How hard is python to learn? Doesn't seem that hard.

On contracting rates, Larry's rule of thumb is pretty good. A lot dpends on the situation you ate getting into. Also, getting involved with failing projects as a contractor can be extremely lucrative.

Yea good point there.

If you know you are going into a failing project and the situation is cool, then it can be great. If you are going into a place where the project is slowly failing and they think they need Skill X to get over a hump, but they are really clueless it is bad

Well, probably no harder than any other programming language? But I would guess like most languages, you aren't going to be really proficient for at least 6 months. Add on the fact that there's layers on top of that which it would be helpful to have experience with (django, mongoengine, django rest framework, django-auth, blah blah)

Like I said, if he'd aced the JS or been really good in some other way, I think we'd have hired him.

You can always do this.

I interviewed someone for a Project Manager position that had something on his resume about competing in local logic puzzle type meet-ups. So I got a bit excited because I like logic puzzles and did a bit of research to find some good ones.

When I mentioned it being on his resume and asked him a couple hard ones, he completely froze up and couldn't answer them. That lasted like 3-5 mins of him trying and not even knowing what math to do to start, etc. Then I asked him the easy one with 5 gallon and 3 gallon water buckets and getting exactly 4 gallons, and he got it in decent time.

That experience really put me off asking them.

So I strongly believe that good programmers do their problem solving at a terminal window using their programming langauge, so I think if anything, I'd ask someone to work on a programming exercise while I was there or maybe while I watched. I know that's more stressful than working alone but I think it's better than whiteboarding or talking about algorithms (which frankly, I am not that good at)

Are you allowing them to use Google/whatever they want as aids?

I guess I don't care but also we're talking about like 30-45 minutes to do whatever I come up with.

I'm thinking you should come up with a standard task you can every candidate and see if as you become an expert in how people answer the question it results in better accuracy in assessing them.

What, like, polishing my balls?

Can I just assign them a JIRA ticket and leave them alone in the interview room for a day or a week?

What's wrong with the questions in the interview thread? Apparently that matrix swirl was a hit.

Depends on what you mean by learn? To be useful and able to push stuff out, probably a week, but if you are going to be a Pythonista, where you are talking about OO, decorators, generators, Standard Lib, fully semantic pythonic code, Pandas, Numpy, etc, then probably a good 6 months to a year.

It's a layered language, meaning it is targeted towards beginners while still offering enough for people to learn something new all the time, so even after years, you realize you haven't fully "learned" Python.

For example, I learned how to add a dictionary to a dictionary generator about 2 weeks ago:

Generators are the fastest ways to build dictionaries, lists, and sets. I wouldn't expect someone who is relatively new to Python to know about all of this, but I wouldn't say that someone who did this didn't know Python:

daveT, serious question... how in the hell do you have such a hard time finding a decent job? I bet there are tens of thousands of engineers that are way worse at their job than you would be making 6 figures.

You know, after my devastating performance in my last interview, I've been really pondering this question. I may write an update to my progress thread, but I'm still working out where the honesty is without making it look like I'm all feeling sorry for myself.

Been spending the past 2 weeks drinking and writing, trying to work it out. I don't know the answer yet, unfortunately.

I think that's fine as long as it doesn't involve using any frameworks. Using frameworks will sometimes require a lot of time spent reading documentation on how to do X which would be nerve racking especially if it is something you haven't done. I would rather do something that shows my critical and logical thinking skills.

There's a start up in the Bay Area where their interview process consist of a hackerrank test and a 3 day hackathon where you work on a small project. Not three hours but 3 days! Ridiculous!

I am confident that I wouldn't have to give someone a coding challenge to see if they understood a software stack that I knew well. If the company is seeking someone that is capable of being productive as a software developer and is willing to tolerate a 6 month learning curve (which seems like a reasonable expectation for a learning curve) then that is a different matter. Seems like the company you work for is looking for a productive software developer and is willing to tolerate a learning curve.

As I wrote this I realize that I guess hiring for specific knowledge/skills is a much easier task than hiring someone who will "grow" into a role. Still I think tests are over emphasized. Which leads me to this.

This is a good post that captures the essence of why testing has its drawbacks.

It is ridiculous and I wouldn't put up with it.

I for example, just learned part of this from you. I've seen the {k: v for k, v in foo} dictionary generator before, although I didn't use it until this job because it didn't work on python 2.6, where I would usually have to use
method.

But I literally did not know you could stack dicts like that, i.e.
and now I'm glad I do because I use that **** all the time. It irritates me that dict.update doesn't return the updated dict so I'm always doing stuff like
Wait. I actually just tried this and it doesn't work. Is it python 3?

I am too. But the guy sat down and said he didn't know python, so now how do I decide if I should hire him based on his general programming strengths or not? I don't know php or node, which in this case was his backend experience.

Yeah, that's Python 3.5 and above, according to this link: http://stackoverflow.com/a/26853961

I would never bother interviewing someone who doesn't even pretend to say he or she knows the primary language we use for the position.

I don't think I have the balls to just stand up and say "well, good day then", after all, I didn't call him in there to interview, someone else did. Maybe that guy (my boss) thinks it's OK? Hell if I know.

lol'd

So by having a hackerrank question and then a 3 day hackathon assignment being the barrier to entry, you are essentially limiting yourself to hiring people who don't value their time and are desperate to get a job?

Seems terrible?

Is the 3 day project that different than 3 days of interviews with different members of different teams? We tend to focus on culture fit as much as technical fit so there are more interviews than hackathons, but it doesn't seem out of the question.

We've brought in a Windows programmer though we're a java shop and have had success with him, I wonder how good a programmer has to be before you get more benefit from having them learn a new language than having a mediocre programmer learn your product?

If a company wanted me to do a 3 day hackathon as the interview, I would need some sort of compensation for that time. Otherwise, hell to the no.