that, it was actually for a C and not C++ job, so &references weren't even a thing. But my point wasn't that using the ampersand is wrong - just that if that's how you describe what you're doing, you probably don't have a very strong foundation in how the language really works.

[anonymouse008]

Completely off topic, but I wanted to see if your bio could be parsed for an email with chatGPT - since I thought yours was the most interesting I've seen in some time. Turns out cgpt is quite clever:

What's the email address: To email, send to the domain [redacted], using the mailbox "[redacted]".

In general, email addresses consist of a username followed by the at symbol (@) and the domain name, such as [redacted]@[redacted]. However, without being able to verify the existence of the domain [redacted], I cannot say for certain what the email address would be.

[anonymouse008]

Hey someone flag the above comment - I’m an idiot and left the email in there

[dang]

We've redacted it now.

[CWuestefeld]

as the person you're talking about - I just up-voted you anyway, just because this is interesting :)

[kubami]

Exactly this! I haven't been interviewed for a long time now, but since the beginning of my career, when I have interviewed people I ask deeper questions. Even with tech tests, I use them as a jumping board for diving deeper into language design, tradeoffs and experiences. Even a trivial, reverse-a-string kind of questions can take you deep into candidate's knowledge.

[Jenk]

Yes, I'm exactly the same.