Hacker Read top | best | new | newcomments | leaders | about | bookmarklet login
Ask HN: Should we waive technical interviews if candidate contributes? (b'') similar stories update story
6.0 points by varunagrawal | karma 439 | avg karma 7.08 2018-08-31 18:43:41+00:00 | hide | past | favorite | 16 comments

If a candidate contributes a code related patch to a company's open source project, should that serve as enough justification to the interviewer that the candidate has technical prowess in order to waive a technical interview?

This would IMO be a better gauge of technical quality over white-board interviews.



view as:

This might suffice to see if the candidate is able to write decent code, but what about their ability to communicate their ideas with other developers, their approach to tech in general, or how they'd mesh with the team on a technical level?

Those aren't part of the technical interview, possibly except for the "approach to tech" item depending on context.

How can "ability to communicate their ideas with other developers" not be part of the technical interview?

How could a developer get a 'hire' or 'strong hire' rating without communicating their ideas clearly to the interviewer?


What if a friend of the candidate is the actual person who contributed the code patch? You can't be too careful these days.

With such a level of cynicism, nobody would ever be hired.

People game the system because a system exists to be gamed.


Then they’d be excellent management material. Motivating friends to do random, high quality code patches. Very impressive.

Whiteboard interviews are as useless as baptizing a rattlesnake... but that doesn't mean contributed code is relevant indicator of future performance. It might be ... but as always "it just depends"

>... as useless as baptizing a rattlesnake

Says someone who clearly doesn't have to worry about the threat of demonic rattlesnakes.


You should waive technical interviews, full stop.

Most people have zero professional training on how to do them properly. So don't.

Using the technical interview approach, Einstein wouldn't get a job as a physicist because he had trouble doing simple math problems. That's not Einstein's fault. That's your fault.

Use methods of interviewing that are time-honored and aren't cargo cult. Conversational interviewing comes to mind.


You should waive unit testing, full stop.

Most people have zero professional training on how to do them properly. So don't.

/s


The question is if technical interviews of the white-board variety usually have any value at all in determining someone's technical prowess or if they're generally just a huge waste of resources.

Quite likely, there's no one-size-fits-all solution for assessing candidates.

The way I approach this is by looking at the candidate's background and trying to make connections to the product the company or team is working on.

Then I have an open conversation with the candidate about these connecting points and possibly related subjects. It usually becomes obvious quite quickly if the candidate knows what he or she is talking about. It also gives you an idea about how the candidate communicates, which often is more important than technical skill.

Technically, that's still a technical interview but it's much more free-form than the rigid white-board interview format. It's more of a conversation between equals rather than a test setting, where one side is being questioned.


I like this way of doing things. Some people are brilliant and very good at general communication but just don't interview well due to various factors. This sort of free form conversation should put them at ease as well.

I feel like what you're asking is borderline to a dumb question. Because yes, it is obvious that someone who can contribute a (meaningful) patch to working code is qualified. At least skip the FizzBuzz tests.

We don't see this happening for tech jobs, and at the worst, interviewers will still ask programming language specific questions. Plus "meaningful patch" might mean different things to different people.

Yes, this is a far better measure of ability than any white board test.

Then if you want to measure communication skills etc then just look over some of his commits and get him to explain them to you.


Legal | privacy