Whippets are much less potentially damaging than glade air freshener. Dentists use nitrous.

Anais,

After what everyone else said, that was definitely a simplistic and I think unnecessary addition to the situation wrt Fubster. I think if he was looking for a generic programming task to be accomplished he would have no problem paying for it at market and finding someone quickly to get him a solution.

The issue here is the people he is targeting are a specialized subset of people who are valued incredibly highly in the same geographic location by large financial and other well funded organizations. The feedback of posting 150-300k job listings for the same skills is a lot different and helpful to Fubster than comparing him to a band looking for free artwork.

This probably seems like nittery, but I think in general the whole "asking people do to things for free" can get lumped into the "trying to hire a specific programmer to accomplish a result, but don't know how exactly to do it" and in the second case it isn't always easy or possible for people to know how to go about it.

yes, i am providing the algorithms/analytical approach. i guess what this entire exchange has taught me is that i'm better off just hiring a contractor to do it, and i can take care of the details on my own.

not anymore they aren't but i get what you're saying. i actually built an event driven low latency trading platform (in c++! gasp) based on equity that ended up just sitting around collecting dust. to be fair, it wasn't nearly as complex as it sounds, it was basically just something that took in a datastream and then rifled off FIX orders based on user input before a news event. most of the work was already done for me by the stream and platform providers. honestly, my biggest hurdle in this whole thing is knowing where to start. i have such little experience working on big projects without some sort of hand-holding that i would 100% for sure **** it all up right from the beginning and have to start from scratch like five times before i got something that wasn't terrible.

no, i was being honest with my motivations when i explained them before. nice that you were able to take time from your already busy schedule to call me a liar though, that's very helpful.

just to clarify, the reason i am going about it this way is because someone who made billions+ dollars doing this exact thing (except with horse racing) told me to "find a partner you can work with," and since everyone in my professional network who is capable of doing this is already doing it, i'm sort of grasping at straws. obviously i didn't realize that there was already a huge amount of dumb startup ideas from people who pitch equity options instead of payment.

also, blackize et al, why do you think that C++ would be a bad choice for something like this, or at least the core "game player" component anyway? is C++ really that much of a tricky specialization? it's been a long time since i've been involved in any big project, but last i heard, C++ was the go-to for things where runtime is important, and if you're going to simulate a million games in a reasonable amount of time, i feel like runtime is going to be important. is python just that sick these days?

also i did whippets in high school fwiw

I'm no C++ expert, and I didn't mean to imply that it was a bad choice, just that it wasn't a clear best choice. Since you're building from scratch you've got the option at looking at alternative language choices which opens up the pool of developers you can work with.

I mean you'll likely see a performance drop if you decide to use java, but there are more java devs out there. Maybe that balances out the speed advantage.

I'm not certain that pypy will fit your needs but it will let you write python and approach C speeds.

Java makes threading considerably easier than C++, do perhaps it could be a better choice.

Anyway, wasn't trying to knock C++ I just wanted to suggest that there are advantages to other languages. You can see huge gains in speed of finding a dev and speed of development while taking only a small loss in performance(it's even possible you could see improved performance)

Yeah that makes sense, I guess I just thought that there were a bunch of C++ devs out there for some reason. Doesn't pypy have threading issues as well though?

I just went for C++ over Java because it's what I know. Probably not the best approach.

As a developer with focus on C++ myself, I've had the impression that the lack of love for C++ here is primarily related to developers being able to do a lot that C++ has on offer with other languages that do not require you to be very intentional about how you use memory.

For a large part of my programming career, I've programmed risk management systems and extensions thereof. The software package has been developed in C++ and, naturally, the API for C++ extensions was very extensive, well-tested and therefore robust. Certain parts did rather lend themselves to be done in C# (services come to mind) which worked well, but I did run into a few hick-ups here and there.

From my impression of working alongside other devs whose financial software package was developed in Java and hence their extensions were Java-built, I could see the massive performance advantage we had in our C++ world. How much the software packages, the languages or simply the set of problems is to blame - that's a difficult thing to answer.
At least from my experience, performance wise, C++ used to kick Java so hard in the nads that it threw up a little bit every time. I freely admit that Java has come a long way to reduce the gap. I still have the impression there is a gap, but first and foremost hate having to invoke a silly JRE with its memory bloat and limitations.
Given my background, I'm obviously in the "C++ rocks runtimes" camp. So take any such statement from me with a healthy dose of skepticism and get a second/third opinion.

I sorta understand why you're given a hard time, Fubster. The reasons have been laid out plenty times (the way the ad sounds and the income potential for the profile you're looking for).
At the same time, I am the type of person who would have looked into that just out of sheer curiosity. Given my limited amount of time available, I highly doubt I could be of much use to your project regardless of the compensation model you are going to settle for.

We can have an informal chat about the project if you're interested and are at liberty to discuss it freely. Also, if you have specific questions regarding the architecture, feel free to PM me. If I can point you in the right direction, I'll gladly do so. If I can't, I'm not afraid to admit that I do not know.
There's no need to spend money on getting (a bit of) advice if it can be given without a ton of work/research.

Manager phoned screened someone for my job, came back and said to me "he said he's 10/10 in html and css, think up some stuff to throw at him when he comes in next week" so that'll be fun.

Also just for fun find and replaced "parseInt" with "~~" in our entire codebase

amirite? My salary requirements are about three fiddy.

but are you 10/10 in html and css? oh and the kicker: he's 5/10 in "javascript" but made it clear he meant jquery.

First time interviewing someone, Grue?

You know, I wrote a response to your post about interviewing up-thread and deleted it.

I've found that 95% of applicants will lie to no ends to get that interview. They don't realize that they are setting themselves up to be destroyed by any interviewer who is somewhat competent. I've never once been in a position where I had to choose between person A and person B.

10/10 sounds suspicious to me unless this guy has some impressive hand-made sites to show for it. Anyone who makes this claim is probably saying what they think I want to hear, will probably be far too polished during the talking part of the interview, and fall flat once they are tested. Humility and the ability to say "I don't know" goes much further than Mrs Perfecto.

I wish I had some insight from interviewing that helps me as an interviewee, but I don't. I know how to write a far better than average resume, but I know the interview just comes down to gut feeling. I've had people come in more than able and willing to take a job but get turned down. The most difficult part is the rule that both interviewers have to say "yes." I don't know if this is a good thing or a bad thing, but the turn over is surprisingly small in the departments that work like this. In contrast, other departments have one BDFL who does the interview and makes the final decision and their turn-over is through the roof.

Yeah it's pretty much a given that Java trades off performance for cognitive simplicity and maintenance. Although as machines get faster and faster that gap is becoming irrelevant for a lot of applications. Obviously HFT would not be one of those.

Thanks for the perspective. I have no problem talking about this project. So far I've got a few people that seem interested but I really have no idea how legit their interest is or if it'll work out or what. I have a feeling that I'll end up going the contractor route before this is all over, possibly with some of the people who are interested now, and I may hit you up for advice about that when I get there.

I've actually seen some "HFT" platforms built on Java, but it's just for OTC stuff and it's mostly for retail traders who want to pretend that they're low latency gunslingers

I've sat in on probably 5 interviews over the past year. We were trying to replace a sr. web services dev (java), and my manager phone screened ~30 people and brought in about 15 for face to face. They all started exactly the same way: fizzbuzz. I'm not a huge fan of it but went with it.

No one ever heard of it, and everyone other than 2 people (the first, and the last) couldn't do it, like 30 minutes of bang-head-on-desk watching people fail fizzbuzz. Everyone had 5-15 years programming experience on their resume. We hired the last guy literally because he could do fizzbuzz (not great, but got through it), and then could describe a REST API in terms of verbs and uris. I mean that was it, he did that and we hired him. This was for a $100k position.

Sellers market.

Oh yeah. 10/10 in both if you let me talk to your manager.

At Google, you are asked to rate yourself on a scale of 0-10 for a variety of skills.
I was told of a funny story where a guy interviewed at Google and rated himself a 10 on Java. The interviewee drew Josh Bloch as an interviewer. The interviewee was not hired and probably recalibrated his 10 after that.

Confirms it.

I'm wasting my time at uni

That's sort of the problem with interviewing at Google:
"How would you rate yourself at Linux?" 10
-- Oh great, we'll have Linux Torvalds interview you!

"How would you rate yourself at Python?" 10!
-- Oh great, we'll have Guido interview you!"

"How would you rate yourself at..."

Aren't you in the bay area?

nah major city in the midwest working for an incredibly unsexy fortune 100 company (for the next week).

I would legit have passed that interview, how would people not know fizz buzz is like THE blah company interview question.

I think the only people who should be able to say "that's easy to do" are the people who can actually do that type of work. A non-programmer that thinks its easy to do a specified programming task is someone to be avoided imo.

I've heard of fizz buzz but never tried it. I am semi-drunk. This took me less than 30 seconds in jfiddle:

I'm sure there are more elegant ways (now that I've read the wiki - if (i %15) is pretty cool). But still, what in the living ****?

But - based on our experiences with "Create an object that represents an N x N grid"--the trivial (or so we thought) first part of what a harder problem that somehow had hung up a lot of interviewees--I guess I believe it.

I think there is some fundamental disconnect with how a lot of programmers process verbal/logical instructions vs. how they think when they write code. Like we were informed many many times that the guy who bombed the N x N object creation is a really good programmer. And I believe the source.

But there was a huge schism between a set of simple instructions and code output. His brain just didn't seem to work that way. Maybe he's turning visual inputs into code or something? He's just really ****ty at turning verbal logic into code?

But when I think about it - he drew out the N x N grid exactly as described. So I know he understood the problem, and was visualizing it. He just got completely lost trying to create an array within an array.

Ok now I'm back to wondering how this guy could possibly be a good programmer. Maybe he's just really good at learning libraries and getting useful stuff out of them w/o really understanding fundamentals. Is that possible that someone like that could be productive and respected by his peers as a decent programmer?

One of the little wrinkles it's supposed to be SUPER AWESOME to figure out is that (i%3==0 && i%5==0) can be replaced with (i%15==0).

I've never understood FizzBuzz. Is this supposed to be "wrong"?

I've seen people be like "That solution sucks. It's not elegant. You're duplicating in code the check for i%5==0 and there are too many nested ifs".

This seems to be nothing but opinion to me. I mean, I think suzzer's version is a bit more readable, but the one above it seems totally clear what's going on and it actually involves slightly less computation, for whatever that's worth.

If it's literally just "can you solve this problem" then sure whatever. It's incredible to me that there are programmers out there who can't, but OK. The second it's supposed to be about "insights" like the i%15 thing or "elegance" in removing nested ifs etc then I think it's super bad.

I think either one is fine as long as you don't have any redundant logic.

But yeah as I noted in my edit - if (i %15) is pretty dope if you actually figure that out during an interview. If anyone ever makes me do fizzbuzz I'll start with if (i%3==0 && i%5==0) then pretend to have a deep flash of insight and change it to if (i%15==0).

Combining them into a single expression only works if the two numbers are prime numbers.