wow, that's insane. because there is basically no way any but the top 5% of boot camp grads are delivering 70k in value....

LOL you think bay area programming is about delivering value. So cute.

hahaha

Maybe they don't hire junior devs with the expectation of getting a perfect 1:1 return on their first year's investment?

This whole conversation illustrates the fact that pay in programming is looking through a fish eye lens. It was immediately obvious to me and others that Noodle wasn't talking about $70k, but about half that (the obvious giveaway is that we'd be e-high fiving him instead).

The alternative? Unless rent is due next week and he needs something pronto, there is no reason to not hold out for something better, in a tech he likes, and something that he isn't calling "crappy" before he even considers an offer.

Speaking of tech we can hate on. Am I the only one that hates Jenkins?

Even the configuration is all wonky. I mean, you get someone who gives up and their <ip>:8000 is exposing their entire setup to the world. Who needs script kiddies these days?

"Please wait while Jenkins is getting ready...."

Oh yes, auto-refresh is on... don't have anything else to do in my life but sit and stare at that **** for 2 hours.

Noodle in the time we've all been navel-gazing about this you could have already taken the job, extracted value, and moved on to something else. Just do it.

Unless you get raped or murdered or something similar, there's no such thing as a bad experience. I even learned a lot, made good contacts, and had a gigantic lesson in how not to act if you're the boss - at the startup in LA with the psycho boss. I lasted all of 4 months before it led to something much better.

Why can't he just take the job and continue looking? He can tell other prospective employers that it's a contract that will end soon and this why he is looking.

Unless he has to drop out of school, I think he should take it.

What's the best practice way to implement mixins in javascript? I'm following this tutorial but it feels really clunky how it handles it.

I considered using underscore but there were some harsh words thrown around here about pulling in underscore for simple projects like this.

if you need a utility method thats not covered by the built in ones (i.e. map filter reduce etc) and you're using it a lot I have no problem with pulling in lodash, way better than defining them yourself i.e. copy and pasting shims.

Anyone spent some time interviewing/hiring offshore developers? I think I've determined the core of my issue which is trying to figure out what an acceptable "word salad percentage" is. i.e. you ask them a technical question that they can't answer well and you'll always get the same thing which is them just throwing words at you and hoping you'll eat it up. We're going to hire a guy who I think was in the 30% range and that seems pretty high. But I've talked to candidates that were 80-90% before.

Interviews are such crap shoots. I feel like all programming jobs should start with a paid 1-week trial. You know after one week if they're any good or not.

Our model is the onshore devs (who are generally very good) deal with the offshore devs. If we see one of them consistently checking in bad code we will say something. But when they're new the onshore devs typically check in code for them for a while (presumably after heavy vetting). So we don't really know, but the system seems to work now that we have a bunch of solid onshore devs with 2-4 years of experience on our site.

I'm gonna double down on this and say that a very low paying job is unlikely to mean a very successful business-unit/company and not only will you be potentially impacting your future earnings (by having made very little if people demand to know) but you will likely be less valued and less likely to be given any responsibility or difficult/fun work.

It isn't of course black and white and there are good and bad, but if you are entering into the world of software development and are capable, you should command a solid earning wage in this economy. If for some reason a company cannot afford to give you that, it signals negativity on many levels.

suzzer,

I think the model that you use of off-shore/on-shore made sense 10-15 years ago, but with the flourishing open-source economy (and tons of other factors), it makes literally no sense anymore. I'm very convinced that qualified teams with all very good players, even while much smaller, are significantly more effective than larger teams (sometimes even an order of magnitude or two larger).

Do you agree or disagree with this?

I don't even think our model makes sense from a budget POV, because the offshore company makes all their money off the offshore devs, and breaks even or loses money on the onshore devs - who do 90% of the actual work. But it was dictated by pay grades way above our heads, so we make it work the best we can.

Yes in general a small team of rock stars is almost always going to be the way to go. Although I will say at the side-job startup right now, some of the devs tend to want to go off in caves and work on their little piece. It's very hard to get them to communicate (exacerbated by them being scattered all over the globe), which is a major impediment no matter how good at coding they are.

I've used off-shore developers in South America (many parts), India (2-3 parts), Vietnam, and Eastern Europe, and around the US. Some of the developers were completely behind a language barrier.

I think in certain areas the "word salad" is kinda cultural and the 30% people are often stuck working and habituating with the 80-90% people, so you have to allow for some of it. Some people do it as a nervous reaction as well but then get more comfortable and speak normally. What is funny is I think this is essentially the same behavior that exists between middle-management and executives in many companies. People who are not really sure wtf is going on but like to follow trends and the latest tech buzzwords that doesn't matter to anyone doing anything real.

So the only reasonable excuse is that they cannot be trusted to self-hire these necessary programmers on-shore?

As for the second point, that is why leadership is so important in a small company. You need to be able to all have a common goal and continually make sure you are all in touch about what that goal is.

There is no reasonable excuse. Someone somewhere got a huge bonus and a bunch of free trips to India (and maybe other perks) because they can somehow technically show our offshore company is delivering big savings. You can always manipulate the books to be able to the higher-ups "See we were delivering X amount of work for $Y, not we're delivering the same amount of work for 20% less!"

But of course no one knows what X really is. Show me someone who says they can accurately quantify programmer units of work in an objective way - which somehow abstracts out the skill level of the programmer doing the work - and I will show you a bull**** artist.

In the beginning the difference was made up by all of us full time employees working our asses off to train the onshore devs (at that point the offshore devs were 100% useless but still billing tons of hours). A few of the long-time devs were still on contract. One guy billed 95 hours one week. Lucky ducky. I worked the same as him but am on salary. Needless to say we weren't seeing big savings then. But I think the C-level higher-ups were always sold that it would take a few years for those to materialize.

Now I think the offshore company makes up a lot of the savings by working the onshore devs to death. They have to work all day, then spend hours every night on calls with offshore. They tend to find someone else to sponsor their visa and jump ship after a year or two onshore. But we've had a few stay for up to 4 years.

And we have had a few onshore devs go back to India and work from offshore. Which is good because they can train the other junior onshore devs locally. Basically we're finally at the point where we have roughly the same continuity we had before the offshore company came into the picture.

And to be fair I guess one of the big motivating factors was that we had to turn down a lot of work pre-offshore company, because we just didn't have the capacity. But that's mostly on my lazy-ass bosses for barely bringing us any candidates to interview.

if you give some details on what you're trying to do, i can probably give a functional solution that better than anything you'll get with mixins.

this might help too: http://raganwald.com/2015/06/17/functional-mixins.html

Btw in my previous post:

You can always manipulate the books to be able to the higher-ups "See we were delivering X amount of work for $Y, not we're delivering the same amount of work for 20% less!"

Should have been:

You can always manipulate the books to be able say to the higher-ups "See we were delivering X amount of work for $Y, now we're delivering the same amount of work for 20% less!"

Got this "code challenge" today as part of the interview process, thoughts? seems a bit excessive..

Anyone have recommendations for load testing tools? Ideally something SaaS to avoid setting up a bunch of servers to simulate the amount of concurrent traffic I want.

I don't need to simulate huge load, just ~1000 concurrent requests sustained for a minute or two.

Grue, that doesn't look like more than an hour or so of work unless you want to go crazy with styling and tests right?

Assuming my time estimate seems about right, I'd be willing to do something like that after face to face interviews, but not before. Before face to face interviews I think it's too much of an imbalance in amount of effort put forth by each party.

think it will be a lot more than an hour but I'll keep an eye on time. Obviously I can just do it during my day job...

I forget the discussion we had last time when suzzer was negotiating for the contract position - what did we think was a reasonable rate for a senior/super-senior contractor these days?

Jenkins (like most on-premise/self-hosted stuff these days) is for companies that have dedicated operations teams that can keep things running and secure. For everyone else there's TravisCI/CircleCI/Codeship/etc.