I really think it's great. For this project, like I said, I had to build installer packages. I actually built deb, rpm and a windows msi. The build "system" is just a dockerfile and a small bash script. Anyone, anywhere in our org, can just say
docker run osquery-build-windows
or
docker run osquery-build-linux
and it'll work perfectly. They don't have to have anything installed on their machine except docker.

Likewise, to test the installer packages, there is a dockerfile/script that goes through the motions of adding our repo to your system, installing the package, running it, and uninstalling it. Same - there are zero dependencies on your system except docker.

Adding these to build system like jenkins or bamboo is trivial, because again, no dependencies. It's a one liner that will take me a minute or two at most.

A really nice thing about it, also, is that whenever you stop and restart a docker container, it reverts to it's base image. Anything you did in the last run is gone, unless you explicitly save it outside the container. This means that I could test installing and uninstalling a package hundreds of times without accidentally contaminating my system. You could do this with a normal VM and a checkpoint, and with some VMs you could probably even automate it, but this is a lot easier, and when I'm done, I don't have to ship someone my VM image for them to be able to do it.

It's not complicated either. There are a few concepts that take a little while to get comfortable with, but after that, making new dockerfiles or docker-compose files is quick and easy.

It's almost certainly using we sockets which will allow two way commination over a persistent connection. I don't think these requests or events show up in dev tools or whatever. You could try running something like Wireshark to verify

Yeah docker feels like plumbing to me. Problems that will be completely solved by a few turnkey services or maybe standalone but still like AWS. So all that knowledge will be useless in a few years unless you work for amazon or the turnkey standalone vendor.

I've never debugged sockets before, but Chrome DevTools has a tab on its Network thing for websockets, so I'm guessing they'd show up?

I've sort of worked around my problem anyway, so don't need to worry about it anymore.

Docker used be virtualized on mac, but now it's native.

I'm like 90% sure it uses hyperkit. I don't know if that counts as "native" or not.

It uses osx hypervisor, so I think that is considered native. Not sure if that's what you're referring to as hyperkit? In the past I've used a version of docker for mac that integrated with virtualbox. That was slow and crappy.

It turns out that hyperkit is a hypervisor framework, so I guess we're both right.

Web sockets + ui/backend code integrated (RoR for example).

job interview this afternoon.

I looked up the company, they have about ~20 crappy mobile arcade style games for android OS. I am familiar with developing with android and am going to mention a game I've been developing on the side.

I'm going to ask about a few of the games they have in their portfolio and bring up some issues I saw people complaining about in their reviews and mention some ways I think I could help those. Good idea or no? It's an "informal" interview so I imagine the conversation is going to be more two-way than I am used to and am looking for talking points.

I'd say no. While that sort of thing can work, I think it's better left to people that have a good bit of industry experience (even then I'd probably caution against it unless you're going for mgmt or product role). You, having minimal experience, are likely to offend even if you have great points.

Holy damn I was blindsided by that interview.

He led me to believe I'd be talking to him (the owner) so I expected a non technical interview followed by possibly talking to his head dev.

Right off the bat it started off with a whiteboard interview with a pretty grumpy guy that seemed to have a ****load of industry experience.

The first problem was given a crossword, and a n sized "slot" where a word could go, he wanted me to find all possible words that could fit in that slot.

I got the base case very quickly, when the slot is empty. Then he asked what would happen if there were already letters in there from other words, and I realized I had to change my function to look at a substring and compare the strings vs each word in the dictionary.

I got the logic fine but I was extremely nervous and messed up one of my if clauses and wrote '&&' when it pretty obviously should've been an ||, and it was pretty embarrassing. He had to point it out to me, and kinda chastised me because he (correctly) said that's a really hard bug to find and the cause of a lot of grief in software.

I did catch a few things he pointed out, like he said I had a memory leak somewhere and I found it right away.

Then the second problem he gave me was a math/logic puzzle where given two lightbulbs and a n-story building, he wanted me to optimize the number of drops required to find a safe story to drop the bulb from without breaking.

I got the right answer in like 2 seconds which I felt he was impressed by. Then he wanted me to write a formula and optimize it, and I couldn't really generalize it so he had to help me. But, when he asked me how to find the minimum of an equation, I just blurted out "find the derivative and set it to 0" which I could tell he was really impressed by but honestly I just guessed. He said most interviewees cannot answer that question. He expected me to do a lot worse because I told him I'd never taken physics and he didn't like that, but when he asked me why it's often important in game physics that the final answer to this type of problem is a square root I was able to tell him because acceleration due to gravity is meters per second squared, and I could tell he liked that too.

He had me finish writing out the problem and went into the other room to speak to the owner and came right back and said the owner wanted to see me. So, I think I might've done well enough, but it felt like a trainwreck to me and I was very flustered. My glasses broke mid-problem too, lol. Just fell right off my face. I think I played it off ok.

The owner asked me how the whiteboard interview went and I told him very honestly that it was my first one and I was extremely nervous, I made an embarrassing mistake but the head software guy was very nice and patient.

All in all could've gone better could've gone worse. The owner is a silicon valley veteran and they are trying to corner the market on "niche" arcade games designed for flip phones and they are trying to transfer those into the android marketplace. He liked that I am willing to work part time and lived close and am really flexible. He said he'd let me know.

I'm going to practice the **** out of hackerrank puzzles and maybe get an interviewing book, I never want to look that stupid again.

String puzzles are my absolute kryptonite too. ****. since my earliest days of learning to program I have hated strings with a passion. Probably something I should practice.

I know this is from long ago, but this situation arose again today and I just then realized forget address it.

say you make like 40 commits on your local branch. then you go to rebase that branch from orgin or master or develop or whatever you call it. for every single commit you will need to fix the merge conflicts. so you will end up making the same fixes up to 40 times (in this situation.)

Sounds to me like you did pretty well on a very tough interview, esp for a part time role. I wouldnt sweat it, everyone is nervous in those.

I got the lightbulb q in my first interview too, def did not do so hot on it but luckily the dude held my hand through it. I didnt have to actually code it though, just explain the solution

Sign up for leetcode and solve as many problems as you can. This will help you land a good job way more than side projects.

wtf @ that interview. Pretty sure you really don't want to work there.

I'm lazy and haven't been pursuing opportunities. It's close, I have no plans for the next year, it's small and in the area I want to end up in (games) so it seems like a good place to strive for right now.

Basically I need experience, confidence, and something to write on a resume so I don't really care how crappy the company or team is.

That **** is industry standard unfortunately, and I hate it. I interviewed for a lot of jobs this time around because I already had a job, I just didn't like it, so I was taking my time trying to find something decent. Every interview is like the toughest day of work I ever had, and most of them are nearly a full day, like 6 hours at a minimum.

I like it best if they let me code on a computer, less if I have to whiteboard, and the least if I have to explain verbally. Go figure.

I haaaaaate these problems like the drop the light bulb thing. At least the "how many ping pong balls can you fit in a 747" or "how many piano tuners are there in canada" have gone out of style. And if anyone asks me why man hole covers are round or about light switches and bulbs in other rooms, I am getting up and walking out. Someone once asked me that god damn midget in an elevator question.

I once got accused of cheating in an interview, i.e. I guess looking up the interview questions on glassdoor or something. It was just something I'd spent a lot of time thinking about myself for other reasons.

From your description I think you did well on the interview. Some interviewers are *******s and want only 100% correct instant answers but for the most part people are asking questions that are meant to be hard, to make you stretch, so they don't expect you to just nail it. Esp if this is pretty early in your career, which, it sounds like it.

Working in games is fine, a lot of my friends did it, but a word of caution: everyone wants to work in games, which tends to mean that if you don't want to do something there is someone standing behind you ready to do it. This means that industry standard is overwork and underpay, not to mention rush changes, crunch times around release that are unbelievable, etc. And you should get used to the idea that you might spend a year on a game that never sees sunlight.

Most people I know who did it are glad they did, and glad they don't any more. It's a young man's game.

+1

This is standard now a days. I imagine gaming industry would put more emphasis on this kind of stuff.

Plus solving those type problems will help you write short and clean code in my opinion. Once you get in a good habit of it, the hardest part will just be thinking about the algorithm of how to solve the problem.

Side projects are great, but only if you can get feedback. If it's open source projects where you can get your code reviewed by great programmers, then that's awesome. But not many people have that amount of time to invest.

For work I had to read some bad PHP code and re-implement it in Scala. Turns out the original developer was trying to guesstimate a solution to an equation in a horrific and ugly way. I replaced it first with binary search and later improved it by using Newton's method. So yea, I think leetcode/hackerrank did help me in recognizing what I needed to do.

Yea I was waiting for others to chime in but that's one of the most ****ed up interview experiences I have heard of.

If I was in jmakin's shoes I 100% would have let slip a "No wonder people complain about your crappy games having issues".

Why was the interview messed up?

Curious about that as well. Didn't really sound bad to me at all, except for the interviewer being a bit grumpy.

I also don’t get why the interview was particularly bad.

Interviewing sucks most places. I’m not even sure there’s that strong of a correlation between a ****ty interview and a ****ty job at this point.