Fork me on GitHub
#jobs-discuss
<
2020-07-16
>
sveri07:07:51

> because those programmers aren’t good at writing on a whiteboard and explaining their work out loud while coding I have been doing live coding over the last years again and again and simply talking about what I am doing is super hard up to impossible, simply because my mind is busy thinking about a problem and how to solve it and not explaining it. One would have to be prepared for that specific problem, solved it already or have a script to say something useful.

practicalli-john15:07:09

I concur that thinking and writing/typing can be very difficult and have over 80 hours of YouTube video if anyone needs evidence 🙂 Sometimes you can get into the flow and the right information just comes out of your fingertips, but that's very much the exception and I dont believe I've ever had that during an interview. I do enjoy white boarding in interviews when its big picture things, overall architecture and design. Coming up with detailed solutions is very flawed and I have failed because of adding too much or two little redundancy at the whim of different interviewers. Overall I enjoy a whiteboard session more than doing a coding test, especially when I have several people in front me me allegedly pairing with me but actually just stilling their and not interacting. This seems to be what interviewers think pair programming is in some companies.

sveri16:07:01

I remember, back when http://livecoding.tv or something similar was a thing and accessible to everyone, I actually had a few recurring viewers and indeed enjoyed talking about programming and the projects I was working on, but yea, in that time I did not really finish anything. On twitch I find it kind of harder to engage with the public. Do you have live viewers on youtube that you talk to?

practicalli-john18:07:43

I don't think YouTube does anything to assist or encourage participation. It's used mainly for reach of audience and it was connected with Google hangouts which I initially used. I now use OBS to stream to YouTube, so it would be easier to use other services, but haven't seen anything that would make a switch worthwhile. My live broadcasts were trying to be a study group with lots of interaction, but I think it's not an easy medium for many to actively engage in. I still find it uncomfortable after doing this for years. I do get feedback in the chat, although there is a delay in the broadcast of about 20 seconds, (for caching I think), so that doesn't help. The best way to get interaction is of course to have several people as part of the broadcast itself, like apropo cast with 4 people and other podcasts where there are several people as part of the conversation.

Neil Ashton13:07:28

@jr0cket: thanks for adding the links to those FP-oriented job sites. Are there other such sites that didn’t make that shortlist, or is that two-item list pretty much exhaustive?

practicalli-john14:07:11

I added functional works as I know they specialise in functional programming jobs and have had a good experience with them. Also their website is written in ClojureScript. There may be other companies that specialise and often general recruitment companies have functional programming roles too.

practicalli-john15:07:09

I concur that thinking and writing/typing can be very difficult and have over 80 hours of YouTube video if anyone needs evidence 🙂 Sometimes you can get into the flow and the right information just comes out of your fingertips, but that's very much the exception and I dont believe I've ever had that during an interview. I do enjoy white boarding in interviews when its big picture things, overall architecture and design. Coming up with detailed solutions is very flawed and I have failed because of adding too much or two little redundancy at the whim of different interviewers. Overall I enjoy a whiteboard session more than doing a coding test, especially when I have several people in front me me allegedly pairing with me but actually just stilling their and not interacting. This seems to be what interviewers think pair programming is in some companies.

Aron15:07:38

My former colleagues who have several hundreds of interview experience said that this is all moot somewhat because the point is not to examinate how well someone can solve some random problem, but how we could work with them, how they validate their decisions, how they attempt to solve problems. That is impossible to judge without interaction.

danieroux15:07:43

David Hussman (RIP) said that you once you knew that someone could play their instrument - you had to see how well they played in your band.

scriptor16:07:22

Those are the reasons that are always given for those sorts of interviews. My opinion is that 40 minutes in a high pressure environment is absolutely not the right way to see how someone solves a problem

👍 3
scriptor16:07:10

And as much as people talk about seeing how people think, they'll still factor in how much of the exercise was completed

scriptor16:07:50

It'd be much more useful to present people with problems with multiple straightforward solutions and see how they weigh pros and cons of each. If you want to do a pairing exercise at all

scriptor16:07:14

To me that's more useful than seeing how someone solves a leetcode greedy algorithm problem

seancorfield16:07:17

@scriptor This is why I don't give "technical test" interviews because they're never a measure of a candidate's real ability. I like a 30-60 minute discussion, no coding, no computers, no whiteboard. I've posted the mind map diagram I use to guide that discussion several times and it is pinned in this channel I believe (see channel information (i) for pinned items).

💯 3
kenj18:07:08

I performed my first interview last week using the mind map as a starting point and it went well. I can definitely recommend it!

18
robertfw21:07:14

I also used it recently and it was a very useful tool, thank you for sharing it

3
scriptor16:07:52

Yes! I definitely agree more with your style

scriptor16:07:53

As someone who recently went through the whole interview ordeal I've noticed that my success with pairing exercises corresponds almost directly with the seniority of the interviewer

scriptor16:07:55

People with more experience at interviewing others seem to do better at knowing how to break down a problem into multiple stages

scriptor16:07:44

Whereas others tended to mostly go "here's an empty screen. Please write a four-in-a-row game"

Aron17:07:29

Oh yeah, well, anyone pressuring interviewees are sending the clear message that hey, if you come work here with us, we can get you pressurized all day for you to get your salary and feel accomplished. I wouldn't want to work in such an environment.

Aron17:07:58

For entry level jobs where you get people with zero experience, you have to filter out those who literally didn't even try to learn at least a couple of lines of code. And I have seen enough of these to know that technical questions have a place. Just very minor.

Aron17:07:06

I hope no one is judging my former colleagues unfairly, I can vouch for them. Not for the whole company sadly. We had a top manager who was so obsessed with the hiring process that he wrote a software with hundreds of technical questions in a big decision tree depending of former experience. It was extremely bad. He refused to not use it 😞

Aron17:07:13

reminded me of the Peter principle

seancorfield17:07:19

Thirty years ago I worked in the IT department of an actuarial firm. The manager insisted on implemented a standardized "technical test" for screening all new hires. I suggested that all existing IT staff take any such test to get a baseline of data and scores so we could set (management's) expectations about scores for potential new hires. As I expected, nearly all of the current staff "failed" the technical test...

seancorfield17:07:10

Only the team leads passed -- and they mostly had been promoted to roles where they no longer did programming on a day to day basis.

😆 30
sveri17:07:31

Would be ironic if the promoted team leads actually would fail a similar test in their new role. Would be the Peter-Principle in action.

scriptor18:07:48

I wonder what it'd be like if candidates were also allowed to test future colleagues. Like, if the candidate could pick a problem and have someone they would work closely with try to solve it

scriptor18:07:34

It seems only fair since I'd also want to know how my future colleagues solved problems

scriptor18:07:58

I don't think any company would ever go through with that though

scriptor18:07:14

Actually, picking a problem that's unfamiliar to both the interviewer and the candidate might lead to interesting results and better matches real world conditions anyway