Linus is one of the few brilliant jerks out there who is truly brilliant. Unfortunately his antics have made it OK for lots of less talented individuals to think they have the right to treat people like dirt in order to write good code.
He's exceptional in my opinion but I often wonder what the OSS community would be like without behavior like his. I recall not wanting to major in CS when I entered school because the only programmers I knew when I was young were from newsgroups, forum, and OSS communities that I perused as kid...and they all seemed like assholes. I'm glad I changed by mind cause I was so interested but thick skin shouldn't be the bar for contributing to tech.
edit: I'm certainly not defending Linus. His talent makes him a net positive but he definitely should not treat people the way he does.
Flipping off a bad hardware company, writing some scathing roasts of patches... he's obviously not checking himself into staid polite behavior, absolutely. The extent to which this makes him a jerk, or makes him bossy is a connection that I've always felt is tenuous at best.
Even his meanest patch roasts are things that I'd hope people realize can be taken in stride; technical work will have to go in but work can progress, and Linus is not forebearing, he's not enhancing the difficulty of doing that technical work by being directly scathing to the existing work, and generally (what I think most people regards as) humorous and spirited to the rest of the readership.
What I haven't seen is any real attempt to find citeable referencable direct impacts of Linus being flamboyant as he produces his technical outlook. Which contributors have taken serious injury, and how has it impacted them? Certainly contributing is hard and finding out the work needs to be redone is hard: what measures can we assign to how much harder those necessary acts are when it's a "jerk"-y delivery (which as I said is not a term I'm not confident is appropriate or applicable). Rather than focus on Linus's behavior as problematic and the emphasis of it being a deeply stemmed root cause, I'd like to see some hard impact-analysis with good data behind it that gives a better idea of why we should adopt strict tone-policing of the necessary harsh and brutal technical review process.
Patch roasts aren't the issue; roasting contributors is. Telling someone their code is awful is not normally a problem. Telling someone they personally are awful for even having submitted it, and that they should have been "retroactively aborted" or any number of worse statements, is absolutely ridiculous.
Suggested fix: just use the 'seq_printf()' interfaces, which do the
proper buffering, and allow any size reads of various packetized data.
Of course, I'd also suggest that whoever was the genius who thought it
was a good idea to read things ONE FCKING BYTE AT A TIME with system
calls for each byte should be retroactively aborted. Who the fck does
idiotic things like that? How did they noty die as babies, considering
that they were likely too stupid to find a tit to suck on?"
- linus https://lkml.org/lkml/2012/7/6/495
do you want some data on the people who don't even bother dealing with sort of talk? he could easily replace the second paragraph with something like
"This is an extremely bad way of doing things; reading one byte of a time is terrible. Please research the system calls you are using before submitting another patch, it's unacceptable to submit code that hasn't been thought through thoroughly."
"Key, I'm fcking tired of the fact that you don't fix problems in the code you* write, so that the kernel then has to work around the problems you cause[...]I will not be merging any code from Kay into the kernel until this constant pattern is fixed.[...]This has been going on for years, and doesn't seem to be getting any better."
"What I mind is people closing bugs and not admitting mistakes. If Kay had even said "sorry, the excessive output was a bug in systemd, it's already fixed in current -git", that would have been a valid reason to close the bug. And for the people thinking this is a storm in a teacup: this is not the first time Kay has done this, which is why I personally get so frustrated. Kay has done the exact same thing with major bugs that were not fixed anywhere else, and that caused machines to fail at boot time, and Kay happily pointed the finger elsewhere for months at a time and closed bugzilla entries.
[...]
people are (I think) understandably upset that systemd not only screwed up, but then the people involved weren't even willing to say "sorry" about it but instead go "uh, it wasn't our bug, deal with it".
Suggested fix: just use the 'seq_printf()' interfaces, which do the proper buffering, and allow any size reads of various packetized data.
Of course, I'd also suggest that whoever was the genius who thought it was a good idea to read things ONE FCKING BYTE AT A TIME with system calls for each byte should be retroactively aborted. Who the fck does idiotic things like that? How did they noty die as babies, considering that they were likely too stupid to find a tit to suck on?" - linus https://lkml.org/lkml/2012/7/6/495*
do you want some data on the people who don't even bother dealing with sort of talk? he could easily replace the second paragraph with something like*
"This is an extremely bad way of doing things; reading one byte of a time is terrible. Please research the system calls you are using before submitting another patch, it's unacceptable to submit code that hasn't been thought through thoroughly."
I dont know what I feel about this particular missive. But I do have some points to add about it-
First and foremost, if I interpret this correctly there are two different issues Linus is conveying:
1. Kay has let a subsystem he has (shared-?) responsibility over decay from it's normal functions.
2. Debian does whack bad shit to read kernel logs.
So, Kay is not the target of abuse. People who are not kernel devs are indirectly being targetted for insane usages of system facilities. I believe most laymen might mistakenly interpret the email by Linux as being directed at Kay. Only the first two paragraphs have a direct pertinence to Kay. The rest is saying why it's silly that Kay should have to do the work at all. I don't believe Kay is traceable to Debian's almost-certainly longstanding use of dd bs=1 .
Second, Kay's level of commitment ranks much beyond a "contributor" "contributing." He has volunteered to maintain subsystems, volunteered to maintain udev. He is running his part of the show, and is responsible for ensuring those realms are working ok. That relationship is far different than that of a contributor in that Kay has directly taken up a distinct responsibility to be on top of things.
Third, I think it's funny and hilarious. I thought it was funny and characterful the first time I read it. And I'd misread it too: I thought it was all packaged against Kay, laying the single-byte read WTF usage at his feet.
I didn't think, "Oh Kay, what a horrible unabsolvable should-be abortion! I hate him!," I didn't think "GAWDD this guy is so lame, fuck Kay!," I thought "there's probably some really freaky old code from way the hell back when that has some really dumb stupid reason for doing things in a safe 1 at a time way."
Engineers are expected to have empathy. That's the core lesson of all of this. And empathy for happenstance, for the weird misconfigurations that have arisen themselves is the chief empathy our ilk endures and battles with. We have great powers but we are still tied to a shamblingly mortal system of seeming immortal and unchanging nature. The conflict of empathy here is not with each other, it is with circumstance and what has already accrued, and people who are actual makers all can relate and understand- Linus's rage is heavy here because the very bad very old reasons are things which dog us still, and all of us who push are healed by hearing the drive and push towards betterment and away from _that,_ (a non-specific, non-direct that-which-haunts-us, even in the case of /proc/kmsg now). Linux capabilities unlike shipping code don't get post mortems- this bad capability will soldier and go on. But Linus and others give us an outlet where we can empathize with the desire to get beyond, to be better, even as Kay stays up puzzling through seven headfuls of concerns to hack a path forwards.
I WILL NOT DRAW JUDGEMENT EITHER WAY on Linus's choice of writing here. I think it foolish to have conviction this is or is not acceptable behavior. It's deeply human, it's speaking to his base, it's speaking to his true beliefs and assessments and his feelings and passion all, it's sharing a deep seated completely justified mortal pain. Could he have channeled all that out in another fashion? Perhaps. But the words you wrote sbilstein are hollow, mechanized, shallow productions: they are what a machine would produce. I'll leave it for others to debate the merits and disadvantages of the "professionalism" and "responsibility" and "respectful leadership" that must balance against the more charged and unrestrained tact Linus opted to take, but I would say absolutely that as an engineer I feel better about my own empassioned pitched battles with ugly reality by getting to see Linus try to channel his feelings and assessment out, not packaged for efficient mechanized high-compression high-efficacy conveyance, but as someone giving a shit about decisions and what gets built.
It always seems like jerks are better at handling other jerks than non-jerks. A jerk-led project may be more sustainable for that reason in the long run. I can think of numerous situations where contributing jerks (in the minority) did more to harm a project than the person leading the project.
I found this well worth reading. Note the section "Ideas and code are still fair game", which points out that eliminating bad behavior from a community does not prevent project maintainers and leaders from taking good code and complaining about bad code, as long as they don't tear down people in the process. Frequently, when this issue comes up in the Open Source community, it produces flippant responses about how behaving better would prevent telling people their code or even their core idea is broken.
I believe that while Linus' attitude could be better, it is understandable. Not that I'm excusing it, it's just that almost all the times you see someone pointing out him being an asshole, you can go back up the email chain to find the person he's being an asshole to being obtuse. Combine this with a deep care that Linus has for his "baby", and tons of crap trying to get past the radar on a regular basis, and it's understandable his reactions are what they are.
Similar situations arise in communities where denial and/or subterfuge are common tactics, such as climate change denial and creationism. Just look up the term "sea lioning".
That being said, I am for more diverse opinions, which I think you only truly get when you have more diverse points of view, which is why I agree that prejudice is a very bad thing. What is interesting to me, though, is I have yet to see prejudice from eg Linus. He's mean, yes, but he tends to "spread the love" (as the kernel management guide puts it) fairly evenly and not based on prejudice. Of course, I might just not be seeing this because I'm a white straight cis male.
Last but not least, I agree that ideas and code are fair game for criticism. Unfortunately, sometimes people associate with their ideas (and sometimes code too) and take it as a personal insult when their idea or code is criticised.
Nice. Ad hominem. What's next, are you going to call me misogynistic? How about homophobic? Would you be happier if I phrased it as "I'm not excusing his behavior"? Because that's essentially the same meaning, just not as "weaselly" [sic]. Also, sentence did not contain a "but", even semantically.
Once upon a time, I worked for IBM. At this time and in this place, decisions were made at IBM based not on evaluating ideas, but on who talked loudest. I watched a lot of good ideas get shot down, or more frequently just ignored, because nobody vocal felt like supporting them or because someone vocal decided the idea wasn't in their best interests. I also watched a crap-ton of really shitty ideas rise to the top and get a fucking load of money wasted on them because the set of people who can identify good ideas and the set of people who are loud do not have a very large intersection.
Fortunately for IBM, it's a large company with nearly unlimited resources and its position as the most successful technology company will never be seriously threatened. Particularly its office systems, AIX, and PowerPC directions are world-leading.
(For those of you with Asperger syndrome, that last paragraph was sarcasm. And yes, I'm looking squarely at Larry Loucks.)
Your case is fair, and your advice is solid, but no one is sure it's appropriate. At IBM there is a distinct chain of command and people who have to interact with each other and a corporate body every day. A technical mailing list can certainly suffer from domineering people monopolizing the conversation as well, but most evidence we have of this- Systemd debates in Debian, Gentoo- show a much more overt structural deficiency going on in a far more pervasive fashion than what is found on LKML.
Look at the outcome in problem cases- whereas in a contained social environment- a business- the containment ends in creative control ending up in the lands of the domineering group, in mailing lists we tend to rampant infighting and shifting lines of battle amongst larger parties. The 1:1 1:few defined hierarchies and political power-games of a business have few clear places we can draw parity to on a highly-trafficed mailing list: dissent resisting the domineering force is highly visible and contested, and conflict spreads as people weigh in and the conversation threads-out and threads are linked into one another.
You make a very good point, but I think you may be over-weighting the "chain of command" aspect of the corporate body and under-weighting the "flame-fests are fun, but somebody has to make a decision" aspect of a mailing list discussion.
My personal favorite result of the latter are trigraphs in the C standard. Nobody liked them, but nobody argued strongly against them because they were introduced so late.
I'm all for civility, but unfortunately it seems the people who are complaining about the loud assholes are attempting to conflate "being rude" with "being anti-feminist" and therefore "being a racist misogynist".
I'm not about twisting peoples words to push the feminist agenda.
edit: please HN, enlighten me before you down vote. Are you guys against the civility part or do you disagree that people are conflating rudeness with being racist?
Overall, I don't think this post is very good. It tries hard to ride the wave of several important, valuable, and popular sentiments but I think that the ways in which it misses the mark, mars its purpose.
I don't understand why Nick (the author) chose to group a bunch of unrelated stuff in to an otherwise pretty reasonable post on decreasing personal attacks in OSS communications.
The parts about systemic biases and asking speakers about them has essentially nothing to do with the first part of the post. Essentially the entire "What Can We Do About It" section reads like it was written for an entirely different purpose than the rest of the post. It is fine to promote those ideas but to tie the idea of increasing civility in communications as an idea of countering systematic biases against certain demographics is a faulty alignment of ideals.
Saying read a wiki on feminism as a way to reduce personal attacks in email threads is bizarre. The notion that people are "experiencing harassment over your identity rather than being critiqued solely based on the quality of your work." is not related at all to my experience in OSS communities.
I disagree. Men in our society are taught to be emotionally abusive in their relating to others. Environments where emotionally aggressive communications are the norm have a whole lot to do with gender roles in our society. And how someone deals with such an environment is effected by their gender socialization too, such environments are perceived and dealt with differently, as a generalized trend, by women and men. (yes, individual men and women can be on both ends of it, at different times).
> Men in our society are taught to be emotionally abusive in their relating to others.
My parents sure didn't raise me that way. Honestly, a large reason why so many people are aggressive when they communicate is simply because they just don't know how to communicate. It's not a male thing, nor is it a female thing.
It's a "I'm unable to express my unhappiness regarding this scenario in a well-thought out and respectful way, so I'm going to lob some insults at you to make my point" thing.
My dad used to say that if you need to swear then you've obviously lost the ability to express your opinion in an articulate, and polite way. I don't 100% agree with that statement, but it does have some truth to it when you apply it a bit differently.
If you can't tell somebody their code is bad without berating them for even being born it's not because you're a man or a woman, you either have a condition which prevents you from recognizing you're being a dick or you do not know how to communicate like a civilized human.
Being assertive, constructive, and listening well are three ways to be able to communicate like a decent person. Being subservient or aggressive, destructive, and failing to listen are three/four ways to communicate like a person with a lack of social skills.
Torvalds is being dogpiled . This is the third or fourth article I've read in the last month that has tried to conflate Linus Torvalds sometimes being rude with the horrendous, persistrent, ongoing harassment/abuse that is being carried out by completely different people, in completely different industry.
If Torvalds is going to be used, is it too much to ask to stick to the things he has done and not lump him together with death threats, SWATing, etc? Like, catalog what HE said, in their context, and discuss those? Because I really don't think they are comparable, and the conflation is disingenuous.
That is one of the scariest aspects of all this: there is an aggressive Internet mob that does nothing of value for Free SW, but will come in full force any time that an issue like this one is debated in the open, and will stop at nothing to ruin the life of those they identify as enemies.
Encouraging civility and attracting new contributors are very important goals for the future of Free SW. Achieving that will help us create better products that can be useful for more people.
Therefore, having all this harassment and hate come up whenever there is talk about improving the community is tremendously hurtful. It is sabotage, that's what it is.
Since he isn't a nice person and likes arguing, in his own words, why would he care? And why are so many white knights worried about his precious fe-fes? That's what Linus is arguing for, feelings don't matter, only ideas, right?
As far as I can tell, the abuse all originated from a board on 8chan. As I said, I that is so far away from what he does that I don't even see them on the same continuum. Also,
>his silence implies approval.
No it doesn't. He probably doesn't even know about this. If he did, have you considered he might be scared to speak up?
I agree. This is a click bait title. The words abuse and harassment should not be used about as freely as they are in current meta-discussions of online discourse.
If we look at posts of Linus he is an equal opportunity jerk. He bashes people solely because they have made a bad commit or suggested something absolutely braindead even if they should have known better, regardless of their gender.
You can say he is jerk but you cannot say he is misogynist.
"Equal opportunity jerk" seems alright at first thought, but when you consider that people of different races/genders have different social standing, it's clear that some people will be disproportionately affected by others being "equal opportunity jerks".
Plus, being an "equal opportunity jerk" is not a good policy at all — it's downright abusive. No one should be verbally berated in public for writing code that's not up to the maintainer's snuff.
Niceness is about worrying about how other people feel. Sometimes caring about other people's feelings (especially in the short term) can lead to duplicitous, decietful, dihonest, fraudulent, counterproductive or wasteful behaviors.
Being "good" in a sense is optimizing against those behaviors, which usually is an orthogonal concern to niceness, but sometimes leads to distinctly being not nice. I can't say one way or another if Linus is overoptimizing, but given a choice between niceness and goodness, I pick goodness.
> Niceness is about worrying about how other people feel.
I really disagree. Niceness is about CONSIDERING people's feelings. Not overly worrying or bending to them or self censoring but just being aware of your effect on other people. Linus seems to be aware but thinks that considering other people is a waste of time and I think that is a pretty toxic attitude.
I think you're putting the emphasis on the wrong word here. The important thing here is not worrying/considering but "feel".
I would guess that Linus makes a categorical distinction between emotional and physical welfare - because emotional welfare is ill-defined and easy to manipulate from the 'reciever's' point of view, he considers it at worst a waste of time or at best a secondary consideration after definable things, like achieving a software objective.
> Sometimes caring about other people's feelings (especially in the short term) can lead to duplicitous, decietful, dihonest, fraudulent, counterproductive or wasteful behaviors.
I think this is a lie that liars tell themselves. When you say you are lying to spare someone else's feelings -- what you really care about is your own feelings, protecting yourself from having to deal with their reaction to the truth.
I think it's lying and lack of concern for the truth that actually lead to counterproductive and wasteful behaviors. Have you ever worked somewhere where nobody trusts each other and nobody tells the truth about what's going on? I think many of us have. How's that situation for productivity?
Neither the OP nor anyone else is asking Linus to lie about anything.
>Instead, what I do care about, passionately, is helping the best ideas win (where I include "feasible" as part of my definition of "best"). Not the "best ideas from people willing to tolerate extensive personal abuse".
really? another diatribe about Linus? get over it. Linus is the way he is. dealing with him requires thick skin, and he knows it, and we know it, and anyone working with him needs to deal with it.
do you think complaining about him on the internet is going to get Linus to change? what's the point here?
there's also an awful lot of false equivalency and inappropriate analogy going on here. I don't think Zoe Quinn and Anita Sarkeesian have ANYTHING WHATEVER to do with Linus' communication style. Invoking their names here is an effort to summon a twitter shitstorm and its pretty low in my opinion.
I agree in part, as much as someone can hate Linus' style, it's worthwhile to have a toolset to work with people that operate on slightly different social scripts [0] that you want to.
You call this having a thick skin, but I think there is a distinction between not taking offense when directed your way, and not finding the statement offensive in the first place. There is a sliding scale of what levels of language imply what levels of offense and it changes depending on context. And this is just one of many properties in evaluation of environment.
I've started thinking participants in social environments have an at-least-4-tuple for every context: their desired, expected, perceived, and actual social setting. And a tuple exists for every possible context.
The thing that makes for a hostile environment isn't that Linus' posts are "No Big Deal" or a "Big Deal, Indeed", but that people disagree in their desires on how the setting should treat a Linus rant. Especially when there are people that not only expect their opinion to win, but make it imperative to their politics that the culture adapt to them rather than the other way around.
Programming is the most tame, calm, and soft-spoken professional environment I've ever been in. I'm used to adjusting my expectation for social scripts for different environments. I've learned to be comfortable in environments with innocuous swearing and ranting elsewhere, I can take it here too. In the end I don't really care which way we go, but the fighting is certainly getting on my nerves.
really? another diatribe about Linus? get over it. Linus is the way he is.
What a sad, accepting attitude. The fact is that people are more aware than ever than his behavior is unacceptable. They talk about it more and more often every year, and that trend will continue. Eventually, he won't be able to make a public appearance (in person or online) without getting called out on his completely unnecessary and hostile behavior. Then he'll either change or be shunted into irrelevance, and it will be a net positive for the Linux community.
Fame and accomplishment are no license to be uncivil.
he's not uncivil. he's just blunt. the fact is, Linus does the right thing when it matters. you might not like the way he writes emails on the developer mailing list. you might think he swears too much or uses too much hyperbole. it doesn't matter though.
what you're advocating here is censorship and ostracism of one of the most important people in the FOSS world. think about that for a few minutes. you're advocating censorship and ostracism. who's uncivil now?
edit: and now you've downvoted me for disagreeing. more "civil" behavior being demonstrated here.
It does matter if it means otherwise qualified contributors stop contributing or, more likely, never attempt to contribute in the first place. Linus doesn't know how to act any other way, but that doesn't mean there aren't other ways to act that don't harm the overall success of the project but create a more welcoming atmosphere.
It's like he's trapped in this weird world where "being nice" is the alternative to being a temperamental jerk. How about being kind, patient, but forceful? Lots of technical leaders know how to strike this balance and Linux is worse of for Linus' inability to do so.
I'm tired of seeing civility and toxicity being mixed up. Blogs about Linus is just one example of this type of discourse, which comflate being nice and civil with mysoginy, racism, and the like.
Linus is not toxik.
Furthermore, there's this notion that being nice is a free, costless attitude. Just as Nick Coghlan himself admits, being nice is hard and demand energy and time. All things that come in limited quantity. There is also an unspoken assumption in these rant that Linus is constantly being a bully. It's just not true. The fact of the matter is, in a forum, you need to be able to use a large range of attitude to deal with the wide range of people and contribution. You can nicely rebute someone on a subtle topic. It's wasted time to try to be subtle about boneheaded ideas. If you can't go termonuclear from times to times, then your opinions get muddled. The fact is, it is hard to convey the level of opposition by text alone. Using bluntness works. You also have to deal with really disruptive people.
You just can't rule out never to be uncivil. It's a tool, to be wielded when needed. There is hypocrisy in demanding not to be called an idiot when you are.
Is any attempt to get people to consider changing their behavior to be dismissed as censorship and ostracism? We should never be able to say to anyone "Hey, you're being a jerk. It would be better if you were less of a jerk."?
(Is this one of those ridiculous "If you're so into tolerance, why don't you tolerate my intolerance?" arguments?)
the post i was responding to said "Linus should be shunned into irrelevance". that's a fair bit different than simply asking someone to change their behavior.
And if he doesn't change his behaviour, which is increasingly deemed unacceptable, we should continue interacting with him and pretend it is acceptable?
What you're essentially saying is that it doesn't matter how much of an asshole someone "important" is to others, but that it's ABSOLUTELY TERRIBLE for anyone to dare call him out on being an asshole. Do you really think any community benefits from this attitude?
think about that for a few minutes. you're advocating censorship and ostracism. who's uncivil now?
No, I'm saying that uncivil speech has consequences, and that's a desirable feature of existing in a society. In order for the predictions in my post to not come true, you would have to somehow prevent people from discussing Linus's incivility. What mechanism do you propose?
Also, I can't downvote you because it's a reply to a post of mine (this is how HN works). I generally don't downvote posts I reply to either.
I don't take anything seriously that tells me to review the "geek feminism" wiki.
These "geek feminists" tell people they have to follow a moral code they make up on the spot then when someone says something they deem is "inappropriate" they bully them and accuse them of horrible things that aren't true. This actually happens and isn't debatable.
What I personally really don't like about these people is that they go after open source specifically, which leads me to believe many of them have never worked for a large company but only have experience in the open source sector, which by it's nature allows anyone to participate.
I mean, accusing open source of being intolerant or whatever is like an alien race accusing the human species of being intolerant because they observed that --yes-- when you allow people to live in a society freely some people make decisions others (gasp) disagree with.
These geek feminists should all go close their source and only allow people to work with them who they deems is tolerable. A Big corporation --no matter how progressive they may seem from the outside-- is always going to have someone working for them that says some stupid shit. Look at the racist emails in the wake of the sony leak from some of their top people.
I don't want a monoculture of people pretending to be nice. Blaming open source is just a way to push an agenda.
If this "geek feminist crowd" wants to fork off and do their own thing and base it around being nice to each other then it's just a matter of time until their top devs get exposed for being human and they look like idiots. The geek feminist community seems to contain more assholes than any open source community I've ever been a part of, in fact I've worked with a few feminists who have said some very nasty things about people of a certain color. Unfortunately, the JavaScript crowd seems to have a lot of feminsits and I had the displeasure of working with a JS developer who was a rich white straight male feminist asshole google employee. He would talk about how hard minorities have it on his personal blog and then get on github/irc and treat everyone like shit. Pretty typical geek feminist in my experience.
I really don't think you have a very good handle on what you're railing against. Especially because at the end you conflate diversity in FOSS with criticism of Linus' attitude. This just reads like you have a really big axe to grind against your reading of feminism.
I'm not the one whose conflating diversity with Linus' attitude, that would be the feminists.
edit: please hn, tell me where I'm wrong.
My whole point is that making this a matter of diversity is wrong and only done to push a political agenda.
It's not Linus who is saying anything against any race or gender. He's being mean and then some people calling themselves "feminists" are coming in and saying "being mean to people is so much harder if you're a minority, so therefore open source is racist".
I'm not conflating diversity with anything. Just pointing out that that's what the geek feminists are doing.
I'll give it a shot -- upthread someone makes the statement It always seems like jerks are better at handling other jerks than non-jerks. Assume for the moment that this is true, then combine it with the ample research revealing that in workplace situations, men are respected for "jerk"-like behavior and women are looked down on for it. Now you have a situation in which being a jerk creates an environment in which women are less likely to succeed. This is a pretty simplistic example, but a good representation of the type of syllogism that can lead you from "abusive/uncivil behavior" to "anti-diversity behavior". Hope this helps!
He's exceptional in my opinion but I often wonder what the OSS community would be like without behavior like his. I recall not wanting to major in CS when I entered school because the only programmers I knew when I was young were from newsgroups, forum, and OSS communities that I perused as kid...and they all seemed like assholes. I'm glad I changed by mind cause I was so interested but thick skin shouldn't be the bar for contributing to tech.
edit: I'm certainly not defending Linus. His talent makes him a net positive but he definitely should not treat people the way he does.