I have a hard time imagining somebody who truly cares about programming to learn about Joel's blog and not devouring it (I mean the earlier days, when it was still about programming). I could let it pass if they never heard about Joel, but they know lots of other interesting stuff. Or anything else, really - it is the basic difference between somebody just doing their job and somebody curious about everything. If somebody is a Java developer and only reads about Java programming, he might be a good fit for some jobs. But I am different - I constantly read about new programming languages and development trends.
Why should it be good for me to work with somebody who only cares about Java? And what would you take about during lunch - Java development?
I am not saying you should hire based on musical taste, but I am waiting for a solid argument for why you shouldn't.
More specifically: why is "hire who you know/is like you" supposed to be a bad heuristic? If you know you yourself are efficient, it seems likely somebody like you is also sufficient. At least it seems more likely than that a random stranger (or somebody who explicitly isn't like you) is efficient. Please show me data to the contrary if you disagree with that heuristic (especially, as I said in my first comment, as nobody really seems to have a clue how to hire).
Back to Joel: it seems more likely that somebody who knows Joel is a good programmer than somebody who doesn't know Joel. Would you disagree with that assessment? Look at it in reverse: how would a good programmer not know about Joel's blog?
"But isn't there a basic curiosity in some people?"
Someone who does not read the same old blog as you is not curious? That does not make any sense. I checked out Joels blog now (after years) and there does seem to be anything exceptional or different then what is on gazillion other blogs.
If anything, Java guy reading about Java could learn about map reduce, full text search, security, compilers, algorithms, sound processing, text analysis, artificial intelligence and million other things in the process. There are java libraries for all those things and when you follow java news you are primary learning about those.
Mastering full text search library in ruby is easier if you already mastered another full text search in another language. The underlying concepts are going to be similar.
I get it, you care more about language itself, but that does not make you more curious nor better developer. It makes you slightly different.
"And what would you take about during lunch - Java development?"
I would try various topics until I find the one we can talk about. Plus, discussion with someone who knows different things then me is more interesting usually, I learn new stuff from it. Listening to the same ideas I read about yesterday does not make for a compelling lunch.
"I am not saying you should hire based on musical taste, but I am waiting for a solid argument for why you shouldn't."
I'm willing to bet that you don't live up to your own standards of being friendly with people. But I think you don't even understand what I am trying to say.
"I would try various topics until I find the one we can talk about."
I am not talking about meeting once and making small talk. I am talking about meeting every day, for 10 years straight.
Curiosity: if somebody can give me a good reason why they don't like Joel, OK. But if they have never heard of him, it seems likely they are not curious about programming. Btw please mention some of the other gazillion blogs that say the same things? (Refer to the top list he has in the sidebar)
"You might as well hire on color of their eyes."
You think people have as little control over their musical tastes as over the color of their eyes? If they like death metal, it is just a random variation that says nothing about their personality? Or if Gangster Rap is their thing? Musical taste says nothing about personality, is that really what you believe?
> Curiosity: if somebody can give me a good reason why they don't like Joel, OK. But if they have never heard of him, it seems likely they are not curious about programming.
You are conflating with what you are interested in with what everybody should be interested in. Some people just don't care about Spolsky or aren't big fans of his writing. That doesn't mean anything about those people's interest in programming or computers.
The whole point of this article is that companies and teams are creating cliques out of cultural preferences and setting up those that fall outside of their preferences as inferior for not liking what they like. The last several posts you had in this thread reenforced this idea: well gee I found this thing interesting/useful, anyone who doesn't clearly isn't any good.
"That doesn't mean anything about those people's interest in programming or computers."
Of course it does. Joel advocates a certain style of development. If you don't like him, you likely don't like that style of development. If your company works "Joel style", why should you hire somebody who doesn't like that development style.
Sorry, but the criticism here is ridiculous. It's clear people don't even think about what they write.
It sounds like your answer to question "Are you hiring programmer or drinking buddy?" is "I'm hiring a drink/lunch buddy".
If you still work together 10 years later, your topics will be much different then. Whatever you read now is irrelevant. However, with company hires being focused on dinners and lunches small talks, the company is not likely to stay together 10 years later on. Not unless huge changes in culture happen.
As for music taste and personality, I do not see much of reliable and relevant to employment. If they would act like jerks towards anybody who does not share their passion or would insist on playing it loud despise other people objections, yeah I would have the issue. Other then that, nope. As far as I'm concerned most people listen whatever was the thing where they grew up.
You think people have as little control over their musical tastes as over the color of their eyes? If they like death metal, it is just a random variation that says nothing about their personality? Or if Gangster Rap is their thing? Musical taste says nothing about personality, is that really what you believe?
I've always enjoyed finding out what's on the playlist of the great developers I work with, because I am always surprised.
anecdotally, yes, i've noticed very little correlation betweeen people's personalities and their tastes in music. why would you even expect there to be one?
How do you propose to hire somebody you don't understand? It seems to make a lot more sense to judge people based on the things you know.
If you believe X is important for programming, why on earth would you hire somebody who doesn't like/know about X?
It's possible that you are mistaken about X being important. But then you are screwed anyway. But if you start hiring based on things you don't know, you are also screwed.
At the end of the day everything you do is based on your beliefs. Maybe you started an IT consulting company because you believe Rails is the best thing ever. Then why should you hire somebody who doesn't believe that? If you think about it, the whole premise of the article is completely ridiculous.
It seems to make a lot more sense to judge people based on the things you know.
So do that. Stop pretending you are a psychologist, and capable of figuring out what kind of personality is good at this job, because those are things you don't know.
Instead, give work-sample tests. Those are something you can know. You can look at the results the person emits. Then hire based on that. Personality only needs basic things like "doesn't spit on other people" and "doesn't say racial/sexual slurs to coworkers/clients."
There's no reason to think that a psychological test designed by non-psychologists is even as good as random guessing. On the other hand, work-sample tests have been studied and are a very good guide. Commenter 'tokenadult has FAQd this up for us: http://www.focus.vc/tokenadult-recruiting-faq/
So you would also say that it is best to leave all hiring to a computer, as I suggested in another comment (to summarize all comments I received)? After all, human perceptions seems to be too flawed?
A computer can administer a psychological test, and it can administer a programming test. Problem solved?
Other than that, again: you can only act based on the things you know. So the conclusion is nobody is qualified to hire (since the article talks to everybody)? Of course it is possible that there are some clueless people in the business of hiring. But people can only do what they can do. I don't see any practical advice in that article.
"I only know how to do ad-hoc personality tests, so ad-hoc personality tests must be how I can hire."
The very practical advise is to deliver a work-sample test. Tell the person exactly what you are looking for up-front("we want this code to run as fast as possible" versus "we want this code to be very readable") and then measure on what you said you looking for.
If you do a lot of code maintenance, pull up a piece of your code with a bug and have the candidate hind it. If you do a lot of OOP, describe a recent problem and have him design the class.
I think you're missing the premise of the article. It doesn't say don't hire qualified people based on a set of characteristics. The article brings up various examples where people are totally confused about what those signals and characteristics of qualification actually are. Dress code, common taste in music, beer, literature, etc. are definitely not those characteristics.
The article also doesn't say don't hire people you will get along with. You should try to hire such people but a workplace is a professional environment and not an extended family like some places like to paint themselves. First of all such places should be avoided because no workplace is an extended family and second because they are being unprofessional and exploitative if they pain themselves that way. The recent github fiasco is a good examples of what happens when people forget that workplaces are professional environments and github is actually one of the better places where work and life balance is highly valued. So yes if you're hiring a Rails developer then you should look for someone that understands Ruby and Rails but you shouldn't really care if they are a devout follower of the church of DHH because that has no bearing on their ability to do the job.
The "GitHub fiasco" would not have happened if they had only hired people who are like them. Bad example :-)
I personally don't think there is much to be learned from it. It was just a normal situation where employee and employer don't get along anymore - it happens thousands of times every day. I don't think because of that people should now shut down all emotions at their workplace and only focus on their work and never talk to each other.
You can not really demand that every employee has to be happy at every company and vice versa.
>Why should it be good for me to work with somebody who only cares about Java? And what would you take about during lunch - Java development?
Knowledge depth, for one. The time you spend reading about new trends and languages is time not spent getting deeper knowledge about the skills you already have.
Maybe you spend just as much time as he does gaining knowledge depth in your chosen language and spend some of your free time reading about trends and languages, whereas he might spend time with his family instead.
You're still basing this decision on an arbitrary signal like "Reading about trends and languages makes you a better programmer", which is a very difficult thing to prove, the least of all reasons being that there are no hard lines drawn in the sand regarding programmer skill levels and the positions that they qualify for.
But then someone with this same attitude comes along and reads about trends, languages, and computer science developments. That guy won't hire you because you don't read about computer science developments because he thinks those three sources of knowledge makes you an even better programmer. He spends the same amount of time gaining knowledge depth as you do, but spends even more time reading about trends, languages, and CS developments, regularly sacrificing an hour of sleep.
And then someone comes along and reads about trends, languages, CS developments, and hardware hacking[, and social engineering][, and UI/UX theory][, and mathematics developments].
You can see how this can quickly fill someone's entire waking hours in a rat race to be the guy casting down judgment.
>Back to Joel: it seems more likely that somebody who knows Joel is a good programmer than somebody who doesn't know Joel.
You can hang around forums and parrot whatever other people are talking about and blend in without ever having to elaborate or think about Joel's blog (not to mention actually practice any of it). This is the same sort of signalling that SV is using according to the article.
Having that bit of knowledge quickly skyrockets my perceived skill level above "the crowd". I'm an insider now, someone to get a little excited about hiring, and for no more work than it took to read a blog.
Then I start rattling off other things I've seen on here, reddit, and other places that make me seem intelligent and well-read. Now they're tittering to give me an offer.
Then they get a nice dose of reality when they learn I'm still pretty new with regards to actual skill.
If I wanted to be an SV insider, I'd follow the same procedure: read about what signals they are looking for, mimic those signals, and play it by ear the rest of the way. I guess if I had to code actual projects to make it then I can't be as deceptive because I'll actually know how to do something. (To save some of my dignity, I actually have done that!)
@facepalm I would much rather hire someone who reads in-depts books that those shallow blogs. Too much time spend on blogs and discussion forums is not that great thing.
You are right about forums. Although I have the impression many of the people who are very frequent writers on HN are also very good developers. So at least it's not necessarily a negative signal.
But it shows you are willing to read blogs to further your knowledge. Presumably even voluntarily. Would you really say that is just a weak signal, especially in a profession that changes as quickly as IT?
I think it's a weak signal. It should be a nice-to-have, not a deal-breaker. The article's signaling is even weaker in my opinion, and those are deal-breakers of the most superficial kind. It's too easy to gain significant credibility for so little actual work.
>Well I initially said "if somebody doesn't know Joel or Graham, I would probably count it as a minus", not a deal breaker.
Oh, sorry. I still think it can swing opinions too much being what it is.
Well I initially said "if somebody doesn't know Joel or Graham, I would probably count it as a minus", not a deal breaker. But why should it not matter to my estimate whether somebody knows that or not. Would you believe somebody who claims they are passionate about physics but have never heard about Einstein (to make an obvious example)? How would you go about probing a candidates enthusiasm and determination?
I have a hard time imagining somebody who truly cares about programming to learn about Joel's blog and not devouring it (I mean the earlier days, when it was still about programming). I could let it pass if they never heard about Joel, but they know lots of other interesting stuff. Or anything else, really - it is the basic difference between somebody just doing their job and somebody curious about everything. If somebody is a Java developer and only reads about Java programming, he might be a good fit for some jobs. But I am different - I constantly read about new programming languages and development trends.
Why should it be good for me to work with somebody who only cares about Java? And what would you take about during lunch - Java development?
I am not saying you should hire based on musical taste, but I am waiting for a solid argument for why you shouldn't.
More specifically: why is "hire who you know/is like you" supposed to be a bad heuristic? If you know you yourself are efficient, it seems likely somebody like you is also sufficient. At least it seems more likely than that a random stranger (or somebody who explicitly isn't like you) is efficient. Please show me data to the contrary if you disagree with that heuristic (especially, as I said in my first comment, as nobody really seems to have a clue how to hire).
Back to Joel: it seems more likely that somebody who knows Joel is a good programmer than somebody who doesn't know Joel. Would you disagree with that assessment? Look at it in reverse: how would a good programmer not know about Joel's blog?