The high cost of a free coding bootcamp
theverge.com> Once, a student asked about the difference between JavaScript array methods, like push versus concat or forEach versus map, the team lead said they were interchangeable. (They are not.)
Technically, push and concat can be augmented to achieve the same thing. Same with forEach and map. Might not be the best fitted screwdriver but it’ll still kinda work.
Except that push mutates the array in question and concat returns a deep copy. And forEach returns undefined, while map returns an array. If someone claimed they were the same in an interview I doubt they would get hired.
Technically, yes there is a difference in semantics and convenience. pragmatically it depends on the programming style and language features / expressiveness.
Developers were able to map from one list to another list using for loops before functional-style map-like functions. Map is just semantically more concise about what it’s doing, rather than how.
As for mutation versus immutable copy behavior, that’s for convenience. Developers can also clone collections themselves.
A more charitable means of evaluating the class lead’s comment is that the core of an algorithm is not affected by these things.
One could argue multiplication is just repeated addition.
Do you mean a shallow copy?
30k?! You are honestly better off taking two yeas of night courses at a community college, and it will probably cost you half of that for more classroom time.
Bootcamp programs are really crap. Even in a traditional classroom setting, you only get out of it what you put in. I know CompSci majors who graduated only knowing Java and who were totally lost if they had to replace a ram chip in a computer. I knew others who would try new languages every semester and who probably still keep up with tech news in our industry.
Boot camps are more of a filter to find people who are going to put in that effort anyway. You have to in order to be able to make it. But I feel they're also part of this hustle culture bullshit; you're better off taking your time and learning slowly (and more affordably) rather than trying to cram everything in to 12 to 24 weeks.
I'm currently teaching a Bootcamp by one of the major players in this space. 100% in classroom. On week 12 of 24. It's a side / fun job for me. Not my day job.
While I disagree they are "crap", it's definitely a what you get out of it is what you put in situation.
I have students who knew very little about the subject matter, put in a ton of effort and are rocking. I have other students who knew some, put in 0 effort, and are clearly struggling / stopped caring.
I feel that I'm basically a long-term tour guide for the subject matter (Cyber Security in this case). I absolutely cannot teach the ins and outs of all aspects of the subject matter. I regularly remark "I could probably teach a class on this thing we spent 10 mins talking about".
What I feel I can do (and actually do) is give them a subject or concept, introduce it, and let them run with it. I'm then available to answer any questions, provide lots and lots of anecdotes about the reality / theory of theses subjects.
We definitely move fast and skip over more theoretical things...but I think the students want that. I spent 1 day talk about the OSI 7 layer model...in college I spent weeks learning about that.
> I know CompSci majors who graduated only knowing Java and who were totally lost if they had to replace a ram chip in a computer.
The rest of your post notwithstanding, I don't see why a CompSci major should need to know how to replace a stick of ram.
They should know it all: where the cpu and ram is, how the motherboard look like and etc. Nobody expects them to go out there and start fixing computers, however if someone spends 4 years studying the subject and can't even be asked to look inside the box,well then there's not much hope... A couple of weeks ago I was on the sales floor and was about to change RAM on a couple of PCs( it's not my job,just wanted to do it). The sales rep pulled the PC from underneath the desk, took the side cover off,removed the old ram and put the new one in.If sales can do it, I'm confident anyone in CompSci should be able to do so as well.
> They should know it all: where the cpu and ram is, how the motherboard look like and etc.
Why? That's not Computer Science. Computer science is the science of computation, not the science of computers.
> Nobody expects them to go out there and start fixing computers, however if someone spends 4 years studying the subject and can't even be asked to look inside the box,well then there's not much hope...
Hope for what?
> A couple of weeks ago I was on the sales floor and was about to change RAM on a couple of PCs( it's not my job,just wanted to do it). The sales rep pulled the PC from underneath the desk, took the side cover off, removed the old ram and put the new one in.If sales can do it, I'm confident anyone in CompSci should be able to do so as well.
Why? It's about as related to CompSci as doing field-expedient surgery is to Psychology.
> In the seven decades he was at M.I.T., Professor Forrester retained an engineer’s curiosity about how things work, and occasionally voiced dismay that his students were not always so inclined.
> He recalled in 2011 that he once asked students in an engineering class if they understood how the feedback mechanism in a toilet’s water tank maintained the water level.
> “I asked them, ‘How many of you have ever taken the lid off a toilet tank to see how it works?’” he recalled. “None of them had. How do you get to M.I.T. without having ever looked inside a toilet tank?”
Would you hire such a person for embedded software work? Also having a feel for what can go wrong with RAM can be useful in case a computer acts up. Then there are details like price, speed and parity which may or may not play a role when it comes to hardware selection. And last but not least software security researchers have been seen using cooling sprays and taking out RAM bars to extract information from systems.
They don't need to. As astronomer doesn't need to know how to repair his or her own personal telescope either. But if you're into the field, it's nice to know how your tools work.
You may not know how to fix it but surely you'd know they have lenses inside,right?
I think as far as learning to code, bootcamps aren't much better than following tutorials online. I do, however, think that bootcamps are good at getting people who already know how to code to learn how to code well, and code in a job environment. There's probably a good market for classes aimed at people who have already learned basic coding through online tutorials but now want to refine their skills to the point that they are viable in the job market.
There are exceptions, I think, when the boot camp instructors are able and willing to impart significant industry knowledge, war stories, and personal teaching time to students. You don't get this at the vast majority of boot camps (not even Hack Reactor these days), but they do exist. That kind of experience isn't the kind that's easy to pick up, and getting it in a highly condensed time frame to facilitate a career change can be a worthwhile investment.
Source: graduate from one of said rare boot camps, and gainfully employed (not on the frontend, though many from my cohort chose that route as well)
Lambda School is 9 months full-time (40+ hours/week) and 18 months part-time (20+ hours/week).
20 hours for PT? Schedule only has PT down for 15 hours, 3 hour blocks x 5 nights, which is why FT has advantage over PT. Also the 5 extra hours are not explained at the start that we have to do on our own. Makes planning ahead a bit of a hassle. Another reason a lot of PT tell new people to do FT, it is more organized and better program.
Of 20 people I've interviewed who completed a coding bootcamp without prior CS education in the past year or so, I think 1 passed the phone screen (all front end). At this point I outright reject them because of the futility.
The article states that the ISA aligns Lambda with the student - but I thought I read that they generally sell off the ISA, meaning they no longer have that alignment....
I am not going to hire anybody from a free coding bootcamp or a coding bootcamp for that matter. Seems like they turn out people who know how to hack things together based on shallow knowledge.
I hired someone from a bootcamp who was making a career change and it was easily one of the best decisions I’ve ever made. Took a little while to ramp up, but that process helped me be a better manager and helped the rest of the team feel comfortable mentoring someone.
It’s not for everyone that’s for sure. But saying you’ll never hire anyone means you’re missing out on some great candidates. Take a look at people making career changes, and where you can use that domain knowledge to broaden the team in a way someone with CS degree necessarily might not.
Maybe it's an opportunity for people who wouldn't otherwise get the chance. I've met lots of young bright folks stuck in mcjobs, sort of at the mercy of circumstance. Maybe trying a coding camp will turn on a lightbulb.
There will be some filtering required.
maybe can be suitable gig hires to turn out quick cheap startup ideas.
People calling out Lambda is what PG's "Haters" essay is about http://www.paulgraham.com/fh.html
The attempt to portray income share agreements as bad is very large stretch:
> These ISAs are the bedrock of Lambda’s program. They allow the school to market itself as an “accessible” computer science education. But critics, like Sen. Elizabeth Warren (D-MA), have warned that ISAs carry many of the pitfalls of traditional private student loans, “with the added danger of deceptive rhetoric and marketing that obscure their true nature.”
Yeah, except for one massive pitfall that student loans have and ISAs don't: With ISAs you only start to pay once you actually start making money whereas student loans saddle graduates with debt regardless of their job prospects. If the graduates of the Lambda school don't land jobs then the school doesn't get paid. With traditional student loans, the school gets paid even if their graduates are unemployed.
ISAs seem strictly better to me. It shifts the risk burden onto the school, and creates monetary incentives to improve the job prospects of graduates. By comparison, traditional loans put the risk burden on the student and because the university is paid upfront there is not as much incentive to better the job prospects of graduates.
> Yeah, except for one massive pitfall that student loans have and ISAs don't: With ISAs you only start to pay once you actually start making money
That's not that different from federal student loans, since those all have income-contingent repayment plans (as well as pay as you earn, revised pay as you earn, and income-based, which are generally similar but have different repayment formulas) available. It mainly differs in the reduced flexibility with the ISA.
> If the graduates of the Lambda school don't land jobs then the school doesn't get paid.
ISAs are marketable assets and can be sold even while the student is still in the program, and the school gets paid by selling the asset. The new holder of the asset doesn't get paid if the student doesn't make money.
"strictly better" and "carry many of the pitfalls of traditional private student loans" aren't contradictory.
ISAs can be, and are, better than student loans. But they're not a universal good, as demonstrated in this article: if you are on the receiving end of a crap course that doesn't help you get a job is it really good that you'll have to pay for it (when you get a well paying job from entirely unconnected means)?
> if you are on the receiving end of a crap course that doesn't help you get a job is it really good that you'll have to pay for it (when you get a well paying job from entirely unconnected means)?
When the alternative is:
> if you are on the receiving end of a crap course that doesn't help you get a job is it really good that you'll have to pay for it (with interest, even if you're still unemployed after taking the course)?
Yeah, the former is really good.
Sure, it'd be better if the courses were free. But are you going to spend two years teaching computer science without pay? I'm not going to, and I'm going to go out on a limb and guess that most people won't either.
> When the alternative is
Your supposition is that there are two options for everyone:
a) go deep into student debt
b) go deep into ISA debt
and no others. I contend that isn't the case. You could do neither. When you're in danger of getting into serious debt via an ISA to a company that doesn't seem to be competently running courses that will lead to employment it's only sensible to question whether it's worth signing up in the first place.
There's no "ISA debt". There's no interest racking up. You may not even pay anything if you don't get a well paying job.
Sure, an ISA from a bad program is not good. But at least you're not paying unless you get a well compensated job.
> There's no "ISA debt"
?!?! Of course there is! What are you repaying when you get a high paying job?
You're paying a percentage of your income, if you're making above a threshold. This is in comparison to debt which has interest that makes the debt increase over time until paid off - regardless of whether you are employed in a well paying job or even whether you're employed at all.
Yes, under both system money is being paid back to someone. But crucially, there's no interest with an ISA. And furthermore, no money is taken until people get a job that pays a certain amount.
Pull up a chair, DHH is going nuts on twitter about this rn