GitHub's Code of Conduct
matthewhopkinsnews.com
matthewhopkinsnews.com
>Our open source community prioritizes marginalized people’s safety over privileged people’s comfort. We will not act on complaints regarding:
> ‘Reverse’ -isms, including ‘reverse racism,’ ‘reverse sexism,’ and ‘cisphobia’
> Reasonable communication of boundaries, such as “leave me alone,” “go away,” or “I’m not discussing this with you”
> Refusal to explain or debate social justice concepts
> Communicating in a ‘tone’ you don’t find congenial
> Criticizing racist, sexist, cissexist, or otherwise oppressive behavior or assumptions
It's one of the only CoC I've ever read that specifically condones some forms of harassment. No harassment should be acceptable. I've already switched to gitlab and I'm about to go through the list of projects that use this CoC and make sure I don't use their products anymore.
That doesn't mean a project should tolerate any form of hate. It's simply a statement that a code of conduct is not to be subverted and used as another weapon rather than as a form of protection. Also notice that it talks about the "safety" of one group versus the "comfort" of another, not the "safety" of one group versus the "safety" of another. The latter would be quite different.
If you like, look at it as a statement about Bayesian priors: "what situation is far more likely, and should be the default assumption?"
Here's a better replacement that addresses all concerns:
Attacking someone based on their real life identity is not allowed, regardless of what that identity is and furthermore is off topic as this is a technical project, not a sociological one.
It has the same effect of banning all -isms by default, is clear enough and unambiguous enough for even a troll to understand and not being able to rules lawyer their way out of (people's identities are off limit, end of discussion), and keeps off topic nonsense out of tech projects.
Because a few moments of thought to compose a reply will produce a better result than people who have been working on codes of conduct for years and have seen the stunts people try to pull?
Among many other problems, not all forms of unwelcoming behavior have to do with "attacking someone based on their real-life identity". A statement like that might be a start.
And in the process, you've also managed to dismiss any and all discussions of how to positively improve a project to better address social issues, or how to fix existing problems. Congratulations, your code of conduct itself makes people unwelcome.
Of course not - but it would slot nicely as a replacement into the controversial piece of the Github CoC, which is kind of what I was aiming at. Under what circumstances would a contributor's real life identity ever be relevant in the context of a ticketing system?
you've also managed to dismiss any and all discussions of how to positively improve a project to better address social issues
Most of the time, those social issue are irrelevant. We're not talking about behavior on a chat channel, or a conference, or a mailing list, we're talking about behavior on a bug tracker.
What social issues are not being addressed on a bug tracker that are relevant to the projects in question?
Simple amount of effort doesn't guarantee a good result. I'm sure we can all name software projects that added feature after feature as years slipped by and ended up as a bloated, counterproductive mess, much like the CoC we're debating.
>And in the process, you've also managed to dismiss any and all discussions of how to positively improve a project to better address social issues, or how to fix existing problems. Congratulations, your code of conduct itself makes people unwelcome.
Actually, since you mentioned it: why is it the obligation of every random hobbyist open source project to "better address social issues"?
For the most part I'm talking about issues those projects already have, such as being hostile to new users or contributors, and especially to various specific groups. As for "why": much like "Free Software" versus "Open Source", there are ideological reasons and practical reasons. A sample of practical ("Open Source"-style) reasons: because you want more users and contributors, because you want contributors who don't all think like you do, because less energy spent dealing with awful people is more energy spent hacking, and because communities that people have more fun in are communities that grow much faster. Additional ideological ("Free Software"-style) reasons, which apply whether the rest do or not: because being decent to everyone is not too much to ask, and shouldn't have to be asked for in the first place.
Those are admirable goals, but a) they have nothing to do with free software and b) it is not established that these codes of conduct, especially when they enshrine fringe political viewpoints and explicitly racist and discriminatory behavior as this one does, will achieve those goals.
Conversation A:
Person 1 (repeatedly) makes discriminatory comments towards person 2
Person 2 makes only non-discriminatory comments back
Conversation B:
Person 1 (repeatedly) makes discriminatory comments towards person 2
Person 2 same as above, but eventually makes a discriminatory comment back
In conversation B, there was no reason for person 2 to make the discriminatory comment other than their emotions got the better of them and they were dragged down to the level of person 1. But I see no reason why this should be forgiven. Person 2 still made discriminatory comments which violated the code of conduct. We could argue over the appropriate level of retribution they each deserve, but I don't think they should get a free pass. Giving them one encourages a "fight fire with fire" attitude which simply leads to more discrimination.
The solution in my eyes is to have a clear definition of discriminatory/harassing comments and enforce it consistently. Stronger repercussions for repeat offenders, warnings for the occasional slip up.
https://github.com/domgetter/NCoC
..at least when it comes to the hacker ethos, which is something, IMO, we should all strive for. Think of it as the WTFPL of the "thou shalt not" world.
More to the point, I really fail to understand why documents like the one Github is working on are necessary. Most FOSS projects I'm aware of are quite adept at dismissing dickery in all its forms, rather than adopting a futile (and controversial, and agenda-driven) approach to defining it.
(Huh, someone already did something that's a lot more like the WTFPL. Predictably, it's rather acerbic: https://github.com/ciafwywcoc/ciafwywcoc/blob/master/CODE_OF...)
Depends on the projects you work with. Some are good at maintaining a great community (Rust, for instance), and some are quite awful at it, or actively cultivate a hostile community (by welcoming awful people into it with open arms).
There are enough awful people who know how to code that it's worth stating up front that "this project doesn't accept awful, hateful people, no matter how interesting their technical contributions". Because otherwise you end up with projects where "well, we'd kick them out, but nobody else knows how to do what they're doing, so...".
See also http://www.slideshare.net/dberkholz/assholes-are-killing-you... .
Here's my personal litmus test: Would you accept Linus Torvalds on your project? He's known for his technical ability, but also known for taking little to no crap and calling out said crap with extreme prejudice.
If the answer is yes: You clearly value technical ability and Getting Shit Done® over all else.
If the answer is no: You clearly value social communication over Getting Shit Done®.
Neither of these options are completely right or completely wrong, but again, I lean towards the hacker mindset. Getting stuff done should be more important than not offending people, IMAO :)
Which is exactly why codes of conduct do so at length, rather than assuming everyone will agree on a "simple" definition.
> This is not a winnable battle. Some people find being told "no" to be oppressive and horrible.
They'll get over it or go somewhere else. That's not what codes of conduct are there for, and that's exactly why this issue came up in the first place: some people use a code of conduct as a weapon when they're told that their actions are not acceptable.
> Would you accept Linus Torvalds on your project? He's known for his technical ability, but also known for taking little to no crap and calling out said crap with extreme prejudice.
Sure, anyone is welcome, right up until the point where they make themselves unwelcome, at which point they don't get a free pass for it for any reason. Even the kernel community seems to be getting somewhat better in that regard, though sadly not nearly fast enough.
Note also that criticizing code is not criticizing people. See https://www.youtube.com/watch?v=_O2zgNfvsh0 for a great presentation on the various levels of awfulness that often appear in a project.
> If the answer is no: You clearly value social communication over Getting Shit Done®.
http://www.slideshare.net/dberkholz/assholes-are-killing-you...
One person can drive off many potential contributors, and a steady flow of contributors is often more critical to a project than any one person.
I could say the same thing of people who don't like the atmosphere on a given project. There are plenty of projects with plenty of different kinds of communities.
Sure, anyone is welcome, right up until the point where they make themselves unwelcome
Please answer the question as asked and in the spirit it was intended in.
One person can drive off many potential contributors
And one bad CoC that legitimizes certain kinds of bad behavior can do the same. The backlash speaks to the truth of that.
You keep linking that slideshow - not one of the claims made there are sourced or backed up. Considering one slide goes on about the importance of providing numbers, this is especially galling.
More of a general question: Is anyone aware of a FOSS project which was killed due to a lack of contributors brought on by the perceived negative atmosphere of the community?
And doing so means you're perpetuating the problem.
> There are plenty of projects with plenty of different kinds of communities.
And codes of conduct help create more projects with welcoming communities.
> Please answer the question as asked and in the spirit it was intended in.
I did. Contributions welcome; hateful behavior not. Doesn't matter who the contributor is, they'd be told their behavior was unwelcome, and removed if persistent.
> And one bad CoC that legitimizes certain kinds of bad behavior can do the same. The backlash speaks to the truth of that.
Every time I look at this supposed "backlash", I see piles of people I don't want anywhere near any project I care about. So it sounds like the CoC is doing its job.
> You keep linking that slideshow - not one of the claims made there are sourced or backed up.
There are myriad sources and firsthand accounts available if you look, and the numbers from that particular presentation are (to the best of my knowledge) based on a direct study.
> Is anyone aware of a FOSS project which was killed due to a lack of contributors brought on by the perceived negative atmosphere of the community?
Off the top of my head, glibc was forked to create eglibc largely because of the (now former) maintainer of glibc, and many distributions switched over to eglibc as their primary upstream. Shortly after that person was replaced as the maintainer of glibc, eglibc got merged back in and distros switched back to glibc as upstream. Many more cases where that came from.
What problem?
Why the double standard? Why is your CoC fine and dandy and mine is not? What of the people who don't want to waste time litigating sociology and who want to code? Why can they just "go away" as you say and people who want to muddy coding with other non-coding interests not?
My entire point was that there are plenty of communities that don't need this, and so far, you've not shown that the problem it sets out to solve actually exists outside of a few well-publicized outliers.
>I did. Contributions welcome; hateful behavior not.
So is Linus' behavior hateful or not, by your definition?
>There are myriad sources and firsthand accounts available if you look, and the numbers from that particular presentation are (to the best of my knowledge) based on a direct study.
Which study? There are many, many studies that detail the behavior in online communities (and given my recent courseload, I'm familiar with more of them than I'd like ^^). Which one were those numbers pulled from? Nobody knows, because the author didn't bother to include that data.. This is why cites are important.
Random example, if you want one likely to apply to an issue tracker: http://www.kalzumeus.com/2010/06/17/falsehoods-programmers-b... Reporting a bug about fixing such an assumption will almost certainly involve some discussion about identity. And if you tell people to shut up because their concerns aren't technical, the bug won't get fixed, and the next bug like it won't even get reported because you've made it clear that you don't care.
Another example: submitting a patch to change "master/slave" terminology into "primary/replica".
Another example: submitting a patch to avoid assuming male pronouns in documentation, in an effort to not make the default assumption that the reader is male.
> So is Linus' behavior hateful or not
When he's creatively explaining how a piece of code is broken, there's nothing wrong with that. When he's telling people that they should have been "retroactively aborted", yes, that's hateful. As mentioned, I've seen a lot less of that lately.
> Which one were those numbers pulled from?
Life isn't Wikipedia; direct primary research is perfectly acceptable.
The same is thought across corporations: process (an equivalent for potential contributors) is valued over any single employee; corporations don't like to rely on a sole worker. And guess what? It doesn't work in favour of innovation or good craftsmanship.
It may be somewhat different in open source projects, as contributing is way easier and requires much less commitment than changing jobs (the nature of things can change in surprising ways as the numbers grow), but personally I doubt it.
Perhaps someday, if all these various efforts are successful, it won't be necessary to have such things. But it'll be a lot easier to get to that point if every attempt didn't have to put up with this kind of "but my free speech!!1!" backlash.
I'm part of just about every privileged group possible. I don't need a safe space; the world is my safe space. I don't need an outreach program; I have many opportunities others do not. I don't need a group dedicated to my concerns about inclusiveness and being welcoming; I'm already included and welcome by default. The very least I can do is not complain when others try to deal with that, and to recognize that some of the advantages I have will go away in the process, and that's a good thing. Better yet, I can actually help with that whenever I can.
There are many white males on the lower rungs of the socioeconomic ladder that are effectively silenced in this discussion. I suspect a lot of the pushback you hear comes from them.
There are also minorities that have succeeded in the current system by working hard and being smart that are being told their experience is abnormal/non-existent.
I'm quite certain it is, which is exactly why I make a point of listening to people who have more of a stake in this than I do. I notice somewhat more than I used to before I educated myself on these issues, but still nowhere near as much as people who live through them personally.
> There are many white males on the lower rungs of the socioeconomic ladder that are effectively silenced in this discussion. I suspect a lot of the pushback you hear comes from them.
I suspect you suspect wrong, and most of the "pushback" comes from people indignant that their "concerns" are not at the forefront the way they are everywhere else, as well as people who see yet another opportunity to be disruptive. In any case, socioeconomic status is one of many relevant, and it's closely correlated with many other forms of privilege.
> There are also minorities that have succeeded in the current system by working hard and being smart that are being told their experience is abnormal/non-existent.
Read http://geekfeminism.wikia.com/wiki/Clawed_my_way_up , which directly responds to the argument you're making. Such experiences are not "non-existent", but they're also not a sign that there's no problem and everyone else can be safely ignored. There are a lot of other smart, hardworking people who have not had the opportunity to succeed.
Geekfeminism wiki is a political organization, they're not a source I would trust.
> Geekfeminism wiki is a political organization, they're not a source I would trust.
And with that ad-hominem dismissal of one of the most concentrated sources of useful information and references, my interest in continuing this discussion with you comes to an abrupt end.
If you want to actually convince other people, you're going to have to engage with their arguments, not just find ever more elaborate ways of ruling them out of bounds. There is no royal road to victory in a debate, there's just the long, hard slog of talking to a whole bunch of people who disagree with you.
>And with that ad-hominem dismissal of one of the most concentrated sources of useful information and references, my interest in continuing this discussion with you comes to an abrupt end.
While we're at it, "go to the website run by your political opponents and believe everything they say" also isn't going to get you very far. We're discussing a CoC that was largely written by Geekfeminism and/or its allies; it's highly likely that someone who doesn't like the CoC isn't going to be especially trusting of the people who wrote it, either.
Also, on top of that, this entire discussion is nicely demonstrating why the proposed Code of Conduct included "We will not act on complaints regarding: [...] Refusal to explain or debate social justice concepts". This kind of thing takes a large amount of time and energy.
That my point of view must be silenced at all costs is one of the main reasons why I would never support the Geekfeminism people. Contrary to what the parent comment states, he is most definitely not willing to question his own base assumptions and political goals.
There's a difference between silencing a point of view and refusing to repeatedly have the same argument. New arguments need new information, and the vast majority of arguments in that area have been had already. That's some of the primary value provided by their wiki: a repository of past arguments and refutations, to avoid having to revisit them every time.
Project mailing lists and other communication channels, for instance, are the wrong place for repeated introductory conversations demanding justifications for a code of conduct. And that's a very common problem.
Considering I didn't start with the set of base assumptions I have now, and had to learn them over time, I'm always willing to question them. That I don't personally find any of your arguments even remotely compelling towards that end doesn't make those assumptions immutable.
To explicitly state some of the assumptions that I suspect we disagree on: 1) If you refuse to exclude anyone from a project, then you're by default excluding people who are not willing to put up with a hostile environment. 2) Since you can't avoid excluding people, better to exclude people who make the project less welcoming. 3) Don't let the people you want to exclude write the policy. 4) Don't let the people who would fare just fine in a project without a code of conduct drive the requirements; get the requirements from people who wouldn't.
Apart from that, the other reason I don't particularly care about having this argument here is that it's unlikely to have as much effect as having it elsewhere.
As for base assumptions, I see we also differ wildly. Mine are:
* Code quality matters above all else.
* People that don't contribute code don't get a seat at the table.
* There's no inherent value in letting as many people as possible contribute to the project. Quality over quantity.
I'm leaving out anything having to do with feminism because we're not going to agree on anything about it.
There's another assumption you're implying but leaving out: "The set of people who show up and contribute under those conditions will inherently be the best and produce the best results". And that assumption is subject to a lot of confirmation bias; if you don't set out to question that assumption, you'll tend to get results that give you no reason to doubt it, precisely because it's harder to notice the absence of something.
Coding well is damn hard, by its very nature it's not a very inclusive thing. Very few have the capacity to do it well. I prefer to focus on them than on the masses that will never be great coders. I don't think pretending like everyone can and should be involved is useful.
I'm a minority from a poor background. No one encouraged me to code, I couldn't even afford the tools I needed. I would read computer magazines in the supermarket and write psudeo-code on paper at home. I'm now a very successful software engineer. Regardless of how unimportant you think my own experience is, it IS quite possible to succeed if you're talented and from an under represented group, no "safe spaces" needed. Just tenacity, an interest in the subject matter and self-drive.
For the people who do know how to code, there's an additional barrier of dealing with people who (for a few examples) like to shout about how anyone who could possibly write such terrible code should die to save important people the trouble of dealing with them, or who talk about "how to explain to your wife or girlfriend", or who give conference presentations featuring lingerie models and pornstars. Nobody should have to put up with that, or should be told to grow a thicker skin. Not wanting to put up with that does not mean someone doesn't know how to code. And allowing that kind of thing in your project disproportionately tends to exclude some groups more than others.
People are centralizing the philosophy behind a project more than is necessary.
As for communities, no, "fork it" is not a universal solution to problems within communities, conferences, professional events, and other environments. But even in cases where the approach finally ends up being to go start new, more welcoming environments, that's still a case for codes of conduct and similar; even if you're successful in moving a community elsewhere, if the same set (or a new set) of awful people show up with their same behavior, you need something to show what standards a community expects. And much like the Zen of Python, "explicit is better than implicit".
I'm also noticing a conspicuous lack of any response to the actual behaviors I was referring to; you're dismissing that as behavior that someone ought to have to tolerate or go start their own project. This problem ought to be as universally recognized in projects as buffer overflows are. And rather than playing whack-a-mole, it makes sense to try to get better solutions in place. For instance, getting support from major hosting sites, or otherwise creating larger meta-communities of people who agree that this is a bad idea, to avoid having a different set of standards (or lack of standards) in every project. And just as it's a bad idea to write your own software license rather than using a standard one, it's a bad idea to write your own code of conduct rather than using a standard one.
Absolutely. Why should I judge anyone? I'm not forced to use their project. Last I checked lingerie and porn weren't illegal. If someone wants to represent themselves like that, then more power to them.
Not every project is a corporate project and not all are going to come from that politically correct environment.
I would no sooner tell someone that they couldn't fill a project with sexually explicit terms than I would tell a songwriter to not use sexually explicit lyrics.
Nothing stops people from acting that way. It's not illegal, nor should it be. But I'm looking forward to a world where when someone tries everyone else just looks at them funny and dismisses them immediately, so that people stop trying.
Which brings us full circle back to the original point: given a choice (and it is a choice) between which set of people to exclude from a community, I know which set I'd prefer. You seem to have a different preference, and you're welcome to it. I look forward to that point of view losing all credibility and respect in the industry as a whole, and an increasing number of people are working towards that goal. Have a nice day.
If you're attempting to bring people around to your point of view, you're going to repeatedly have the same argument, because you're constantly going to be encountering new people who need to be convinced. If you don't want to do that, perhaps political activism isn't for you.
> the vast majority of arguments in that area have been had already. That's some of the primary value provided by their wiki: a repository of past arguments and refutations, to avoid having to revisit them every time.
To be more accurate, it's a repository of refutations that GeekFeminism asserts answer the argument. The Discovery Institute also has a repository of what it asserts are refutations of evolution, but (assuming you're not a creationist) I doubt you'd rush out and read them all just on my say-so. Same deal here.
Granted. I've certainly spent a pile of time commenting in this thread, and none of it was new. But just as it's reasonable to link to a FAQ rather than answer the same question for the Nth time, it's also not unreasonable to link to a detailed response to an argument rather than reiterating it for the Nth time, especially when it's far better supported and researched than something off-the-cuff.
> To be more accurate, it's a repository of refutations that GeekFeminism asserts answer the argument.
Sure. And if someone really wanted to try to argue against those points, by all means do so (in an appropriate forum and context, not on a project mailing list for instance). But ignoring those arguments because of the site they're on or the people who made them is a perfectly sensible reason to end the discussion, just as if someone said "I'm ignoring everything you just said, my arguments still hold, so there".
I'm disinclined to read a pile of arguments in favor of creationism for exactly the same reasons I'm disinclined to directly get into an argument with a creationist: there's no possible way it's worth my time to do so. If I've found myself making some set of poor life choices that led me to become involved in such a debate, then of course I'd read and respond to arguments made in it, as with any other discussion. I'd also be inclined to consider whether I'm primarily trying to convince that person, or primarily convince the set of as-yet uncommitted bystanders while writing off that person as likely a hopeless cause.
A plain english reading of the document suggests otherwise.
They don't say "all -isms are banned", they say "we will not act on complaints of X Y and Z". In other words, it would appear to legitimize harassment committed by non-minorities.
If they meant your definition, that definition is what should have found its way into the repository.
Put another way, if this is a legalistic document that people are going to be expected to adhere to, it needs to be as unambiguous as possible. I really don't think that's asking a whole lot.