The ethics of unpaid labour and the OSS community
November 16, 2013 5:37 AM   Subscribe

"We've somehow been culturally talked into accepting this arrangement, not realizing how businesses are using it to further extract value from us. Businesses are choosing candidates based on their open source contributions, knowing that they are getting more value for less money out of them. These are candidates that will continue to work on things in their free time because it's something they care about and are passionate about. This is akin to not paying someone for overtime." -- Ashe Dryden talks about how unpaid work on open source projects may help sustain inequality and exploitation in the software industry.
posted by MartinWisse (154 comments total) 49 users marked this as a favorite
 
As much as it appears a well researched blog post and there is certainly an element of truth, it verges on a troll. The Open Source world is a pretty broad spectrum that ranges from personal projects to highly paid corporate contributions that are in a sense lost leader for lucrative contracts.

It's a very good movement in the world, and huge, and there are likely abuses in various niches. But the predominantly young professional men that are the primary contributors are generally doing quite well. I've seen very few jobs where open source contributions are the exclusive requirement.
posted by sammyo at 5:56 AM on November 16, 2013 [8 favorites]


Excellent post. It's obvious that companies use this as a way of screening not just for candidates who like to work for free, but, as Dryden points out, those who don't have any significant family caretaking responsibilities and fit into the predominant "hacker" culture. And all this while maintaining an altruistic, anti-establishment pose. Google supports open source! They're doing it for the good of the community!
posted by Ralston McTodd at 6:12 AM on November 16, 2013 [24 favorites]


the predominantly young professional men that are the primary contributors are generally doing quite well
That's Dryden's point.
posted by Ralston McTodd at 6:13 AM on November 16, 2013 [8 favorites]


It's sort like the internship game. In order to be considered a good hire, you need to be willing to do a bunch of unpaid labor to prove yourself.
posted by Elementary Penguin at 6:27 AM on November 16, 2013 [14 favorites]


I use open source as one of many things I look at when interviewing candidates. Ashe has given me something to think about here, which is that lack of OSS contributions doesn't necessarily equate to lack of passion, or skill.

However, if I was hiring a musician, (or maybe composer is a better analogy), I'd want to hear their music and I would not feel bad hiring the one that devoted more of their off hours to practice.

In my twenty years of working in the field, I've noticed quite a strong correlation with regard to the motivations of great vs. mediocre programmers. The great ones almost invariably love doing it, they have personal projects, they write about it, speak about it, and think about it both in and out of work. The ones that are primarily attracted to the salary never quite seem to get the idea of craftsmanship. To paraphrase a friend: "They don't pay me to write good software - I'd do that for free. They pay me to attend the meetings and to let them tell me what to work on."

So I want to screen for that kind of enthusiasm, and OSS is a good proxy, but not perfect. I'd love to have something better and less discriminatory, but it seems most filters have problems. For me the key is to look along a number of axes and try to get context for everything. I always pair program during the interview if possible, but given the context of the interview, even that can give a skewed picture sometimes.
posted by bashos_frog at 6:30 AM on November 16, 2013 [32 favorites]


I'd love to have something better and less discriminatory

One tool that may help you to get more diversity is joblint: Test tech job specs for issues with sexism, culture, expectations, and recruiter fails. Also online at joblint.org.
posted by MartinWisse at 6:40 AM on November 16, 2013 [17 favorites]


Imagine you were evaluating two candidates, both starting out in the field. Equal age, gender, race, etc. If the only significant difference between them is that one spends 10-20 hours a week on open source programming and the other spends 10-20 hours on building homes for the poor and feeding the hungry - who would you bet on to be the better programmer a year from now?

Now, 'better programmer' in this context might not be the same thing as 'better hire'. But often you have to make decisions based on incomplete information, and you are hiring not just the person who is before you, but the person they will be in a year.
posted by bashos_frog at 6:43 AM on November 16, 2013 [2 favorites]


However, if I was hiring a musician...I would not feel bad hiring the one that devoted more of their off hours to practice.

Software is indeed becoming a "passion" field like music rather than one where workers are treated as professionals. However, you don't hear much about how we need more students to go into debt getting music performance degrees because there is a desperate shortage of oboists.
posted by Ralston McTodd at 6:47 AM on November 16, 2013 [9 favorites]


I guess the "all software needs to have a mandatory legally-supported warranty" trial balloon still isn't floating high enough, but maybe "giving an open source programmer a day job is a violation of overtime laws" will catch on.

It's a mere coincidence that good-intentioned rules designed to help exploited consumers and workers would also incidentally create new barriers to entry for smaller would-be competitors to huge corporations. Funny how often that happens.
posted by roystgnr at 6:51 AM on November 16, 2013 [11 favorites]


She isn't trying to discourage supporting open source - she is trying to say there is an issue with using it to judge people. There is an issue with making it a job requirement as well. Just as much as if they require me to do any other activity in my off hours.
posted by bashos_frog at 6:54 AM on November 16, 2013 [1 favorite]


a "passion" field ... rather than one where workers are treated as professionals

"Professional" fields all have the concept of 'malpractice' - I don't see ever making that work for software, even though I agree that it is actually a thing that happens quite often.
posted by bashos_frog at 6:58 AM on November 16, 2013 [2 favorites]


A professional certification process complete with malpractice may not be a good fit for software, I agree, but what happened to the concept of the skilled professional who goes to work every day and does his job with the utmost pride in craftsmanship, whether he likes it or not? I think that the stereotypical "Millennials", as portrayed in Wall Street Journal articles, believe that this ideal is stifling compared to that of the "passionate", "creative" worker who does his job for fun. I also think that they will discover that they are seriously wrong, probably when they hit their 30s or 40s.
posted by Ralston McTodd at 7:22 AM on November 16, 2013 [8 favorites]


Sadly, I assumed Ashe Dryden was a man, and only just realized she wasn't upon reading bashos_frog's comment. I should probably work on that.

I think using Github profiles to judge a developer quickly has caught on because it is far easier than the deeper evaluation techniques she suggests as alternatives.
posted by ignignokt at 7:25 AM on November 16, 2013 [1 favorite]


I think part of the issue is that software as a field moves very quickly, and companies are generally unwilling to pay for all of the costs of sharpening your skills. They want you delivering features, not improving your market value. But if you do this stuff in your free time, they get to eat your cake and have it, too.
posted by bashos_frog at 7:29 AM on November 16, 2013 [3 favorites]


bashos_frog: " a "passion" field ... rather than one where workers are treated as professionals

"Professional" fields all have the concept of 'malpractice' - I don't see ever making that work for software, even though I agree that it is actually a thing that happens quite often.
"

Isn't that what errors and omissions insurance is for?
posted by Samizdata at 7:35 AM on November 16, 2013 [2 favorites]


Every article about how software development is so exploitative should come with a big disclaimer at the top that says: Software development is one of the most lucrative fields you can go into. Your standard 24-year-old Facebook/Google/Apple software developer makes $150-200k/year.

Why do we keep leaving this little fact out of these conversations?
posted by tylerkaraszewski at 7:35 AM on November 16, 2013 [7 favorites]


but what happened to the concept of the skilled professional who goes to work every day and does his job with the utmost pride in craftsmanship, whether he likes it or not?

You're conflating two things here: 1) craftsmanship and 2) doing things whether you like it or not.

Craftsmanship is about bringing out the very best in whatever you're working on. It is intrinsically rewarding.

As for 2, I know there was a dominant assumption at some point that whatever your company asked you to do was right, whether you liked it or not. This has caused many problems, both in terms of productivity and ethics, and so, I think people—of all ages, not just millennials—have abandoned it. At a micro level, approaches like kaizen and agile, which despite having plenty of their own faults, have resulted in a lot less doing what you're told and polishing turds. At a higher level, yeah, if someone is no longer interested in high-frequency trading, they'll up and quit instead of sticking it out.

This is better for everyone.
posted by ignignokt at 7:36 AM on November 16, 2013 [2 favorites]


150k isn't standard for a dev by any means.

Also, we "millennials" have a lot of different life goals. One thing that is constant, though, is that to get hired you need to make noises about "passion".
posted by vogon_poet at 7:37 AM on November 16, 2013 [13 favorites]


Your standard 24-year-old Facebook/Google/Apple employee make $150-200k/year.

Why do we keep leaving this little fact out of these conversations?


Because the "standard 24-year-old Facebook/Google/Apple employee" form only a small part of the entire software industry and many more people make much less. Most IT jobs pay decent but not spectacular salaries and don't make you a millionaire in your twenties, unless you're lucky.

The focus though on those who do get paid the big bucks is part of the whole "passionate"/"superstar" complex of the industry, where just being a regular, plodding programmer isn't enough, you have to be excellent all the time, while any hardship you encounter if you're not a superstar is brushed off as the price to pay to get a shot at the big money.
posted by MartinWisse at 7:42 AM on November 16, 2013 [24 favorites]


Your standard 24-year-old Facebook/Google/Apple software developer makes $150-200k/year

What about the standard 64-year-old Facebook/Google/Apple software developer?

As for 2, I know there was a dominant assumption at some point that whatever your company asked you to do was right, whether you liked it or not.

How is the new culture any different, except now you have to act like you're super-passionate about it?
posted by Ralston McTodd at 7:43 AM on November 16, 2013


150k isn't standard for a dev by any means.

In industry hubs like the San Francisco Bay Area, and including equity and bonuses, it is. These are the people who are screening based on github profiles -- the people who themselves are "passionate about software" and aren't working on in-house inventory management systems for trucking companies or whatever.

What about the standard 64-year-old Facebook/Google/Apple software developer?

They make more because they're all managers and execs. Yes, there are fewer of them.

Even if $150k was wrong and it's actually $120k, software development is *still* one of the most lucrative fields you can go into.
posted by tylerkaraszewski at 7:45 AM on November 16, 2013


Most IT jobs pay decent but not spectacular salaries

Most IT jobs aren't software developer jobs.
posted by tylerkaraszewski at 7:47 AM on November 16, 2013 [2 favorites]


Related and thoughtful: Why GitHub is not your CV.
posted by Nelson at 7:48 AM on November 16, 2013 [5 favorites]


How is the new culture any different, except now you have to act like you're super-passionate about it?

It's no utopia, but now you can go leave and work at an insurance company or something, where no one cares how passionate you are. No misplaced belief in thinking you must be loyal to company in order to be a "professional" will keep you doing something that you don't like.
posted by ignignokt at 7:50 AM on November 16, 2013


Your standard 24-year-old Facebook/Google/Apple software developer makes $150-200k/year

That is a horrible abuse of the word "standard." The vast majority of developers live outside Silicon Valley and do not make nearly that much, especially at 24. Yes, it pays well compared to most other jobs a 24-year-old might get. But those kinds of numbers are just not the status quo, even for extremely talented senior engineers, in most of the industry.
posted by Tomorrowful at 7:50 AM on November 16, 2013 [17 favorites]


Every article about how software development is so exploitative should come with a big disclaimer at the top that says: Software development is one of the most lucrative fields you can go into. Your standard 24-year-old Facebook/Google/Apple software developer makes $150-200k/year.

Why do we keep leaving this little fact out of these conversations?


The standard Facebook/Google/Apple developer is a young white man. This post was about how hiring practices, such as requiring OSS experience reinforce that status quo. No one is arguing that software developers are under-compensated, we're examining why these lucrative jobs only seem to be done by the small subset of the population that does all the high-paid high-status jobs.
posted by martinX's bellbottoms at 7:50 AM on November 16, 2013 [13 favorites]


And what keeps these lucrative jobs filled almost entirely by white men? Maybe the article has some ideas about that.
posted by ignignokt at 7:51 AM on November 16, 2013 [3 favorites]


Even if $150k was wrong and it's actually $120k, software development is *still* one of the most lucrative fields you can go into.

This is still a ridiculous argument. Hey, the financial industry is insanely lucrative -- top bond traders at big investment banks make millions every year.

Oh, p.s., your bank teller probably doesn't. But they don't count! For reasons!
posted by indubitable at 7:51 AM on November 16, 2013 [16 favorites]


I'm on contract currently at a software/web agency that offers paid overtime. The staff is mainly 30 and 40somethings with kids and houses at least an hour away. There's very little late night work, as far as I can tell--the social norm is to put in a good eight hour day and go home.

Everyone in the industry I've mentioned this to is deeply impressed, less because it means death marches are paid marches, more because it signals that death marches are to be avoided in the first place, and the company has given itself a strong incentive to avoid them. And with proper planning, it costs the company almost nothing to offer this benefit, which is a huge differentiator to potential hires. Businesses figuring this out is one of the things that will drive change.
posted by fatbird at 7:54 AM on November 16, 2013 [24 favorites]


I'm a software engineer and for many years a manager of engineers. I like it when people have contributed to open source during their "free time" because it shows they have a passion for software development even when they aren't being paid to do it. It has never occurred to me, that by choosing someone with that experience that I was somehow trying to get something or some value for "free". My only aim is to get the best person for the job, and open source involvement is a small part of that consideration. Also, I'm not quite sure why the article takes a shot at meritocracy. I should hire a worse candidate because this person came up with less free time or money?
posted by ill3 at 7:57 AM on November 16, 2013 [3 favorites]


This is still a ridiculous argument. Hey, the financial industry is insanely lucrative -- top bond traders at big investment banks make millions every year.

Oh, p.s., your bank teller probably doesn't. But they don't count! For reasons!


That's not analogous. "Software Developer" is a specific position. "Financial industry" isn't. The janitors at Google don't make six figures either but they're not software developers.
posted by tylerkaraszewski at 7:57 AM on November 16, 2013 [4 favorites]


I should hire a worse candidate because this person came up with less free time or money?

I think the point is that someone without umptyzillion OS projects is not necessarily a worse candidate.
posted by rtha at 8:03 AM on November 16, 2013 [18 favorites]


Like bashos_frog, I found this challenging because we do look for Github repos or other open-source projects in resumes. We don't require them, and I suspect that few employers do.

What proxy isn't subject to privilege considerations though? I'd much rather look at open-source contributions than only hire kids who went to Stanford or MIT. Some amount of pre-screening is necessary; I've gotten 70 resumes for one senior JavaScript developer position.

As a woman who runs an engineering team, I absolutely want to encourage diversity in our group, and I'm glad this issue was raised because it wasn't something I'd considered. But I'm not sure there's a more fair way to evaluate developer resumes; Dryden's proposed solutions don't solve the pre-screening problem, and starting people as contractors isn't helpful and usually runs afoul of employment law, as one commenter points out.
posted by nev at 8:04 AM on November 16, 2013 [7 favorites]


> I like it when people have contributed to open source during their "free time" because it shows they have a passion for software development even when they aren't being paid to do it. It has never occurred to me, that by choosing someone with that experience that I was somehow trying to get something or some value for "free".

And apparently it still hasn't registered with you. The point is that the only people who can afford to spend time doing open source work are the people who are already affluent - overwhelmingly white males.

I have a personal example here. I recruited a young man to help me with my open source project. He's very smart and really interested - but he's also from a poor background, so he had to spend a great deal of time at his day job at a Chinese restaurant, where he's already a manager (at 19).

When I found out that this "managerial" position pays him a princely $9.73 an hour, I offered him $100 a week to drop one day of work and dedicate it to my project. He jumped at this... asked me several times if I were serious. Of course I'm fucking serious - I wish I had more money to offer him but I'm doing this just because I love the idea. And I have another goal - to get him out of the stupid restaurant job, because he deserves so much better.

And I honestly think he'll do it - because he's going to have a good track record. BUT if it weren't for the fact that I can afford to just blow this money on him, he would NEVER be able to find the time to make a serious contribution to open source, and since he's not going to be able to afford to go to school, that would have been the end of it.
posted by lupus_yonderboy at 8:06 AM on November 16, 2013 [36 favorites]


And what keeps these lucrative jobs filled almost entirely by white men? Maybe the article has some ideas about that.

It might. My point is that in these discussions nobody bothers to mention that the jobs we're talking about here are constantly rated among the top in "most lucrative careers" lists, and that may be material in the discussion.

Someone else said:

No one is arguing that software developers are under-compensated

Which is true. We were silently ignoring their compensation, which is generally high, entirely.
posted by tylerkaraszewski at 8:08 AM on November 16, 2013


Which is true. We were silently ignoring their compensation, which is generally high, entirely.
No, we weren't at all. We agree that software development is a great job, if you can get it. Which, for many reasons, seems to be much easier if you are a white man from an affluent background.
posted by martinX's bellbottoms at 8:11 AM on November 16, 2013 [6 favorites]


It might be a bum deal for some developers, but there is the other side of the coin...open-source software is a goddamn miracle for under-resourced users. If it wasn't for open-source software, only rich people could put up web servers, learn GIS, set up a VOIP network, geocode a million addresses, visualize data, you name it.

On balance I think that the presence of high-quality open-source software is reducing the burden of the unprivileged.
posted by Hollywood Upstairs Medical College at 8:19 AM on November 16, 2013 [18 favorites]


While it's clear that a lot of the software industry in America is a white guys niche, I've never seen any conclusive explanation. I've certainly worked with great female and non-white-guy programmers, the skew seems constant. Why are women not going into tech? Babies? Abusive environments? Preference to "helping" jobs? Life balance preference? The open source skew is perhaps a symptom, but the mystery of "white programmer privilege" I think remains something of a mystery.
posted by sammyo at 8:23 AM on November 16, 2013 [1 favorite]


Why are women not going into tech? Babies? Abusive environments? Preference to "helping" jobs? Life balance preference? The open source skew is perhaps a symptom, but the mystery of "white programmer privilege" I think remains something of a mystery.

I can speak to math more than programming, but many girls are pushed away from math long before babies enter into consideration (and if that's the reason, it's a structural problem we should obviously be trying to fix). They're taught they should prefer 'helping' jobs, regardless of whether they'd make an awesome mathematician or programmer and like doing that better.

My high school math team was dominated by boys, but the handful of girls were, on average, better (at both math and the skills required for math team, which isn't a perfect overlap). But the mediocre boys stuck around basically for their egos--as (nerdy, unpopular) boys, they were 'supposed' to be good at math, so it never really occurred to them that they weren't that great. The girls who would have made equally mediocre math team members never turned up in the first place--if they were socially isolated, math team only would have made it worse, and, besides, no one was telling them they ought to be good at it, so it never occurred to them that they ought to try it out, despite the recruitment efforts of teachers. (Same pattern with chess, really, though the chess team was generally immature, which was probably off-putting to girls, who I think tend to mature socially faster. Same with the computer programming classes, which were more evenly balanced gender-wise, but the egos of many of the boys were apparent.)
posted by hoyland at 8:33 AM on November 16, 2013 [7 favorites]


I disagree that OSS contributions done on my own free time are labour. I do create value, which is sometimes useful to the commons and sometimes only to myself, but it's not captured by any person or entity and I have almost complete freedom to pick things that are challenging and/or useful to me. The article's claims on diversity may be more central to its thesis, but the title and the quote about overtime are a bit of bait that's not supported by the rest of the article.
posted by Tobu at 8:36 AM on November 16, 2013 [7 favorites]


As someone who works on and contributes to OSS as part of $DAYJOB this gives me some things to think about. Interesting piece, thanks.
posted by scatter gather at 8:46 AM on November 16, 2013


While it's clear that a lot of the software industry in America is a white guys niche, I've never seen any conclusive explanation. I've certainly worked with great female and non-white-guy programmers, the skew seems constant. Why are women not going into tech? Babies? Abusive environments? Preference to "helping" jobs? Life balance preference? The open source skew is perhaps a symptom, but the mystery of "white programmer privilege" I think remains something of a mystery.

I just... I don't even... sammyo, I will assume you're coming here to ask that question with the best of intentions. There are vast swathes of the internet containing thoughts on exactly that question. Indeed, no "conclusive" answer, just like there is not really a "conclusive" answer to "why is there racism?" There are a ton of factors at play, some perhaps biological, many obviously cultural. This thread is not the right place to go through that conversation, but MetaFilter alone has many past threads that have delved into the different aspects of it. I encourage you to read those and come up to speed if you feel you are not already.

The topic of this thread, IMO, is both a symptom and one of the many cultural factors of the technology industry culture that contribute to the dearth of women, minorities, and underprivileged individuals in the field. As a woman working in the technology industry I think it's important to look at powerful influences on how employees are selected -- such as unwritten "requirements" that are often only relevant to a small subset of privileged people -- and evaluate what can be done to make that process more fair. I think that's what this thread should be discussing, particularly given that we've already seem some hiring managers are reading this and re-evaluating their own processes. I'd like to continue down that path rather than rehash overarching "women in tech" discussions that have already happened frequently here. There is an obvious issue here: this increasingly popular hiring requirement tends to select for a particular demographic. How do we solve that? Address the hiring requirement, or address growing the diversity of that demographic? (Or both?) Either way, *how* can it be done fairly?
posted by olinerd at 8:57 AM on November 16, 2013 [23 favorites]


I know of no idea, system, plan, political system or other thing that doesn't have benefits and drawbacks. These are competing interests. Generally, best solved with measures that reduce the inequality the system of open software might create, without getting rid of its benefits.

Computer coding seems like a strange economy to me.
posted by Ironmouth at 9:00 AM on November 16, 2013


IME the big sv companies are as Indian and Chinese as they are white. Still 90% men, though.
posted by jewzilla at 9:02 AM on November 16, 2013 [9 favorites]


Ralston McTodd: " Google supports open source! They're doing it for the good of the community!"

I know you're being ironic, but for all Google's faults, Summer of Code is one of the coolest things ever done by a software company. Each year I've seen dozens of young Indian students turn from lost lambs into clued-in hackers in the space of a few months.
posted by vanar sena at 9:03 AM on November 16, 2013 [9 favorites]


Well fuck, I hope contributing to OSS in "free time" doesn't become obligatory, the industry doesn't need any more reasons to be hostile to anyone with any kind of reasonable family life outside work.
posted by Artw at 9:36 AM on November 16, 2013 [9 favorites]


My own fantasy pre-screening bar for candidates is that they must have made Wikipedia edits. I can usually tell if someone has the urge to write high quality code, but I want to know if they're driven to write good documentation!

Any edit: +1 point. Correct spelling and grammar: +3 points. Proper cites: +2 points. A useful contribution that shows they can be consistent with house guidelines & style: +10 points.

More seriously, it is a mistake to require open source contributions. But working on open source is a relatively big positive signal.

When hiring someone I need to see enough code to be able to judge their abilities and taste, but I don't care if it's from an open source project, or their solution to a take-home assignment I've given them, or any place else.

Tobu: Agreed. The diversity & privilege angle is interesting. But I don't buy the idea that companies hire people who work on open source in order to get "free work" at all. That's not how it works.
posted by jjwiseman at 9:37 AM on November 16, 2013


As someone with ongoing contributions to various OSS libraries in the form of code, bug reports, mailing list participation, etc., I have to say that in general, this is certainly a case where people know what the deal is in terms of their expected benefit from the work, but that being said, it is truly staggering, and I mean staggering how much money has been made on the backs of so much free labor by so few highly skilled people. We're talking trillions of dollars and entire industries.

It seems untenable over the long term, but even today, the myth of the truly solo and unsupported OSS developer is growing ever more mythical by the year as more and more companies devote resources to supporting the OSS libraries they rely on. As software ages and its deployments age with it, it becomes increasingly less likely that you can rely on the traditional OSS models for support and progress. This means that at some level of success, you simply must have in-house expertise, and that can't be had on a volunteer basis.
posted by feloniousmonk at 9:39 AM on November 16, 2013 [2 favorites]


My own fantasy pre-screening bar for candidates is that they must have made Wikipedia edits.

Oh god. Mandatory time spent interacting with one of the worst, bitchiest toxic communities on the planet.
posted by Artw at 9:42 AM on November 16, 2013 [17 favorites]


I should hire a worse candidate because this person came up with less free time or money?

I think the point is that someone without umptyzillion OS projects is not necessarily a worse candidate.


Which is why I said open source work is a small part of my consideration.
posted by ill3 at 9:48 AM on November 16, 2013


but what happened to the concept of the skilled professional who goes to work every day and does his job with the utmost pride in craftsmanship, whether he likes it or not? I think that the stereotypical "Millennials", as portrayed in Wall Street Journal articles, believe that this ideal is stifling compared to that of the "passionate", "creative" worker who does his job for fun. I also think that they will discover that they are seriously wrong, probably when they hit their 30s or 40s.

the problem here is that what's wrong when you're in your 30s/40s is not necessarily wrong when you're in your 20s when many/most do not have family and other pressing financial concerns. I mean, I could reverse engineer this logic and say that nobody in their 30s/40s should be allowed near a certain kind of independent creative project because they'd be way too likely to bring other concerns to the workplace. Indeed, I've seen it happen -- low budget movies rendered un-produceable because a union got involved and started making demands that the producers couldn't begin to meet, and thus they either had to bring in other partners (who homogenized things in order to justify the increased expenditures) or the movies just died.

Which isn't to say that I don't see the union's position. It makes perfect sense to demand enough for workers that they can do more than just eat-sleep-create ... except many a cool and important creative project happens only because people are willing to work/live this way. Seriously, take a hard look at movie/music/art history in general and you'll see all manner of essential works that only happened because young people (free for the time from concerns other than their own immediate survival needs) did their jobs "for fun".

And I'm damned glad that they did.
posted by philip-random at 9:49 AM on November 16, 2013 [8 favorites]


My own fantasy pre-screening bar for candidates is that they must have made Wikipedia edits.

What would this tell you in particular that you could not get from a basic 10- or 15-minute editing test done as part of the interview?
posted by rtha at 9:49 AM on November 16, 2013 [1 favorite]


It's also worth noting that in my experience, "we only hire people with OSS contributions" is just ego-driven bluster that startups spew out to make it seem like they have The Best Technology People Available. This is so vitally important to the image startups portray not just to the market but to their investors that they will go to any lengths to reinforce it.
posted by feloniousmonk at 9:50 AM on November 16, 2013 [3 favorites]


What would this tell you in particular that you could not get from a basic 10- or 15-minute editing test done as part of the interview?

Whether they had the wherewithal to do it without being asked or told.
posted by jjwiseman at 9:51 AM on November 16, 2013 [2 favorites]


It's OK to tell your employees what you'd like them to do
posted by RustyBrooks at 9:53 AM on November 16, 2013 [9 favorites]


Lets face it start ups are generally horrible and exploitative work enviroments that will punish you for not devoting your whole waking life to them, usually on an illusory promise of them making it big one day (they probably won't) and that the money from that trickling down to regular employees (not guaranteed) - best avoided if you want any kind of sane life.

People who glorify that shit are just the worst.
posted by Artw at 9:54 AM on November 16, 2013 [7 favorites]


Sure. I just prefer the ones that feel there's something a little off in the universe until they've written good documentation on their own initiative.
posted by jjwiseman at 9:56 AM on November 16, 2013 [1 favorite]


Your standard 24-year-old Facebook/Google/Apple software developer makes $150-200k/year

Data point here. I was a self-taught software dev, hired through my contributions to open source projects at that age. My first salary in NYC was 50k. I wasn't exactly Google-level quality though, but I don't think most devs are.

I probably could have reached six figures in that industry if I had been one of the people whose hobby AND job is software. A few years in I had lost interest in contributing code as a hobby though and once that happened I kind of figured it was game over for climbing the ladder. I just couldn't compete with people who came home from work and coded, when I wanted to come home from work and hang out with friends and cook delicious food. I felt immensely guilty about it, both because I felt I wasn't living up to my potential and that I was becoming a "free rider," but I could't make myself do it from 9AM to 11PM anymore.

What would I tell a 19 year old woman about getting into software? I might have to be honest and tell them that they have to have a monk-like devotion to the craft and if they want more work-life balance and to enjoy other hobbies they might want to consider another career.
posted by melissam at 9:56 AM on November 16, 2013 [16 favorites]


While I don't think that hiring someone who works on open source is exploitative, I just remembered that Chris Anderson, whose tweet motivated Ashe Dryden's blog post, is at the center of some controversy over open source exploitation.

Anderson is CEO of 3D Robotics, which develops & sells small drones. Their hardware and software is open source, which I think has been a big overall benefit to the field, and they have hired many people who started out by contributing to the software in their free time, but there are a few people who feel it's wrong for the company to be selling products based on open source contributions or that they don't reward the contributors enough.

I don't find those arguments very compelling in this case, but it is a little additional context for this discussion.
posted by jjwiseman at 9:58 AM on November 16, 2013 [5 favorites]


> I like it when people have contributed to open source during their "free time" because it shows they have a passion for software development even when they aren't being paid to do it. It has never occurred to me, that by choosing someone with that experience that I was somehow trying to get something or some value for "free".

And apparently it still hasn't registered with you. The point is that the only people who can afford to spend time doing open source work are the people who are already affluent - overwhelmingly white males.


No, I understand what the article is saying. I just think it's a ridiculous argument. While I don't disagree that the majority of open source contributors are affluent white males, I do not think being affluent has anything to do with being able to contribute to open source. A classic case of correlation does not require causation. I would argue that contributing to open source has a very low bar to step over in terms of wealth to participate. Certainly lower than going to the right university, certainly lower than belonging to the right social clubs. All you need is a desire to contribute, programming skill and an internet connection. I'm sure many will say only the affluent have the free time to dedicate to these types of projects, but as someone that has worked 2 jobs at once and had time to contribute to open source, I have a hard time believing that it's not doable for the average person or even the economically disadvantaged person. In fact, if one adds up what is required to contribute to an open source project, surely knowing how to program is the hardest bit, because that requires time and some source of education. I think it might make more sense to say knowing how to program is a affluent white male privilege, than saying contributing to open source is, or if open source is, it's because this particular programming skill is required. Once you take the programming out of the mix, which IMHO is the real bar one must step over in terms of privilege, then you are comparing those privileged enough to know how to program who have chosen to contribute and those who haven't. As someone who has directly hired probably 100+ engineers, I can tell you in my experience those with open source experience tend to be the better candidates.
posted by ill3 at 10:03 AM on November 16, 2013 [4 favorites]


All the software companies (most based in SV) I've worked for were not overwhelmingly or even especially white in either the head office or my office. Male, yes.

Companies don't expect loyalty so much as 100% of your time while you're their employee. They don't hold anything against a dev for hopping to a new company because it happens so often.

Going into debt for a computer science degree isn't the worst idea. Chances are excellent that anyone with that degree will at least be able to get a comfortable middle-class salary, and pay off the debt quickly.
posted by Pruitt-Igoe at 10:07 AM on November 16, 2013


When I was young and poor, I sure as shinola didn't have the time or energy for contributing to OSS projects, and I only have a career now because of some very unique and unreproducible events. The reason requiring OSS contributions increases inequality is because who has time for that when you're trying to feed yourself? Mostly young, affluent males.
posted by fnerg at 10:08 AM on November 16, 2013 [2 favorites]


It's true that the bar for open source contribution is very low, but in this case correlation is enough: Any hiring preference that reduces diversity is at least worth examining and thinking about.
posted by jjwiseman at 10:08 AM on November 16, 2013


I think the point is that someone without umptyzillion OS projects is not necessarily a worse candidate.

I know that's what some want the point to be, but it's at best misleading, willfully or otherwise. If I have a candidate without an OSS track-record who's objectively superior to one with one, do you really think I'm going to hire the inferior candidate just because of some vague handwaving that I'll get more labor out of them? That's deeply stupid; in fact, my experience is that the opposite is true...the guys with OSS side projects like to wrap the day so they can go work on "fun" rather than "work" stuff.

On the other hand, given 2 candidates that are close, why would I *not* be more interested in the guy who's work product I can look at and evaluate? The other guy might be better, sure, but lacking some other objective measures, how do I know that at the interview stage?
posted by kjs3 at 10:09 AM on November 16, 2013 [1 favorite]


Student 1: go to college, do coursework, go to job that pays for college, sleep.

Student 2: Go to college, do coursework, hack around with OSS stuff because your parents are paying for college, sleep.

Big difference right there.
posted by fnerg at 10:12 AM on November 16, 2013 [28 favorites]


Also some of the narratives about "why care about software devs they make 100K+ etc." fails to take into account the large contribution from foreign developers, often from emerging economies. I was reading about this stipend program for women contributing to Linux and it's $5000. To me that doesn't seem like a lot of money when I could get a freelance contact for a $10,000+ project, but then I looked at the list of women who are getting this stipend and a lot of them come from places where $5000 is a ton of money and also places them in a fairly good position for employment-related visas to richer countries.
posted by melissam at 10:18 AM on November 16, 2013 [4 favorites]


Also, 3D Robotics' concept of open source was stuck in the early oughts (subversion source control, closed mailing lists) until just recently, and I doubt many of the people who were contributing for free and then eventually hired would have even had github profiles, making Chris Anderson's "reject anyone who doesn't have a @GitHub profile?" tweet a little ironic.
posted by jjwiseman at 10:28 AM on November 16, 2013 [1 favorite]


I think any sort of real requirement for OSS contribution, if pervasive enough, will ultimately become self-defeating as the value of any given contribution decreases with the rush to have something, anything, to pass that gate. To be honest, if a person's contributions are mostly docs and small tweaks, that's cool but contributions aren't going to bear on the hiring decision in any meaningful way unless they include something substantial, and that's just not very common for any number of good reasons. Most projects do not require substantial contribution and in fact, most projects are most likely to reject them outright.

We really shouldn't talk about expectations around OSS contributions without discussing how incredibly difficult it can be to interact with an established OSS project without getting torn to shreds by the smarter than thou people in charge or the peanut gallery on the mailing list. If you have a substantial contribution to make to an existing OSS project, getting the code accepted will be harder than it was to write, likely significantly harder, almost without a doubt. It's very easy to talk about making valuable contributions to OSS, but the reality of approaching a project totally cold is very different from the reality of making some contributions to your buddy's project or pitching in on something you found out about at a hackathon or an annual conference where you met the people in charge and established a prior relationship.
posted by feloniousmonk at 10:29 AM on November 16, 2013 [9 favorites]


I hate to be the guy who is here representing the side of affluent white male privilege or sound that I'm arguing for that, as that is not who I am. I think there is a diversity problem in tech, but I think it has nothing to do with open source (if anything open source is a tool for leveling playing fields in general) and has everything to do with education.

One thing I notice is that people in this thread keep saying, "what if the person doesn't have time to contribute to open source, because they have 3 jobs or are paying for school, etc..." I think what people are missing is that often people contribute to open source, do so because it is fun for them, relaxing, a hobby, an escape. I say this because I am one of the people that this is true for. It seems to be described here as only a different way to work. I posit that it's not a matter of people choosing between the second job and open source contribution, but perhaps choosing between open source contribution and watching TV or going to a party. Even the poorest students I've known with a full load and a job paying the bills, had time to watch TV or go to parties. This again sums up why open source contribution is a useful indicator for me, because it shows people who code for the love of coding rather than just a pay check.
posted by ill3 at 10:30 AM on November 16, 2013 [4 favorites]


This again sums up why open source contribution is a useful indicator for me, because it shows people who code for the love of coding rather than just a pay check.

What is wrong with coding for a paycheck? Are people who love it necessarily better coders?

As long as this is a benchmark it will be hard for tech as a field to diversify. Because passion for it often depends not on education but from things you picked up in childhood and those things are often gendered, particularly toys, movies, books etc. that are tech-related.

I think this explains why software is less diverse than hardware within tech because I think in hardware there is less hagiography around people who are passionate about it and also the reality that most people who work installing and supporting it probably are never going to build hardware anyway.
posted by melissam at 10:36 AM on November 16, 2013 [5 favorites]


The companies want to hire people who have a blindspot in self-recognition that allows them to believe in a lotto jackpot payoff to all those with Jobs' or Gates' or Page's abilities which they of course possess. This delusion enables worker bees to stomach eating tons of shit to inflate their employer's profits. Capital extracts surplus, efficiently and maximally. Also: see stock options not worth the paper they are printed upon.

Surely there are better places to work.
posted by bukvich at 10:37 AM on November 16, 2013 [1 favorite]


There's actually a bit of a backlash to open source contribution in certain parts of the valley nowadays. If you look at the last 20 years of the industry, there's really only been one open source related company that's made it as a stand alone entity (Red Hat) - everyone else was either acquired or died. Looking at Red Hat's numbers, most of their scale out in terms of revenue happened via value add software that they wrote - not support and services (which is the ostensible OSS way to still eat).

We're not far off from the whole of the industry realizing that the way to get rich is to USE open source, not write open source (see Google/Facebook/Twitter - though even they devote huge resources to shoring up the open source projects upon which they're built). Once the industry as a whole sees the imbalance in compensation and rewards clearly, the natural reaction is to turn the spigot of free labor off.
posted by NoRelationToLea at 10:39 AM on November 16, 2013 [1 favorite]


What is wrong with coding for a paycheck?

Nothing. I'm a person who does exactly that.

Are people who love it necessarily better coders?

In my experience, not always, but in general, yes.
posted by ill3 at 10:49 AM on November 16, 2013 [4 favorites]


I recently went through a rigorous interview process for a ultra hot startup only to be offered a job paying 22% less than what I make writing bullshit CRMs or whatever at the huge megacorp I work at. I was at one point interviewed by my prospective "team" and I had more years of experience than them combined. They weren't even qualified to ask me to whiteboard stuff.

Even after all that I'm still considering it. I do love it, I've devoted 40+ hours a week for my entire adult life. I been doing this professionally for over 20 years. We going to quibble over the fact I don't program in my downtime?
posted by Ad hominem at 11:02 AM on November 16, 2013 [9 favorites]


Ralston McTodd: "What about the standard 64-year-old Facebook/Google/Apple software developer?"

I imagine Rob Pike and Ken Thompson are well compensated for their work on Golang.
posted by pwnguin at 11:02 AM on November 16, 2013 [2 favorites]


Even the poorest students I've known with a full load and a job paying the bills, had time to watch TV or go to parties.

True. Current students in CS may have few excuses for not having OSS contributions. What about those of us for whom GitHub came too late? I'm not even 30 yet, but the ability of my generation of engineering and CS students to contribute to OSS is very different than the low-bar-to-entry of GitHub et al these days -- so what about those of us who are still too young to be senior engineers but too experienced to be entry level who now have many more obligations than "TV and parties", such as commutes, pets, families, young children, etc; is it fair to expect them to devote significant portions of their free time to OSS just because the tools for involving them in the community are now available? Most of my ~30 year old friends who do have GitHub profiles have them because they work for companies that use or develop OSS, not because they spend their free time writing code.

Having free time to write code is a privilege. I mean, you did see this, right?
posted by olinerd at 11:13 AM on November 16, 2013 [9 favorites]


melissam: "I was reading about this stipend program for women contributing to Linux and it's $5000. To me that doesn't seem like a lot of money when I could get a freelance contact for a $10,000+ project, but then I looked at the list of women who are getting this stipend and a lot of them come from places where $5000 is a ton of money and also places them in a fairly good position for employment-related visas to richer countries."

Yea, $5k is not very attractive to many US or European students, so GSoC mainly functions as a price floor for those people. So while GSoC seems on the surface to be a way to fund student developers to get more involved in their favorite open source project, it ends up being more of a recruitment tool for developers in developing countries. My mentoring org ends up fielding a number of proposals from people who've never used our software, and don't necessarily have github profiles, or even passing knowledge of git.
posted by pwnguin at 11:24 AM on November 16, 2013


Clearly what is needed is for Amanda Palmer to hold a benefit concert to raise awareness for this issue.

(I keeed.)
posted by radwolf76 at 11:26 AM on November 16, 2013 [1 favorite]


The article makes a good case is that OSS contributions are affected by privilege, and are not actually pure indications of merit. However, this is also true for basically any signal a prospective employee may try to send to employers. Does your family have sufficient emotional resources so that you can spend your time in school comfortably learning instead of anxiously dealing with trouble? Privilege. Does your family have sufficient financial and social resources so as to develop your niche academic and non-academic interests, thereby paving the way to your admission to an elite institution of higher education? Privilege. Does your family have the social capital so that you can present well and "fit in" in an interview at a company where most people are upper-middle-class? Privilege. Do you have the resources so that you can develop a "passion" for something? Privilege. Did you have the connections/academic credentials to get a job at a highly recognizable/prestigious firm early on in your career? Privilege. Etc etc etc etc etc. The whole world runs on privilege.
posted by leopard at 11:30 AM on November 16, 2013 [16 favorites]


My Github profile is empty of public projects. If someone wants to use that to not hire me, they are welcome to hire someone else. Same with "pair programming during the interview." If you want me to sit in a room full of people and do ACTUAL WORK ON YOUR PROJECT- that is, if you want free labor during an interview and fail to grasp the simple human common courtesy of having a private conversation in a quiet room, fine: I will walk out and never look back. There are tons of jobs.

What most of these trendy tech interview techniques test for has nothing to do with programming, and not even that much with who will work for free- they test for conformity. The tech community runs on absolute groupthink. It's not that people don't want diversity as far as race, gender, etc -- though that is a huge problem. No, they genuinely don't want people who think for themselves. How else is something like Snapchat going to be valued at greater than THREE BILLION DOLLARS with a straight face? How else can something as mind-shatteringly moronic as "Coin" (self-link) come out, make one video of a white man walking down the street, and have people lining up to pre-order it? This is where you get not giving people jobs because of what kind of phone they've personally chosen. This is where you get awful programming choices made to accommodate the latest buzzword (hi Backbone.js!, hi Coffeescript!) that is a complete waste of time and does nothing useful except 1) help the programmer survive the inevitable nerd dick-measuring contests (What? You don't want to use that new framework? CAN YOU NOT FIGURE IT OUT?) and 2) Get that buzzword on the person's resume so he can get the next job with the next group of idiots who put that buzzword down as a requirement because a blog said it was "hot."

For something that's allegedly a "science," the lack of critical thinking is staggering.
posted by drjimmy11 at 11:31 AM on November 16, 2013 [8 favorites]


fnerg: "Student 1: go to college, do coursework, go to job that pays for college, sleep.

Student 2: Go to college, do coursework, hack around with OSS stuff because your parents are paying for college, sleep.
"

Student 3: Go to college, do coursework, get a job that pays for college hacking around on OSS stuff, sleep.
posted by pwnguin at 11:34 AM on November 16, 2013


Over the years, I've gradually gotten less and less patient for people who think software development is a pure meritocracy. In my experience, it tends to go hand in hand with successful people who believe that their success is entirely due to their hard work and has absolutely nothing to do with actual luck - anyone else could do what they did, regardless.
posted by rmd1023 at 11:43 AM on November 16, 2013 [8 favorites]


Further, the real idea seems to be something akin to "If you don't have anything to hide, you wouldn't mind being spied on." If you're a "good programmer" and "one of us," why would you mind being asked for more and more invasive things?

The idea is not so much that having Open Source code samples on Github* gives you more information about the candidates that have them, it's more trying to get everyone to do it so that if anyone does NOT have these samples, there MUST BE SOMETHING WRONG WITH THEM and they should never be hired under any circumstances. The same goes for LinkedIn- it's a loathsome, utterly useless, invasive piece of shit that openly encourages discrimination by more-or-less forcing people to post photos with their resume. But they got enough people to use it to tip the balance to, "HE DOESN'T HAVE A LINKEDIN? WHAT IS HE HIDING??"



*Note again the downright obsession with stupid trends and minutiae. I have been emailing people code samples for years- naturally if you want to hire a programmer you want to see his code. But suddenly these are worthless because they're not part of a certain ideology and posted on a certain website.
posted by drjimmy11 at 11:44 AM on November 16, 2013 [8 favorites]


> If you want me to sit in a room full of people and do ACTUAL WORK ON YOUR PROJECT

I find it nearly impossible to believe that any company, no matter how bad, would actually get interview candidates to write code which they expected to be able to use afterward.

Just the time it would take to get up to speed is too great - the coding standards issue - too many reasons to count.
posted by lupus_yonderboy at 11:53 AM on November 16, 2013 [2 favorites]


I think there are a couple of things that contribute to this kind of hiring practice. Firstly, people in skilled professions are naturally biased towards people who came up through the ranks the same way they did. If you've contributed to OSS and feel like that is part of your success, you are naturally going to look to duplicate that success with the entry level people you hire, regardless of whether or not there is any truth to the idea. Secondly, and more importantly, the profession of software development is undergoing a process of rationalization not unlike what medicine went through, but no one really wants to admit that software is still in the stage where barbers are cutting your hair one day and performing surgery on their neighbors the next.

On top of all of the other swirling reasons why this is a complex subject you arrive at one fundamental problem: because of the changing nature of the profession, you can't even take two jobs with the same title at the same company and confidently say they have the same requirements of the people that do them. This is a big reason why hiring processes get cargo culted so heavily.
posted by feloniousmonk at 11:55 AM on November 16, 2013 [3 favorites]


I would argue that contributing to open source has a very low bar to step over in terms of wealth to participate. Certainly lower than going to the right university, certainly lower than belonging to the right social clubs. All you need is a desire to contribute, programming skill and an internet connection.

You also have to decide that this is a worthwhile use of your time, and if you don't already have friends who contribute to open source, or if you're not in an academic program (college etc.) where you see other people contributing, the benefits can be really unclear.

Then you have to figure out how to find a project: one that is suited to your interests, has opportunities to help that are within your skill level, has sufficient documentation, includes some way to talk to other contributors that you know how to use (hopefully they explain what an IRC client is instead of just saying "here's the channel"), and has contributors willing to help you get started. This is hard, and hopefully you don't have too much impostor syndrome holding you back.

You may also have heard that open source projects tend to have grouchy maintainers and contributors, so you have to decide you have the energy to tolerate that. Or if you haven't heard about that, you'll probably quickly find out - somebody will tell you that they don't have time to help you and you should just figure it out, that your contributions don't fit the undocumented style guide, etc. - and you have to have some stubbornness (and feeling like you belong there) to work through that and not get discouraged. And if your gender or other characteristics don't fit somebody's mental model of a normal contributor, you're more likely to get dumb lulzy comments directed at you on the IRC channel, or people thinking you're probably not worth the time to teach how to write tests so that you can contribute.

I am really lucky that I had the resources to go to conferences where I met people who convinced me to help with their open source project: OpenHatch, which is about helping people interested in open source figure out how to contribute to projects, through running workshops on college campuses, providing online "training missions" that teach tools like using diff and patch, running a directory of projects that want new contributors, having a kind and friendly IRC channel where people can ask questions, and generally being a public voice that says "yes, you are totally right to want to contribute to projects, this stuff is hard and somebody wants to help you!" So yay, if you care about these things, and if you want a friendly project to contribute to (in code or non-code ways, or with money) that is making things better instead of worse, come hang out with us.
posted by dreamyshade at 11:55 AM on November 16, 2013 [22 favorites]


When hiring, open source is a helpful way to see a "portfolio" of code. It's problematic that in the software industry, there's really no way to see a candidate's previous work—unless they have open source. I've never seen it used as a mandatory requirement, but instead as more of a tie-breaker. Myself, I have very few open source contributions because I simply don't have time, but I've had no problem getting jobs since I moved into software development because I am willing to produce code samples of various length upon demand (small one-offs I've done for clients previously or that I've done to facilitate my job, rather than code that the company used).

See, for people who haven't done hiring or worked in the industry, it might seem nuts to say that most applicants for software development jobs simply can't program, but it's true; our industry, given its low barriers to entry and relatively high compensation, attracts a remarkable number of people who claim to be programmers who can't solve problems at even a child-like level. The existence of some evidence that the candidate can actually do the job is... very helpful.

"Why Can't Programmers...Program?" is the classic in the field.
posted by sonic meat machine at 12:04 PM on November 16, 2013 [7 favorites]


What is it about software dev jobs that make it so a candidate who does it in their free time is assumed in some way to be "better" than candidates who don't? None of the jobs I've had or fields I've worked in have this assumption built in. For example, no editorial job I've ever applied for or gotten has demanded a portfolio of editorial work that I've done as an unpaid volunteer.

I think I know one answer is that this is kind of just how programming grew up, with enthusiasts enthusing and innovating and playing even if there wasn't any money in it, and now it seems "normal" that that's the best way to do it and to determine the "best" candidates for paying gigs. But "we've always done it this way" is kind of a terrible way for a field that's supposedly about creativity and innovation to grow its talent pool. And then it's embarrassing to see the hand-wringing over a lack of diversity in that pool.
posted by rtha at 12:07 PM on November 16, 2013 [3 favorites]


What is it about software dev jobs that make it so a candidate who does it in their free time is assumed in some way to be "better" than candidates who don't?

Passion. People who do it for fun (and yes, programming can be fun) are usually better programmers in general, not just in unpaid contexts, because they are driven to accomplish things and improve their craft. Furthermore, due to strict NDAs and NCAs, showing (as a candidate) or seeing (as an interviewer) even a tiny snippet of a competitor's code is legally horrific.

Plus, if you're a bad editor, you're not going to be able to stick with it long. The industry is small enough and connected enough that it just won't fly. There are a lot of programmers, and a lot of places that need programmers, and terrible programmers can just float from one contract to the next leaving a hideous wake of destroyed morale and code quality.
posted by sonic meat machine at 12:15 PM on November 16, 2013 [4 favorites]


Plus, if you're a bad editor, you're not going to be able to stick with it long.

Ha ha ha ha ha! /gasps

Ha ha ha ha ha ha /cries

Yeah, no. Look, passion is great. I love passion. I've served as a volunteer reader and done light copywriting for friends. My point is that getting a job as an editor is not dependent on my doing that kind of unpaid stuff. People have passions for finance stuff and all kinds of other things that they may also get paid to do, but most of those fields don't shut you out if you have no unpaid portfolio to show. They have other ways of figuring out who's got the chops for the jobs they're hiring for. I have a hard time believing that software dev stuff is *that* exceptional.
posted by rtha at 12:39 PM on November 16, 2013 [8 favorites]


The article seems to make at least one fairly extreme assumption about how exactly employers tend to use OS contributions as an indicator. It emphasizes volume of work to a degree that doesn't jibe with my (admittedly very limited) experience. I've seen employers giving attention to even quite small samples of work, when it's been well done and indicative of ability to come up with useful solutions independently.

Another issue I have with it is actually highlighted if you read the comments to Chris Anderson's tweet: A lot of people already agree that something as simplistic as just looking at a Github profile is neither a necessary or sufficient test.

The article has many good side-points (various kinds bullying in some parts of OS land, for example) but the main thesis just rings to me like a false representation of what's actually going on in the OS community in general.
posted by Anything at 12:46 PM on November 16, 2013 [1 favorite]


Plus, if you're a bad editor, you're not going to be able to stick with it long.

Ha ha ha ha ha! /gasps

Ha ha ha ha ha ha /cries

Yeah, no.


Do you honestly think that a person without even the barest semblance of an editor's basic skills—reading, say—could get a job as an editor? These peoples' equivalents do get jobs in the tech industry.
posted by sonic meat machine at 12:48 PM on November 16, 2013 [1 favorite]


Also, open source is a "pay it forward" kind of thing for many people. If we didn't have open source, we'd be stuck with proprietary tools, proprietary languages, proprietary operating systems, and the world would be infinitely poorer for it. Compare learning a Microsoft language like Visual Basic to learning an open source language like C or Python; the difference is night and day, not least because learning Visual Basic required, for much of its history, an up-front purchase of hundreds of dollars.

Where would the Internet be today without nginx, Apache, PHP, Python, Ruby, Rails, Javascript, Linux, Wordpress, Drupal, Mediawiki, MySQL, Postgres, redis, svn, git, and even (maybe especially?) the humble GNU tools? I'd expect it to be a much worse place.
posted by sonic meat machine at 12:58 PM on November 16, 2013 [8 favorites]


When hiring, open source is a helpful way to see a "portfolio" of code.

Absolutely. The fact that the code is open-source is neither here nor there for me as an employer. A portfolio is absolutely expected in most creative professional jobs, and software development is a creative job. It is writing.

...no editorial job I've ever applied for or gotten has demanded a portfolio of editorial work that I've done as an unpaid volunteer.

And I would not demand the same. But the problem with asking for a closed-source code sample is that it's one more step that I need to do to vet a candidate. A Github URL is easier.

(And yes, I prefer to review code on Github. It's not trendy to say that their UI is superior to Google Code or, shudder, Sourceforge, and it's incredibly valuable to see not just how the individual writes code, as in a standalone code sample, but also how she or he responds to bug reports or pull requests.)
posted by nev at 1:09 PM on November 16, 2013


I think it's possible to say, on the one hand, yes, open source software is a net positive for society, and on the other hand, it does depend on unpaid labor, which makes contributing to OSS a potentially problematic thing to judge people on. For instance, I can definitely see how it would amount to judging someone based on their out-of-work "lifestyle," while ignoring that not everyone has equal access to that "lifestyle." Also, institutionally valuing that kind of dedication to programming above other interests may make it easier for programmers to be exploited (see for example the video game industry, which we just had a post about).
posted by en forme de poire at 1:16 PM on November 16, 2013


These peoples' equivalents do get jobs in the tech industry.

If this is so, then the problem is not that people aren't demonstrating passion by contributing to open source projects. It's that employers can't figure out (apparently) how to weed out the incompetents. Nearly every editorial job I've gotten has had its candidates take a short test for skills before the interview step. Used to be you went to the office to do it; now, I gather, it's often an online test. I mean seriously, if companies are hiring people who are so incompetent that they don't even know how to, I dunno, close tags or write a very basic query, solving that is not "demonstrate your passion!"

Personally, I don't want to work for a place that cares that much about what I do when I'm not in the office if I'm not being hired as a very senior and public-facing employee.
posted by rtha at 1:20 PM on November 16, 2013 [3 favorites]


It's that employers can't figure out (apparently) how to weed out the incompetents.

Very true. At my company, we do give programming exercises, and require coding during the interview, but even these are very costly; if two employees at an hourly rate of $60 spend an hour interviewing a candidate, and five each spent fifteen minutes reviewing a programming exercise, that's a total of about $200 spent just to reject an incompetent. Shorter/cheaper methods are very helpful, and providing a Github URL (or StackOverflow profile) is just a way of making yourself stand out from the slush pile.

The interesting thing is that, at a lot of other companies (including my immediately previous job), coding during the interview was viewed as somehow insulting. Ridiculous, and I campaigned against it, but eventually just gave up and left. There's only so long you can fight against entropy being forced upon you by management.
posted by sonic meat machine at 1:31 PM on November 16, 2013 [1 favorite]


I think it's possible to say, on the one hand, yes, open source software is a net positive for society, and on the other hand, it does depend on unpaid labor, which makes contributing to OSS a potentially problematic thing to judge people on.

Forgive me if I'm misunderstanding something, but this comment and others, including the TFA, seem to involve a lack of accounting for the option of writing software that is already useful enough for the author him/herself to get written, and putting that up for others to improve -- where the labor pays for itself in the form of utility, and the mutual gains from sharing are just a bonus.
posted by Anything at 1:31 PM on November 16, 2013 [1 favorite]


The article makes a good case is that OSS contributions are affected by privilege, and are not actually pure indications of merit. However, this is also true for basically any signal a prospective employee may try to send to employers. Does your family have sufficient emotional resources so that you can spend your time in school comfortably learning instead of anxiously dealing with trouble? Privilege. Does your family have sufficient financial and social resources so as to develop your niche academic and non-academic interests, thereby paving the way to your admission to an elite institution of higher education? Privilege. Does your family have the social capital so that you can present well and "fit in" in an interview at a company where most people are upper-middle-class? Privilege. Do you have the resources so that you can develop a "passion" for something? Privilege. Did you have the connections/academic credentials to get a job at a highly recognizable/prestigious firm early on in your career? Privilege. Etc etc etc etc etc. The whole world runs on privilege.
Do you lack anxiety disorders that prevent you from functioning? Are you neurotypical and not autistic, dysgraphic, dyslexic? Do you have health privilege? Are you subject to illness, either physical or mental?
posted by This, of course, alludes to you at 2:00 PM on November 16, 2013


The great ones almost invariably love doing it, they have personal projects, they write about it, speak about it, and think about it both in and out of work. The ones that are primarily attracted to the salary never quite seem to get the idea of craftsmanship.

This is just a false taxonomy. Really.
posted by polymodus at 3:06 PM on November 16, 2013 [5 favorites]


Also, the correct way to look at this phenomenon is in terms of economic externality. You'd think programmers would have thought to think in terms of a rigorous model in the first place—and its convenient omission is what's revealing about the prevailing mentality.
posted by polymodus at 3:09 PM on November 16, 2013


Very true. At my company, we do give programming exercises, and require coding during the interview, but even these are very costly; if two employees at an hourly rate of $60 spend an hour interviewing a candidate, and five each spent fifteen minutes reviewing a programming exercise, that's a total of about $200 spent just to reject an incompetent. Shorter/cheaper methods are very helpful, and providing a Github URL (or StackOverflow profile) is just a way of making yourself stand out from the slush pile.

So it's not passion-as-proxy for competence, it's that it's cheaper to find people who are not obviously incompetent by enforcing a cultural norm like working on open source projects. That makes more (and more horrible) sense than the passion explanation, anyway.
posted by rtha at 3:23 PM on November 16, 2013 [5 favorites]


It's no utopia, but now you can go leave and work at an insurance company or something, where no one cares how passionate you are.

No, this is standard at normal companies now, too. I've worked in completely non-sexy global conglomerates all my career and getting lectured in team meetings about how you have to have a passion for your work is perfectly normal, even when that work is capacity planning and forecasting.

Affective labor is not just a thing in service industries and tech companies. I think it's revolting, personally.
posted by winna at 3:25 PM on November 16, 2013 [16 favorites]


Forgive me if someone else mentioned this, but the strangest thing about all of this is the assumption that most OSS is done in peoples' spare time. Uh, no, not for most important projects. Most people I have known that worked on OSS extensively (myself included) did it as part of their paid day-job. I work on an OSS project that I only joined due to using it during a day job, and while now I continue to do it in my limited spare time, I never would've gotten onto the project if my company had not basically paid for me to spend time on it. And that is true for every single major contributor to this project. GitHub toy projects are OSS, I guess, but they're more like writing samples or something. Still unfortunate that they're considered "required" by some companies, but I really don't think they are the same thing as true OSS development, which mostly happens by folks paid in some way to do it.

So really, this isn't exactly filtering for "spare time" coding. This is filtering for "working for a company that allowed you to OSS stuff, or that encourages you to submit back to OSS projects that you are working with". Or else "knowing enough about current culture to bother creating a GitHub and putting some toy something up there", which is I suspect what most employers actually look at.
posted by ch1x0r at 3:41 PM on November 16, 2013 [2 favorites]


It's no utopia, but now you can go leave and work at an insurance company or something, where no one cares how passionate you are.

For me this would, in fact, be utopia. Where are these mystical jobs where I can go in, do my work competently, get paid and go home with time to spare with friends and family? I'm sick of being told I need to be passionate about what I do. I'm willing to work hard because I want to get paid. Why on earth is that not enough?
posted by never nice at 3:48 PM on November 16, 2013 [10 favorites]


Could not be more wrong. Requiring "passion" (or at least far more than 40 hours a week) to enter and advance in a profession probably the greatest pro-egalitarian force a profession can have.

Larry Ellison's kid could breeze into any posh marketing job at any health care company in the world, but that company wouldn't hire him as an graveyard shift ER doc in the lowliest of its hospitals unless he'd sweated out the same grueling residency shifts as the kid of the weekend janitor at Oracle had to do.

In my field (finance) it couldn't be more clear. When it was defined as "123" (take deposits at 1%, lend them at 2%, and be on the first tee by 3 p.m.) it was a game of pedigree almost to the exclusion of anything else. Now that's its gated by SATs, 100-hour-week analyst programs, and requires long days, nights and weekends even for people who've been doing it for decades, the social backgrounds are fabulously more diverse.
posted by MattD at 4:00 PM on November 16, 2013 [4 favorites]


If you extract social production from consumer empowerment, what's left?
posted by parmanparman at 4:14 PM on November 16, 2013


that company wouldn't hire him as an graveyard shift ER doc in the lowliest of its hospitals

ER docs work all kinds of shifts, it's not a hierarchy where you earn a certain shift or something -

--kid of an ER doc
posted by sweetkid at 4:17 PM on November 16, 2013 [1 favorite]


I have been hired before in part due to past work in open source projects relevant to the company's field, and with the expectation that I continue to support that work, but I considered that part of my work; I switched to other stuff in my spare time-- stuff that was fun, not work. If a business wants to be part of an open-source ecosystem, they need to contribute to it, and they should have employees do so as part of their job, to the extent that it benefits the company or its customers.
posted by thefool at 4:21 PM on November 16, 2013 [1 favorite]


I think I will link to dreamyshade's comment whenever I see an AskMeta question about "how to break into software" get facile, hand-wavey answers to the effect of "it's super-easy, just go contribute to some projects and put stuff on GitHub."

Also, I'm confused by the fixation of many users on the purported whiteness of programmers. I have lots and lots of coworkers with south and east Asian ancestry. I haven't seen what software companies look like in the deep south, maybe the purported disconnect between the racial composition of the workforce and the local population is clearer there. From where I'm sitting in Silicon Valley, employment in tech correlates much better with family finances, top-of-the-line academic pedigrees, and other things having to do with class.
posted by Nomyte at 4:44 PM on November 16, 2013 [1 favorite]


This is just a false taxonomy. Really.

I'd like to hear an elaboration of this. I've seen what I described play out exactly the same way in 3 different countries with different cultures and vastly different economic status. I know judging code is pretty subjective, but in all places the code I thought was elegant was produced by the people who would talk your ear off about technical details over beers after late nights in the office - not the ones who left at 5:00 sharp to go watch sports (or similar).
I currently work at a place with great work/life balance, and we all leave around 5, but the best code is still produced by the ones who are going to meet-ups and hacking on stuff on the side.
posted by bashos_frog at 4:53 PM on November 16, 2013 [2 favorites]


I find it interesting that I could correctly guess ahead of time that this was probably coming out of either Ruby or Python and not, say, JavaScript, C#, Java, PHP, C, Objective-C, any of the other big commercial languages. What's up with this? Backlash from Ruby and Python being trendy & attracting "bros" where that is a thing?

Also, I'm confused by the fixation of many users on the purported whiteness of programmers. I have lots and lots of coworkers with south and east Asian ancestry.

I've been thinking more and more these past few years that in America "Asians" and "Indians" are starting to become subcategories of "White" as has been well-described for "Jews" and "Irish". (I wonder if it will still be called "White".) I'm kind of getting that sense from several perspectives in IRL discussion and in media - from White straight-up self-admitted racists saying "I think the Asians are on our side," from Asian-Americans describing themselves as "yeah, basically White," from having 2 Indian-American Republican governors, from diversity advocates attacking the Whiteness of tech fields (It's pretty clear in this article who is not considered a "person of color."), and from the blatantly obvious Affirmative Action "Asian penalty," that last which is just like COME ON.
posted by save alive nothing that breatheth at 5:13 PM on November 16, 2013


I've been thinking more and more these past few years that in America "Asians" and "Indians" are starting to become subcategories of "White" as has been well-described for "Jews" and "Irish".

I've been wondering about this too. I'm not from the USA, so I'm often blind to or baffled by the rich tapestry of the racisms and race-tensions here, but my current working assumption is that "white" doesn't mean white so much as "races/ethnicities/cultures currently over-represented in social success/power", while "minority" doesn't mean minority so much as "races/ethnicities/cultures currently over-represented in poverty".

Being Not From Here, I can't tell if my current interpretation is unspoken but agreed, or offensive, or missing some key idea, or what, so I generally just keep quiet on these topics and read relevant articles when I find them, but I'm currently in the dark about what is meant in cases like this where noticeably mixed workforces are described as "white".
posted by anonymisc at 5:34 PM on November 16, 2013


For me this would, in fact, be utopia. Where are these mystical jobs where I can go in, do my work competently, get paid and go home with time to spare with friends and family? I'm sick of being told I need to be passionate about what I do. I'm willing to work hard because I want to get paid. Why on earth is that not enough?

never nice, I know developers, both competent and not, at a few different companies that clock in and clock out after eight hours max. I've been able to work at a couple of jobs like that myself. I've worked in Chicago and Boston. I can't speak to Silicon Valley, but I think you should definitely look around, focusing on companies that are not built around selling software or services.
posted by ignignokt at 5:49 PM on November 16, 2013 [2 favorites]


Being Not From Here, I can't tell if my current interpretation is unspoken but agreed, or offensive, or missing some key idea, or what, so I generally just keep quiet on these topics and read relevant articles when I find them, but I'm currently in the dark about what is meant in cases like this where noticeably mixed workforces are described as "white".

Most people will have heard of Twinkies and Oreos, "acting white," that awful fad of "things white people like" a few years ago, and so on. Minority identities are complicated. With the possible exceptions of activists and social scientists, the categories people use to reason about these identities are probably pretty problematic and inflammatory to various audiences.
posted by Nomyte at 5:52 PM on November 16, 2013 [1 favorite]


This is just a false taxonomy. Really.

It's... really not. Programming, at a high level, is complex and interesting enough that you can study it for years and always have something new to find. People who find their field interesting, and who use their spare time to learn about it, and expand their understanding—they get better. People who got a degree from Directional University of State in "Information Systems Technology" and never read any of the seminal texts of the field stagnate at a low level. They remain junior level programmers at large companies, move on to become managers, or drift between contracts at low-tier companies with bad hiring practices.

(If you ever hear a manager say that he doesn't code now, but "he's paid his dues," you know that he was one of these types of programmers. He's also likely to be a terrible manager.)
posted by sonic meat machine at 6:58 PM on November 16, 2013 [3 favorites]


So it's not passion-as-proxy for competence, it's that it's cheaper to find people who are not obviously incompetent by enforcing a cultural norm like working on open source projects. That makes more (and more horrible) sense than the passion explanation, anyway.

It's one signal among many. Like I said, I've never seen a company that uses open source contributions as an absolute metric. I remain gainfully employed despite my anemic Github, for example. I also am not particularly socioeconomically privileged, except as my career in technology has allowed me to be.

Another thing that you might be overlooking here is the incredible amount of risk that hiring an incompetent boob represents in software, and the importance of someone fitting in socially. Hiring a NNPP can be an absolute catastrophe for a programming team; they can do a lot of damage by writing bad code, but they can also destroy the productivity of your other engineers by requiring constant handholding (and bug-fixing, after the fact). If the management doesn't bring the axe down quickly enough after a bad hire, morale will be destroyed among the good engineers. If management brings down the axe too quickly after a bad hire, morale will be destroyed among the good engineers.

Psychology, surprisingly enough, is often the determinant of success or failure in software projects. (Think about the poor bastards working on healthcare.gov. Think that was a death march?)
posted by sonic meat machine at 7:13 PM on November 16, 2013 [1 favorite]


This has been a really interesting discussion. I know I've said my piece over and over again in this thread (and I realize this thread is not specifically about me - but as someone that hires engineers and values open source work, I feel I'm in the minority point of view here). To summarize one more time, I do not require someone to do OSS work to get a job with me, I just think it's a one in the plus column. Also, I am not saying that passion is a proxy for competence, but surely passion isn't a negative? Is there anyone that would argue in this thread that all things being equal that passion for the field is a bad thing? If someone enjoys the field so much that they deem to work on it in their spare time, this is a fact that should be ignored, or even a cause for concern? I don't require these things, I don't judge solely on these things - but if two theoretical candidates are the exact same, except one also likes to code in their spare time - is there anyone here that if it was their company and their money they wouldn't choose the person who also programs as a hobby?
I will say that someone that came into the industry during the dot-com era, and did the whole 100 hour weeks for a few years (under other people's management) that I definitely strive to find people and create organizations that have some kind of work/life balance. I've found that above 50 hours a week, the returns diminish very rapidly, never mind the fact that's it's not good for people and their families. All that being said, I think someone can have a decent work/life balance, work and code as a hobby. I mean the main open source thing that comes up when a prospective employer Google's my name was something I worked on for 3 weeks in 2001 - that seems to be enough for most people....
posted by ill3 at 7:34 PM on November 16, 2013 [1 favorite]



I've been thinking more and more these past few years that in America "Asians" and "Indians" are starting to become subcategories of "White"


Sometimes. In economic/educational categories, yes, in "stereotypes about what R U" Asians/Indians are not quite white people yet. As an Indian American, I identify more with "white privilege" type feelings than other "people of color," and yet, when around both groups of white people and groups of black or Hispanic people I'm reminded that I am different in some way based on background, economic class and sense of identity. It's sort of a strange spot.

But more to the specific point in this thread, no having Asians/South Asian programmers doesn't fulfill your diversity requirements because the reality of race in America is more complex than "we've got some brown people here so"
posted by sweetkid at 8:29 PM on November 16, 2013 [3 favorites]


What is it about software dev jobs that make it so a candidate who does it in their free time is assumed in some way to be "better" than candidates who don't? None of the jobs I've had or fields I've worked in have this assumption built in. For example, no editorial job I've ever applied for or gotten has demanded a portfolio of editorial work that I've done as an unpaid volunteer.
Software is a weird, schizophrenic industry.

Most programming jobs are primarily maintenance work. These are companies that already have the software built, and they just need to make some small adjustments and keep the gears turning to support the business. The "developers" are really just there to change the oil and rotate the tires. A lot of these jobs are at non-software multinationals, banks and the like.

Programmers who actually create software systems during their day job are in the minority. There just aren't a lot of programming jobs where your main responsibility is architecting something from scratch. In a large company, the architecting work is done by some manager type while the actual coding is carried out by the lowly peons below. It's really only possible to do both the "big idea" stuff and the coding if you're lucky enough to get in at a startup before they get too big and descend into Dilbert-esque politics, or you work as a freelance/consultant type on a project where you have a lot of creative control.

Bottom line, most programmers are maintenance programmers in their day jobs.

But nobody wants to hire a maintenance programmer. Interviewers can sniff out candidates who obviously only did 1% of the project they put on their resume, and "supporting" a system is not likely to get you much in the way of career advancement.

Unfortunately, a lot of the skill of programming is forged in the fires of some sort of creative role. If you're a programmer and you care at all about advancing your skills, an open-source project can be a great way to put yourself in that creative position voluntarily. Not to pad your resume, but to actually develop real skills that you aren't going to acquire fixing bugs in decades-old code at your crappy day job.

Open source projects also usually get started up organically due to some on-the-job need. Take Firebug for example. Firebug was, in its heyday, an indispensable tool for anyone who had to do anything on a webpage. It was built by people who worked in that field for their own convenience, and anyone else who cared to could contribute. And they did it for free. Just 'cause.

So I think comparing it to working for free as an editor is an apples-to-oranges comparison.

As far as I know, there is nothing like OSS in any other industry. Software has its own weird cultural norms, partly because of the nature of the work. It's incredibly easy to share code and contribute to existing open source projects. Very often, contributing to OSS is coterminous with doing your job well. I have contributed patches to open source projects to fix bugs I found while using them in the company's proprietary code.

Personally, though I think it's stupid to require anyone to be a OSS contributor as a prerequisite for employment. That's antithetical to the organic origins of open source. The point is to collectively build tools that are freely available to make everyone's job easier. It's probably hard to imagine why anyone would do this if you work in a field where sharing tools is not really possible.
posted by deathpanels at 9:04 PM on November 16, 2013 [10 favorites]


My contributions to open source products (that I did during layoffs and weekends while working manual labor), were one factor that allowed me, as a person who has never had a college credit to his name, to get a decent middle class job with benefits.

Poverty is hard. It is harder than learning to program is. I am not going to pretend I pulled myself up by my own bootstraps, I was simply lucky enough to find a hobby that other people would pay me money to indulge in.

Open source maintainers are grumpy because users come to us demanding fixes or features. If you are asking someone for a favor, and not offering money, a few pleases and thank yous are called for.
posted by idiopath at 9:53 PM on November 16, 2013 [4 favorites]


I agree with deathpanel's assessment in a lot of ways, especially as an assessment of the current state of the field. But two things about it bother me:

First, "Unfortunately, a lot of the skill of programming is forged in the fires of some sort of creative role". This is thought to be true, but it's a way the field is misfocussed, and that it needs to grow out of. He's right that most code work is maintenance work--or either maintenance work and extension work, where you're adding onto existing code. The skills that make one able to do this work well are undervalued, or only paid lip service--like being able to read code, to diagram it, to pull apart a codebase and understand it, such that fixing bugs or implementing the next feature is much less risky and much more directly related to improving the quality of the code. Rather than looking for the coder to write code for OSS projects, we should be looking for coders who go into existing OSS codebases and refactor or document them. Between two candidates, one who started her own project and one who spent her time modernizing a lagging project, the latter is, to me, obviously the superior candidate for the likely work. Want to really test a candidate in an interview? Don't give them a problem and see how they solve it. Give them a two page function with no comments, bad variable names, and confusing code paths, and ask them to figure out what it does and then to clean it up.

Second, I think not enough attention is being paid to the fact that, even when you do get to do the "creative" work, where you're figuring out something from scratch, the bulk of the labor is still boring, uncreative detail work requiring a long attention span and a willingness to slog through the unglamorous parts. It's not that hard to conceptualize an API or a novel way for a client and server to interact--but it's hard and unrewarding to actually work out the implications of that conceptualizing, implement a significant part of it, test it reasonably well, and then add sufficient documentation that someone else can make use of it. This is the craftsmanship side of programming, which is a willingness and ability to polish it properly, and which is arguably more important than one's architectural skills.
posted by fatbird at 10:01 PM on November 16, 2013 [9 favorites]


I dunno. Who has time to look at candidate's Github profile?

I feel like I can reliably sort resumes into "genius", "complete idiot", and "other". Geniuses can find more interesting work for them that pays more money than joining our team, usually, so I'm just trying to figure out which of the "others" are good enough. We haven't done any hiring in a while, so I haven't had a chance to use this yet, but stuff like that tells me what I need to know in fifteen minutes. Can you look at a short piece of code and tell me what it's trying to do and how you can make it do that better? I don't care what you do in your free time.

I'm good at what I do, I'm a professional, but I'm not the best. Almost nobody is the best, that's what it means to be the best. Companies should try to hire professionals, who often have well-rounded lives. That's who my company hires, anyway, and we've been very successful with our product.

I do think it's a startup thing, like somebody mentioned above...if you don't make any money, tech team skill level is something you can try to sell. Also, death marches from 23 year olds who don't know any better might be the only reason you're still afloat, so you have to select for that somehow.
posted by Kwine at 11:36 PM on November 16, 2013 [3 favorites]


There's so many assumptions in this piece and in this thread I don't even know where to start picking at it.

1. Your github repository is a portfolio of your work. We're not asking artists to not have something potential employers can't look at for fear of subtly adding bias. We're not asking the same of architects or writers or any of a million creative businesses.

2. Open source is at its best when you have a requirement, you write code to fulfil that requirement and then you share the stuff you made with the world. If I create a PHP library for processing graphic images in a certain way because I need it, and I then allow other people to use that library, I'm not being taken advantage of.

3. I dread to think what the internet would be like if it wasn't for the hours people have spent making for free the things that are on the internet. This includes open source software like Rails and the Apache web server. But it also includes the thousands of blogs and pictures and animations and podcasts and songs that people (women and minorities included) have created gratis. We're not here, sitting around, wondering if Allie Brosh (writer of Hyperbole and a half) only got that book deal because her privilege gave her the free time to draw and write hilarious stories about her dogs.


We need to look seriously as to why women aren't represented in the world of open source. Unfortunately, the linked article misses that by a margin. And it also manages to paint sharing the secrets of your success (the code you write) as a bad thing. That's just depressing.
posted by zoo at 1:17 AM on November 17, 2013 [11 favorites]


Now that's its gated by SATs, 100-hour-week analyst programs, and requires long days, nights and weekends even for people who've been doing it for decades, the social backgrounds are fabulously more diverse.

I believe that finance is more diverse than it used to be, but I think this probably has more to do with more diversity in admissions at the places financial institutions tend to recruit from (e.g. HPY), as opposed to some kind of intrinsic meritocratic quality that only working for free can evince.

that company wouldn't hire him as an graveyard shift ER doc in the lowliest of its hospitals unless he'd sweated out the same grueling residency shifts as the kid of the weekend janitor at Oracle had to do.

Except residents get paid. Even pre-residency medical students get loans that cover their living expenses. And another problem with both of these analogies is that residency and analyst internships are explicitly part of a training program with multiple and clearly defined steps -- all practicing doctors must do a residency. It's also sort of ironic to bring up gating on tests like the SAT/MCAT/whatever, because one of the things people have been saying in this thread is that explicitly testing for programming knowledge could actually be more egalitarian than viewing GitHub as your "resume."
posted by en forme de poire at 1:22 AM on November 17, 2013


It's... really not. Programming, at a high level, is complex and interesting enough that you can study it for years and always have something new to find. People who find their field interesting, and who use their spare time to learn about it, and expand their understanding—they get better. People who got a degree from Directional University of State in "Information Systems Technology" and never read any of the seminal texts of the field stagnate at a low level. They remain junior level programmers at large companies, move on to become managers, or drift between contracts at low-tier companies with bad hiring practices.

So people who didn't go to MIT or Stanford can never "find their field interesting" or "use their spare time to learn about it" to "expand their understanding" and "get better"?

I love that, in the midst of this thread full of people promising that anyone can be a software developer if they are passionate enough, we still have an assurance that people who lack the privilege to get a name brand education at the right kind of school obviously can never be in that category. But it's not because they weren't wealthy enough to go to the "right" school, or because they were a first generation college student who had to live at home and work two jobs while help taking care of their siblings and therefore went to the school closest to their home. No, clearly it's just because they aren't passionate enough.
posted by hydropsyche at 6:13 AM on November 17, 2013 [8 favorites]


So people who didn't go to MIT or Stanford can never "find their field interesting" or "use their spare time to learn about it" to "expand their understanding" and "get better"?

Not what I meant. I have a degree in an unrelated field. However, there are a lot of low-end schools that turn out people who barely understand the magical incantations that they put into the white box that shows up when they double click the "DrJava" icon, let alone how to actually solve problems.

There are diamonds in the rough, and there is often a good professor or two at any university, but the interface between the profession of software development and the education related to it is abysmal.
posted by sonic meat machine at 6:18 AM on November 17, 2013


It just makes me sad how much the goalposts continue to move for women in science and engineering careers. First it was, "Women just don't have the interest or the ability, but there's nothing wrong with that -- vive la difference!" (And it still is to some extent.) Then it was, "Okay, you have the interest and ability, but do you really want to work 40 hours a week without a stay-at-home spouse to do the housework and childcare?" Now it's, "40? We actually meant 60, and of course you'll be programming for fun in your spare time unless you're one of those unpassionate people who would be better off at an insurance company. Oh, and 'childcare' now includes make-your-own-baby-food and breastfeeding on demand."
posted by Ralston McTodd at 7:02 AM on November 17, 2013 [8 favorites]


Are you suggesting that it's the fault of the employer that a husband is not doing the childcare?
posted by Anything at 7:16 AM on November 17, 2013


I am suggesting that employers create an environment that is far less accommodating to outside responsibilities than the work culture of decades ago. Back then, jobs could simply exclude women; now I suppose they have to come up with something more sophisticated. I'm sure it's more complicated than that, but that's what it looks like from here.
posted by Ralston McTodd at 7:31 AM on November 17, 2013 [1 favorite]


I am suggesting that employers create an environment that is far less accommodating to outside responsibilities than the work culture of decades ago.

No - You're suggesting that employers, specifically STEM employers are deliberately increasing the hours needed from their employees in order to exclude women.

Is there a conversation to be had about the hours expected by its employers of their employees? Yes! Absolutely! Should we talk about how we can best incorporate the needs of primary caregivers (of the young and the old) into the workforce? Yep. You bet.

If anything, Software houses treat their staff better than pretty much any industry out there. And that includes those middle-class industries we currently seem to be concentrating on.

I look forward to having exactly this same conversation when someone reports that teachers are more likely to get work if they're the sort to organise after-school clubs and how this entrenches the hopeless sexism of the teaching business.
posted by zoo at 8:02 AM on November 17, 2013 [5 favorites]


I will also say, I think "years of experience" is a horrible tool for assessing a programmer's abilities. I quit my job at a giant corporate web company because I was tired of watching knuckle-dragging idiots with "ten years of experience" get promoted into positions of power and then proceed to make bad decisions that cost the company, and their underlings, money and hours.

Talent is the ability to a) recognize a mistake, b) think of a fix for the mistake, and c) implement it. This equation applies to any skill.

I've found that the only way to get better at programming is to write your own programs, your scratch, and to observe the mistakes of others and learn to recognize them as mistakes.

Re: Ivy League snobs who won't hire anyone who hasn't read Knuth:
You know how many times 99.999% of programmers will need to implement a B-tree? Zero times.

We should place more value on the skill of deciphering old code. In most software jobs, you are vastly more likely to need to refactor some ancient Perl than you are to need to write a tree sorting algorithm.
posted by deathpanels at 8:34 AM on November 17, 2013 [2 favorites]


as a software developer who does OSS stuff, i will never, ever ever make wikipedia edits.
posted by localhuman at 8:36 AM on November 17, 2013 [3 favorites]


i also think that the industry is obsessed with recent trends. (have you used node, have you used grunt, how do you install dependencies) I install dependencies with yum and I think the best webserver is apache. yay for javascript and all its friends but who the hell thinks running node as a server is a good idea?
posted by localhuman at 8:44 AM on November 17, 2013




I've implemented a b-tree before. It was 25 years ago, mind.
posted by zoo at 8:51 AM on November 17, 2013


Idiopathic:

I wouldn't trust or listen to a single word that comes from Larry Ellison's stupid greedy mouth.

"Fashion-driven" probably means "not supplied at great expense by Oracle"
posted by zoo at 9:01 AM on November 17, 2013 [3 favorites]


"Fashion-driven" probably means "not supplied at great expense by Oracle"

Fashion driven means I showed up on the first day of my contract and got told that the design agency was handing us a completed front end built with requirejs, backbone, handlebars, and sass/compass, among other things, all wrapped up in a grunt project responsible for building it, which of course required npm and ruby to be installed for a django backed website. The team lead looked around and asked "does anyone know any of this stuff?" Previously they made PSDs available for download. So it was a step up in one way. Whether or not Ellison said it, it's true.
posted by fatbird at 9:50 AM on November 17, 2013 [5 favorites]


You're suggesting that employers, specifically STEM employers are deliberately increasing the hours needed from their employees in order to exclude women.

I'm sure their motivation isn't explicitly framed as wanting to overtly exclude women, but in the startup/VC area, there's certainly a cachet associated with white male ivy dropouts[1] in spite of actual data showing that women's involvement has a positive correlated with success.

[1]: "... all seem to be white, male, nerds who've dropped out of Harvard or Stanford and they absolutely have no social life. So when I see that pattern coming in ... it was very easy to decide to invest."
posted by rmd1023 at 9:53 AM on November 17, 2013 [3 favorites]


I hate it when I miss a stupid grammar mistake in one of my comments until just after the edit window has ended.
posted by rmd1023 at 10:02 AM on November 17, 2013


I've been in the industry for 15 years now. While I see this open source hiring thing mentioned in the sort of startup/tech press, I haven't seen it in action at larger companies (Google doesn't seem to use it as any significant hiring factor).

More importantly, I have seen zero correlation between skills on the job and OSS/hobby programming outside the job. Many talented people also spend their free time coding, but many others spend it with family or traveling or volunteering or just gaming or whatever. If I think of the best people I've worked with over the past 15 years, I'd say less than a quarter make any OSS contributions. So to the extent companies may be using this, it's a terrible metric in my experience.

(I also think that the big companies actually have decent work-life balance if you choose it, I've had no issues with that in the 10 years since I left startup land and that has been spent working for 2 of the biggest names in tech).
posted by wildcrdj at 11:06 AM on November 17, 2013 [1 favorite]


"Fashion-driven" is definitely true, even if The Great Satan did say it. JavaScript is currently in vogue (how, given its many flaws, I have no idea). MongoDB was The Thing before that. Ruby on Rails and its pathological attitude toward database integrity is still with us, but winding down. Meanwhile, the JVM putters along, powering pretty much everything of any size...
posted by sonic meat machine at 11:18 AM on November 17, 2013 [1 favorite]


If I think of the best people I've worked with over the past 15 years, I'd say less than a quarter make any OSS contributions. So to the extent companies may be using this, it's a terrible metric in my experience.

I don't think advocates were really suggesting it's a metric that highlights most of the good people, they're suggesting it's a filter that bad programmers are unlikely to pass. That lots of good people don't pass it either doesn't affect that utility when your purpose is to winnow down an overwhelming number of applicants - it may be they feel they can afford to throw out a lot of babies with the bathwater, as long as it helps them end up with a baby and not bathwater.
In a market where openings exceed candidates, then I expect (or would hope) it would take a back seat even for people who advocate it.
posted by anonymisc at 12:44 PM on November 17, 2013


In a market where openings exceed candidates, then I expect (or would hope) it would take a back seat even for people who advocate it.

The irony is that, in the current environment, openings do exceed good candidates.
posted by sonic meat machine at 12:59 PM on November 17, 2013 [1 favorite]


(Which is to say, I agree with you, but I'm scarred by an economy where some employers were using filters like discarding all applicants who are unemployed. OSS contributions is empowering by comparison. Unfortunately.)
posted by anonymisc at 1:00 PM on November 17, 2013


Doing quite well working on a JavaScript/NoSQL driven site at the moment, FWIW. Doing what it's doing at the capacity it's doing it an Oracle solution would melt.
posted by Artw at 1:08 PM on November 17, 2013 [1 favorite]


(quite a bit of Java in there too, traditionalists.)
posted by Artw at 1:08 PM on November 17, 2013


Lessons from a year’s worth of hiring data, from a few months ago, has some interesting things to say about looking at projects to filter resumes, apart from considering fairness:
To game the system, applicants start linking to virtually empty GitHub accounts that are full of forked repos where they, at best, fixed some silly whitespace issue. In other words, it’s like 10,000 forks when all you need is a glimmer of original thought...Outside of that, there’s the fact that not all side projects are created equal.

...Telling the difference between these kinds of projects is somewhat time-consuming for someone with a technical background and almost impossible for someone who’s never coded before. Therefore, while awesome side projects are a HUGE indicator of competence, if the people reading resumes can’t (either because of lack of domain-specific knowledge or because of time considerations) tell the difference between awesome and underwhelming, the signal gets lost in the noise.
posted by dreamyshade at 1:33 PM on November 17, 2013 [6 favorites]


yay for javascript and all its friends but who the hell thinks running node as a server is a good idea?

People who are doing work in which they don't want computers to work as hard as Apache makes them work. Node is not magic – it just does less stuff by default and uses up fewer resources.

Grunt, on the other hand, is more of a comfort thing. Shell scripts and makefiles work fine for me, but I am not going cast aspersions at people's motivations for using it just because they're not writing code or build scripts in a way that is different from what I know and love.
posted by ignignokt at 3:01 PM on November 17, 2013 [1 favorite]


Trying new things isn't about fashion. It's about widening your experience and growing your toolbelt.

OK - Well, maybe it's a bit about fashion. But honestly, the ecosystem is better for it.

I don't understand why you would compare a VM to a scripting tool to a database to a web framework to a programming language though.
posted by zoo at 3:12 PM on November 17, 2013 [1 favorite]


I'm not comparing one to the other. I'm saying they were the big fads in sequence (or, more accurately, blended into each other over the last few years).
posted by sonic meat machine at 6:42 PM on November 17, 2013


Great Q&A with major Drupal Contributor "Webchick" Angie Byron:
"You're a new parent, yayyyyyy congrats still! Has it changed your viewpoint on contributing to open source or Drupal? Thoughts on work/life balance?"

It definitely made the challenges with the "do-ocracy" model of development a lot more real, rather than merely empathetic/conceptual, for me. While I love the aspect of do-ocracy which empowers people to just jump in and make change without having to ask anyone for permission (and we've seen some truly amazing people do truly amazing things with this model), it also certainly heavily favours "he/she with the most time on their hands," and can heavily disenfranchise everyone else. Whether that time is because someone's paid to work on Drupal, or because they're single and/or don't have a lot of outside hobbies, or because they are living completely unsustainably and putting Drupal over their own mental health and family (I have been guilty on all three counts at varying times :P~), people with more time tend to have a louder voice in the project. It's really tough for people like e.g. parents, freelancers, hobbyists, who simply can't put in those kinds of hours to keep up. I think the more we can do to make it easier for people with limited time to jump in and find relevant info/contribution opportunities (e.g. the old "topic pages" from the Prairie Initiative), the more we can help offset this.
posted by melissam at 2:37 PM on November 19, 2013 [3 favorites]


I understand the desire to write code for the sake of writing code but much of the time this is some seriously mundane shit that you ought to be paid for. Most people aren't making huge breakthroughs with huge dividends, they're just enjoying the process of going through the motions, finding more elegant approaches, learning new languages, etc. Ripe for exploitation. Many of those passionate programmers become dispensable dime a dozen suckers to be chewed up and spit out. We're creating value and if you believe your product has quality, get compensated.

I understand the desire to know that somebody is passionate but in many cases they really want to know how far they can push you before replacing you with another young buck. People who figure out work life balance and mature with time, learning why it's not at all good for them to be 18/7+ code monkeys are going to be better leaders, rock stars, etc. Of course you can't tell this to a hot shot. In a way their aspirations are as unlikely as joining the NBA or being a famous musician, it just so happens to be a great day job. Get paid. Open source has huge value but don't be a buster ass mark.
posted by lordaych at 3:32 PM on November 19, 2013 [1 favorite]


And I think a lot of devs who are underpaid professionally lack confidence in their work and accept a lower salary because of it, then reframe it as "I love to do it, just pay me whatever!" Really you are saying you're desperate and your work doesn't demand fair compensation. If you are getting a break and proving yourself, do it. Then get paid. You'll find out how invaluable you are when you demand more scratch and show them how you've paid your salary many times over already in productivity.
posted by lordaych at 3:37 PM on November 19, 2013


« Older From Afghanistan to Christmas Island   |   Push My Buttons Newer »


This thread has been archived and is closed to new comments