> determine if a candidate cares enough to code on their own
Which he is arguing that you are not a bad developer for doing other things outside of work. The vast majority of people need to not code to regenerate.
I can use myself as anecdotal evidence: I'm the goto guy when it's particular difficult to solve a bug, when it's a new framework, or when the teams are unsure how to solve a problem. I love coding, I constantly get compliments for my code and solutions. I do rarely code outside of my work. I used to before I got a "real" job after my degree, but now I (almost) only code at work. I usually think about solutions and architectures outside of work though, and I write a lot of pseudo code on paper if I get an idea for an algorithm - but I do not make binaries or contribute to OOS or a public Github repository. I have a spouse, children, friends, and hobbies that simply doesn't get as much of my time as they deserve.
I would never pass the bar at a startup that's looking for the typical startup stereotype developer. I'm 100% certain that I make more business sense to hire than all of your average workoholic startup developer, but I would never be considered. I have seen so many developers with great personality and glowing Github repo fail so miserably because they have poor work ethics and are unable to do "the boring stuff" that needs to be done in a business.
I can make a sensible plan, estimates that aren't completely bonkers, and I can tell when we're no longer on track to meet the deadline the same week it starts to slip. If I make a promise I keep it, or let you know that I'm unable to. I can also work completely agile, and not some "agile-but" that is so prevalent in the industry. In short there is soooo much that you need to consider when hiring. Screening based on a Github account is excluding developers that would be just the guy/gal you're looking for.
I'm not seeing a problem with this. Not because you are a bad developer, but that you don't match the type of person they want to hire at that time.
> I'm 100% certain that I make more business sense to hire than all of your average workoholic startup developer, but I would never be considered.
What's your experience as a "startup developer" that you feel justified in making that claim?
> In short there is soooo much that you need to consider when hiring.
Of course. But I don't see Github being the only thing to consider, nor do I see it as being something that should be ignored. It's a piece of the puzzle.
You're a programmer. A good one at that. But have you ever stopped to consider they need someone more than that for where they are at? Someone who can be more than a programmer?
While I understand that many people enjoy writing code and do so for fun, there are a large number of fantastic engineers that don't (myself included). Just as I wouldn't give someone preference for a robust profile, I'm not going to penalize someone for lack thereof. There is only so much time in the day and if they can do ten hours worth of work in eight hours and choose to spend their off time with their family, friends, or hobbies, good for them. Knowing how to decompress and enjoy both work and life is just as valuable to me as crafting solutions to engineering problems.
Citation needed. It seems to me that the majority of people don't.
This is covered by the OP, but if you are using Github as a part of your hiring decision, you are going to be putting at a disadvantage many different kinds of people who do not or cannot contribute to open source or personal projects. The OP argues that Github is not necessarily a useful indicator of the things you mentioned and if Github becomes a very influencing factor for many companies hiring decisions, then it becomes de facto required and means you must work more as a programmer than you already do. This extra work is more able to be done with those that have free time and more personal resources (e.g. income), so some types of people are probably being excluded from your hiring process, leading to more a of a developer monoculture.
But I do find it funny that they don't think that Github should be used as a resumé but their "conferencing", tweets, books about diversifying and podcasts should do.
Talkers or Makers?
I write OSS on my free time because I feel like other alternatives might not be good enough. I spend a couple of hours in the morning, weekends and etc. It's hard and time consuming. I can take two days if just figuring out what the selector should be called or argument names.
Right now I am getting paid for OSS, but I even spend my free time on it because I'd like to make sure it comes out good.
It's very slow and time consuming, and sometimes keeps me up at night, but I LIKE IT. And I like acknowledgment to it.
These sorta articles invalidate what I make just because I FEEL like making them. Not sure what to feel about that.
I have friends who make stuff to Go-lang because they want to learn the language and make stuff that doesn't exist. These articles invalidates their passion.
And saying bullshit like LGBT or other genders are actively being excluded... well it makes me hard to take you seriously. I'm not white, and I am part of LGBT.
Also, I'd challenge you to find where the article says or even implies that you shouldn't be making stuff because you FEEL like it. Quite the opposite, in fact. You should be making stuff because you feel like it and not because you have some expectation of getting a better offer or being able to show off your GitHub profile at your next interview.
As for the white, male bias...not sure what to tell you. It's real. The statistics don't lie. I'm glad that you have had success as a non-white member of the LGBT community, but as the saying goes: the plural of "anecdote" is not "data".
Actually, Ashe has that on the Hire Me Page.
>As for the white, male bias...not sure what to tell you. It's real. The statistics don't lie. I'm glad that you have had success as a non-white member of the LGBT community, but as the saying goes: the plural of anecdote is not data.
I understand this is a fact, but how this can be used against Github as a resumé or against OSS on your free time, I have a hard time seeing that. It comes off as incredibly stupid.
The thing is, something can be both a filter against 'underrepresented groups' AND a valid criterion for hiring.
Requiring a CS degree probably biases in favor of white and asian males and against black females. It is also a very valid filter for hiring programmers (pace the self-taught).
Not much can be done about this fact except for increasing the pipeline of underrepresented groups into CS degrees, recreational programming, OSS participation, and so forth.
Maybe it's because I'm getting old, but I've learned something... stop caring what others are saying and doing. If you feel like making shit, then make it.
I hope one day we will be able to distil the ultimate rule: "being smart and proactive is discriminatory".
OSS as a hiring filter biases your selection process toward people who are either paid to do it at their current job, or who have few enough prior commitments in the rest of their life to be able to afford to do it outside of work.
I'm not sure whether it biases towards white men or not, but I'm pretty sure it biases towards younger people and against people who are in enterprisey companies (or who work on private contracts).
Also please read my above comment. https://news.ycombinator.com/item?id=6740089 The article links to another article about why it favors white men.
What?
> You just get what other people think is useful. Aside from which, GitHub displays a lot of useless stats about how many followers you have, and some completely psychologically manipulative stats about how often you commit and how many days it is since you had a day off.
You get social proof, some measure of conscientiousness, and a crowd-sourced filter. All on one page. That sounds great. I don't see the problem.
Companies want to hire influential workers. It's relevant. So is putting in the work.
And products which people find useful are vastly superior to those that the author finds well engineered. I'm guessing many company owners will share that view.
I'm actually lucky enough to fall into the former category, but I feel that if you only hire people with a respectable presence on github, you're disregarding a large number of people for something beyond their control that doesn't accurately reflect their ability.
What you spend your free time on is not beyond your control (at least when you're a software dev and presumably paid enough to actually have free time). And you might decide to spend it on other, unrelated hobbies. That's cool, but you should be aware that unless you can present that hobby as interesting, unless you can sell it as something that helps you grow, you are making yourself look less good that someone doing the same kind of work, and then doing OSS in their free time.
A company is not going to disregard you for not having a respectable github presence if no other candidates have a github presence either. But if there's a clear disparity, why should they take an extra risk by hiring you, rather than hire the person that can show them their actual contributions?
Maybe I see the world through rose-colored glasses, but I don't see why sane employer would like to forbid contributing to open-source project if e.g. this project is used for work and has some bug. And there are always bugs or inconveniences if you use something heavily. For some industries it is even hard for a developer not to be on Github, because a lot of software she uses is on github, and in a lot of cases the best way to fix the issue she's working on is to contribute to open-source project that caused this issue, instead of e.g. working around it by some hack.
I agree that discarding people without respectable github presence is bad (unless you're looking for a very specific kind of people, e.g. for marketing purposes). But I can see how not being on Github at all (even with minimal presence) could be a bad sign in some cases. Of course, industries are very different and this shouldn't be a general rule.
I've seen it lots on GitHub, and the problem seems to be growing. The guy that parachutes into a project -- one with money-impacting code already deployed on 1000's of real world systems used by ordinary people -- submits a pull request where he's modified every single file in the project, making only superficial changes to formatting and whitespace, and probably the result of an automated tool, and most importantly DOES NOT THOROUGHLY TEST IT beforehand... yet, from GitHub's perspective he's now (a) an accepted code contributor, (b) modified lots of files -- prolific! -- and (c) has a high count of lines added/removed -- IMPACTFUL! -- and (d) very very busy beaver "working" on lots of projects. "Gosh, we want him!"
But does he/she understand threading, leaks, races, parallelization, good architecture, good documentation, automation, reproducability, security, scalability, strategy, pacing, production support, risk mitigation, algorithmic complexity, tradeoffs, market priorities, DRY, YAGNI, edge cases, BATNAs, etc etc? Well... er... maybe not so much. But his/her eyes just light right up when you mention you have a foosball table and a keg in your cubicle-or-overturned-door "office"!
I don't see how those two points are related. A measure can be good or bad independent of who you choose to measure by it.
Github(not necessarily just having an account but actually looking at what they have done) is a good measure of skill because it allows you to see the person's work output. You know exactly what they contributed to a project and what type of work they are interested in and are familiar with.
Now just because it is a good measurement of skill doesn't mean that it is a good hiring filter. The two are similar but not the same thing. A good hiring filter filters for skill but doesn't filter for other attributes like race and gender. It is kind of like saying measuring how much weight a person can lift is a bad measure for determining a person's strength because we only measured white guys. It is a good measure of strength, just misapplied if you are trying to find the world's strongest person.
That's what we'll get.
The Beats were a genuine movement of talented, hard working literary types. They were hip. They were cool. They were who everyone wanted to be, so people started emulating their dress, their mannerisms, their speach.
But they were not driven by the same passion. Their motivation was different.
Ginsberg said of the Beatniks: "If beatniks and not illuminated Beat poets overrun this country, they will have been created not by Kerouac but by industries of mass communication which continue to brainwash man."
If we let companies continue to use GitHub as a hiring tool, the OSS community will gradually become overrun by people attempting to achieve the same success as the OSS pioneers by mimicking their actions...but lacking their same motivation.
It's hard for me to say this. I like GitHub. I like Open Source. I am proud of what little I have managed to contribute to that community, and I think it is to my advantage if a company considers my GitHub profile when making a hiring decision. That said...I'd rather give up that advantage in exchange for keeping GitHub as an open community to share source, instead of a cauldron for the latest start up founders to pluck their next engineering hire from.
The people ahead of the curve have to be more motivated, but if we get more people sharing, more eyes to find and fix bugs, more people from other walks of life that understand a bit more of the hacker mentality, more acceptance of sharing code as a normal part of life, then the world is a better place.
Yes, it's a shame that not everyone who plays the piano has the dedication and vision of Rachmaninov, or that everyone who paints is Picasso, but I think the world is a better place for amateurs and dilettantes. Sure, we'd perhaps have better art if it were restricted to those who must, but we have a better society because those who are moderately interested can.
It's not a good thing. Submitting proper bug reports is hard, and contributing to an existing project without making more work for others can be hard as well.
The best software does the most with the least amount of code while still being easy to read--and when a library or app gets there, it's time to stop.
The problem with a hypothetical swarm of people trying to pad their Githubs (the same way premeds pad their CVs, for example) is that they'll flood otherwise stable projects with garbage, or start rewrites for no other reason than they think they can do it better. A lot of good projects are going to get fucked up this way.
It's worth noting that James is a UK based developer. Most of his points may be completely valid in the US but they are completely at odds with the UK climate. I've worked with an inordinate amount of tech companies in the UK and consulted with many more and I've yet to come across any that hire based primarily on the merits of a candidates github repo.
In fact, I'd argue his understanding of the request for a candidates github URL is completely wrong. One of my first requests when speaking to a potential candidate is for their github/bitbucket profile so I can see some actual code examples. I know that it's rare to find a true indication of someones ability from just their github contributions but for example, it can help me discern whether a ruby candidate adheres to ruby best practices or not. A small example of a minor indicator that someone might be a good fit for our company.
A man with significantly more experience in the industry than me once put it better than I ever could:
The more information I can get, the easier it becomes for me to make a decision, and to feel comfortable in it. And the best sort of information to give me is to show me what you can make, whether that's code for a backend position, or interesting design/UI for a front end position. It also gives us another thing to talk about during interview; and more opportunity to ask about design decisions and rationale.
Some of the best people I've ever sourced didn't have any publicly accessible code but if they had, I would have found them a hell of a lot quicker.
The problem with this is that you can't really make a logical decision when the code (on Github/bitbucket) is removed from its context. Maybe none of the code follows best practice, but there is a compelling reason why this is so. If your decision on whether to pursue a candidate or not does not include a discussion with the candidate about their available code, you are missing an integral piece of the puzzle.
You are absolutely right. I can honestly only ever think of one circumstance where I dismissed someone based exclusively on what I saw in their github repo. I felt guilty so I phoned them a week later to discuss my thoughts and it turns out my suspicions were correct but that was categorically an edge case.
There are two ways to read this: (1) as a marker; (2) as a filter. Would be interesting to hear your thoughts on the interplay. My guess is that there is some game theory on signalling at play on both sides here.
If you're constantly tweeting/blogging good quality content about an area of technology that is relevant to my business then of course I'm going to take notice. That, however, doesn't necessarily make you a better candidate by default, it just means you were easier to find.
1) Motion detection algorithms for security cameras in outdoor environments with a lot of ambient noise that needed to be ignored.
2) Medical informatics data model with a recommendation engine component for predicting which ICD/CPT/SNOMED code to use for optimal chance of insurance coverage.
3) R&D, implementation, and deployment of an EAV/CR [2] open schema based analytics platform using the neo4j graph database for persistence
So far I've had the luxury of only using meatspace networking to find interesting jobs to work at, but I stress over suddenly finding myself unemployed in the future and having to face the modern interviewing process with social networking activity used as a screen.
Can anyone describe how one's GitHub is used in an interview? Is a candidate with an interesting, non-trivial personal project on GitHub considered, or are candidates judged on their contribution to others' projects?
[1]Started working on a neo4j data science project involving gathering, refining, and translating film, actor, and director data into a graph data model for efficiently answering deeply recursive queries.
[2]http://en.wikipedia.org/wiki/Entity%E2%80%93attribute%E2%80%...
It isn't much different than how a conversation would go after a live coding exercise to review the work, with the exception that the candidate in the case of a prior GitHub work will have had the advantage to clean things up and perhaps has gone through several iterations already.
I realize this is anecdotal, but 95% of the public code in my git repo is stuff I did as fun. I didn't care about being right, or care about memory issues or being thread safe or anything other than 'I wonder if...'
In that instance, and I suspect a lot of other people public where they aren't contributing to something using the repo code as a talking point is counter-productive.
Why did I do it that way? Because I wanted to see what would happen.
What the fuck am I reading? Is this asshole author trying to inject racism into something that is inherently race-, age-, gender-, wealth- and location-agnostic? Github doesn't care who you are, what you code or in what language you code it. And yet here we are with a dumbfuck author who, for the sake of clicks, just absolutely needs to shit on white men, that have been the bane of all tech-interested non-white men since the dawn of time.
When I hire I ask for code samples first and the person's name last. Shit, I'd hire the person based purely on code, put up anywhere on the net, and an anonymous chat, conducted fucking anywhere, and only require their name and social security number when signing the contract.
Shit I'm pissed off now. Some people just can't accept that technology is the ultimate equalizer and instead have to project their white guilt into everything.
The point is that those who have the free time and opportunity to contribute to OSS projects are more likely to be of some means. Given the economic realities of this country, those with some means are statistically speaking far more likely to be white males than anything else (which is confirmed by the data).
The conversation that has been occurring recently is about how having a Github profile is now becoming a de-facto must-have because hiring managers are increasingly seeing it as a shortcut filter.
It's great that you would hire someone based on code put up anywhere on the net, but that's not what people are commenting about. People are saying that Github should not become mandatory, for many reasons, among them because you're limiting your pool of candidates to those with some means.
Is that a concern when hiring? Isn't the biggest concern false-positives? Hiring bad people is the worse hiring mistake. The argument around github seems to be that while the true positive rate is very high, there is also a high false negative rate (rejecting qualified people because they aren't on github). From a hiring perspective (purely arguing economics), that is an acceptable trade-off. Your upside is limited by a smaller pool of applicants (that are still qualified), but your downside is greatly smaller.
It is not about donation, it is about cooperation and net efficiency, and a good share of open source maintainers and contributors are very well paid to do what they do - they are not hobbyists donating weekends to the cause.
Maybe your github profile isn't a good portfolio by default, but it's an excellent service for creating your portfolio. On top of that, even a poor github profile provides more signal than most CVs I've read.
Yeah... that's not what they mean, and you know it.
Whether it (your CV link, internally — extracts data from github? — or externally) points to your github profile or not is your decision.
Why is this still being discussed ?
Websites, people. Websites.
Show me that you take the time to register a domain, configure a host, log in to that Linux box, and build the Web.
A more pertinant question is: Why would you hire a Web Developer who has no place on the Web? It has to be Your Domain, Your Place on the Web. You. What genuine rebuttal is there against this? (There has to be a pretty Good Reason to hire a Web Developer who does not have a Website.)
Forget the "online portfolio" debate; build the Web and actually be about that.
I will not put my real name online in a public forum. Especially not linked with my CV that has a lot of personal information on it. I am afraid of harassment and stalkers, and I really want to control the amount of personal information that gets out there. I don't have a LinkedIn for that reason. I know a lot of that can be found if someone is digging hard enough, but I am extremely uncomfortable with just anyone finding my CV with a Google search. I am really uncomfortable with my work history being public information. In theory I could make that information password protected I guess.
Does it limit me? Probably.
My ideal, if you search for my name in a search engine and NOTHING comes back.
I want to be fair, but "do you have work you can show me?" is about as close to ideal as I can think of.
I mean, if you had a Ruby job opening and you got applications from Aaron Patterson and 20 other people, would it be unfair to put Patterson's application at the top?
There might be another applicant as good, but 1) you have to look harder and 2) you can't really tell until after you've hired them.
Picking a sure thing over a gamble is pretty understandable.
Honestly, those skills are very valuable indicators and do not require someone to be giving up their nights and weekends to OSS.
Open source is not about donation, it is about cooperation and cooperation makes economical sense for a lot of business. Yes, there are the free-riders, but they are the copycats, we are the ones setting the pace.
Startups tend to be pushing the envelope, people from all genders from all over the world are being PAID to fix bugs and implement new features during the business hours, not over weekends. I do, I'm Brazilian and not white by north American standards, so this whole "github bias" thing is BS based on a naive assumption.
I recently went through a job change, and to enhance my chances, I spruced up my Github profile, and put a link to it at the top of my resume. It's not the most active profile, but all the work there are wholly self-initiated projects, all of which tackle some interesting, often original problem. One of the projects randomly ended up with a few stars without me posting that link anywhere. Having code out there would give me a little leg-up over the intense competition, right?
Of the dozen or so tech phone screens, not one asked about it. Only one asked me to talk about a personal project, and it was abundantly clear he had not looked at the Github link.
At least he asked. One phone screen was with a recruiter at a famous open source company, and even she had not looked at it. To be fair, it was a recruiter call and not a technical one, but I thought she'd at least acknowledged it.
Most of the phone screens turned into onsite interviews, and again, nobody even mentioned Github. Well, it did come up in one interview, where the interviewer was doing a deep dive into IP networking details, and I told him in an offhand manner that, y'know, I had code out there that did exactly what he was asking...
The companies I interviewed with ranged from a couple of startups to a bunch of US technology giants in a variety of fields, and many of them get discussed on here regularly.
My advice is, don't put too much stock in your Github profile to help your job prospects. For whatever reasons, maybe the same as TFA's, practically nobody cares. Instead, invest in practicing the same old Google/Amazon/Facebook whiteboard-a-problem style that everybody and their grandmother uses, even if the work involved is nowhere near the kind of work (some people at) Google/Amazon/Facebook do.
I have a hard time thinking of a better way to see how good someone is at all of those those thing than to look at code they've written in the past.
Also, the title of the article is "Why GitHub is not your CV" yet all it talks about is how it can't replace an interview. Which I don't think many people advocate.
As for the main argument, I think it's flawed. Github (or any other online repo) is an important asset and Open Source contributions matters. Especially if you don't have prior job experience. It's a very good way to show that you can actually write code and you will be hands on from the first week, or day, or hour.
Nobody is using Github as their only hiring filter.
On the other hand, Github is just about the best generic indicator of useful contributions of a coder back to society: visible code that someone else found useful. If someone wants to provide whomever is interested with more custom evidence of their work, that's great, and there is a way to link to your homepage for that. But Github stats do not lie, and the constrained profile format is a feature, not a bug.
willing and able to work for free
Me working for free for an employer that tells me what to do does not lead from me working in my spare time doing what/how I want.
I'm not sure I really understand what the author is trying to say. Is it that OSS encourages a culture of being exploited and that GitHub is inadvertently pushing that culture? Or is it that recruiters don't know how to assess the code they see and this is somehow GitHub's fault? Or is it that working on code on your spare time is really a bad thing? Or maybe it's that only white men create GitHub accounts.
It produces false negatives because lots of people who are good coders and who deserve to be hired don't have impressive github accounts for various legitimate reasons. I completely agree with this point.
However, to employers, false negatives are not that big of a problem. In some sense it's morally unfair that I exclude so many people who "deserve" the job, but if in the end I land a satisfactory candidate, I'm happy.
So to employers, the real question is, does github produce false positives? If my screening practice (I think we all agree that githubs and resumes should be used mainly as a screening step) is to ask for githubs instead of CVs, do I let better candidates through? If you read the article carefully, there is very little evidence presented for this; the author mostly argues that using github produces false negatives or weak signals.
Well said. This is also valid for stackoverflow, where your "top" questions are often silly ones that everyone can understand, while the ones you consider valuable are buried in anonymity. Some sort of showcase feature would be nice, in both cases.
- How well can you do this work? - Will you blend in (can you endure our work environment)? - Is it easy to work with you (you are not a dick, are you)?
So get over it, everything you do online IS more real than your resumé. Why? Because we are an online company and your online reputation is more important than a single document traditionally written in a corporate mumbo-jumbo style where the author tend to exaggerate his accomplishments.
When I ask for your github profile, I'm not asking for free work - FOSS is not about free work. I want to see your pull requests. Why? Because from time to time we do pull requests, and do it not because we are good Samaritans but out of sheer greed.
We use an open source technology stack. We push it's boundaries, we find bugs, we fix bugs. We find components lacking, we implement the desired feature and contribute back. We port code from other languages or create it from scratch and open source because it is cheaper than maintaining a closed source library. Is it free work? No, it is cooperation, and it makes sense economically.
Do you have a patch accepted in a high profile FOSS project? This means:
- You are doing serious work based on FOSS stacks - You have some communication skills - You have coding skills, can write tests and documentation - You can fix bugs and implement new features
I want to see the issues you open: are they well documented, with reproducible test cases?
I want to see your stack overflow profile: do you write good questions and good answers? Because a good question is 90% of a good answer, and good answers is what our customers demand.
Do you write really odd blog posts or troll around in online forums? Sorry, I guess you will not fit our work environment.
I had joined several group projects, and started one of my own, back when I was first getting into the job market. It helped a lot to tell any employer that you had been fiddling with this new-fangled "Linux" and "Apache" and "MySQL" stuff for years, versus the older Unix guys who had been using the same industry tech for 10 years. Really, open source was just like a new fad to older hiring types, and knowing something about it meant you had a different mindset.
But now everybody uses GitHub and it means jack shit.
Github-as-the-new-resume is galling for plenty of other reasons.
Today someone told me to review someone else's CV. My last comment was: "This CV does not mention a github account. This either means that this person does not have it or that he doesn't think of it as relevant enough to be on his CV". It was not a deciding factor, but it was listed on the "cons".
There's many good things about it, and reasonably few downsides to trade off against, but overall if you've got a GitHub I'd be more interested in seeing that than knowing what grades you got at school 15 years ago.
That said, my own presence on Github is pretty anemic, so maybe that's another form of bias!
I cannot say that I agree with much of the rest of the article, and I agree with even less of the article on which it is based. But summarily rejecting non-GitHub candidates really is a bad idea.
This has traditionally been important in hiring people for creative work, though prior to GitHub and similar sites there wasn't necessarily a single particularly good/common way for this to be done in the software development field.
Forgive my naiveté, but why is this a topic of debate now? Did something radically change? Weren't you all doing this before Github existed?
Those in hiring positions need to use various indicators of what would make someone a good hire. For most of history in the tech world, this simply meant a standard resume/CV, followed by a phone screen and then an interview where the candidate may be asked whatever questions are deemed relevant by the interviewer. Could be tests, exercises, etc - depends on who you were talking to.
Fast forward to today, where we have more ways that a candidate may show 'indicators' of talent. What about participation in user groups and meetups? Anyone can go to these things, but people who choose to go to them may have a bit more curiosity or interest (and admittedly free time) than those that don't, and that curiosity often goes hand in hand with talent. Not always, but again it is one indicator.
A healthy Stack Overflow reputation score might be another indicator. People who know nothing about programming probably won't be able to rake up major points there.
What if someone wrote a book about a technology? Another indicator probably, and if the book became a best-seller that would be even a stronger indicator since others are judging the material as worthy of their money.
As the author points out, all of these things take time, and many in the industry don't have that kind of free time. Understood.
Experience working at a known entity with a high barrier to entry is another indicator. We know that if someone passed the grueling interview process at certain firms, chances are they will get past our process as well. Another positive indicator.
There will also be some false positives. Candidates that belong to several meetups and have very active GitHubs may not be able to code.
Someone who has never heard of GitHub (or say Node.js or Mongo or whatever may be current and newsworthy at the time) will probably be given a negative indicator Would you consider hiring someone who had never heard of these things? Perhaps not.
Candidates without families that we might expect to have more free time may not choose to spend it at meetups and building GitHub repos either. I don't think we should immediately assume that they are less qualified than the ones that do, but I don't think any will assume they are more qualified.
The author says "you can't judge code without talking to its author". Perhaps you can, but I don't think anybody is necessarily suggesting that you should. No one is hiring candidates based on their GitHub activity alone without interviews, just like no one is hiring anyone based on their CV alone without interviews.
Can we just agree that all of these positive indicators are just indicators? As long as we don't use the absence of them as a negative indicator (as a measure of fairness to those who lack the time or desire), we are not doing anyone a disservice.
I concluded it biases to: folks who most have the free time and the mindset and the kind of desperateness to be doing it. Or income-increasing ambition. (Like the kind that's strongest when you're unemployed or under-employed.) Therefore, it biases (again, only statistically) to young males around college or high school age. Most likely living in their parent's house, or in a dorm, or perhaps out in their own place but without a spouse, kids, sick family members, or any significant hobbies other than computers. And without a long established work history "out in the real world" and thus lots of fellow coworkers/clients/managers to help keep them fully employed out in that pesky meatspace. And I'm not saying this is a bad thing necessarily. Or that this pattern is universal and without exception. I'm just saying this is the pattern I see in the majority of cases. I didn't sense any inherent pro-white bias, or something pro-male. With avatar icons and code commits and pull requests the Internet truly doesn't know or care whether you're a dog. (Only whether you're a smart enough dog that can interact and deliver.) It was just pro-... a certain set of qualities. If a certain demographic is more likely to have those qualities, then, that demographic will be disproportionately present and active there on GitHub. It's not a conspiracy, just the physics and economics of it.
And there's nothing necessarily wrong with it. But it's important to realize it's at play. And to neither blindly assume that anyone without a GitHub presence, or an active one, or large one, is somehow incompetent/unqualified at software engineering. Or that everybody active on GitHub is a young white male in their parent's basement. Instead, treat each person as an individual. And realize that the entire world hasn't fully converted over to your wonderful new "rule of thumb" about hiring. Yet. And perhaps never will. And that too is not necessarily a bad thing.