On the sickness of our community

Posted Oct 9, 2014 12:12 UTC (Thu) by samth (guest, #1290)
Parent article: On the sickness of our community

Count me as another lomg-time LWN reader who is disappointed in this article. Lennart makes excellent points about the problems that prominent people in our communities create, and this article basically denies that this could happen. Corbet has basically always been a defender of the Linux community against all enemies, but a true friend has to be willing to admit when there's a serious problem.

to post comments

On the sickness of our community

Posted Oct 9, 2014 12:24 UTC (Thu) by corbet (editor, #1) [Link] (24 responses)

I am sorry that people are disappointed. Let's just say it would have been a lot easier to just let this go by and not say anything.

But look: over the years we have built a community that lets many thousands of people from an incredible range of backgrounds work together to build a common resource. It is cooperation and collaboration on an unprecedented scale. In my mind, to paint the whole thing as "quite sick" is simply wrong.

I guess I don't know how many times I could say "yes we need to do better" in the article without burying everything else.

On the sickness of our community

Posted Oct 9, 2014 12:43 UTC (Thu) by niner (guest, #26151) [Link]

The point people seem to take issue with is the "those people are not part of our community" stance. It probably should have been "those people _should not be_ part of our community". By painting it like they already are outside, you deny that there is a problem within the community in direct contradiction with the "yes we need to do better" part.

Calling it "a sick place to be in" is obviously going to far. But that was just a headline. While reading some parts of Lennart's or Kathy's posts, the words seem appropriate though. I understand that you wish to remind of the good parts of the community at the same time. Using those to set standards for _all_ parts of the community would still recognize the accomplishment while making absolutely clear that there are still issues to address.

On the sickness of our community

Posted Oct 9, 2014 14:01 UTC (Thu) by rsidd (subscriber, #2582) [Link]

Jon, I am sure you posted this article fully expecting the reaction, which would be at least double the average Lennart-related article. Which means you thought carefully about it. I have my views on Lennart's post, let me just say I find nothing wrong with yours.

On the sickness of our community

Posted Oct 9, 2014 16:59 UTC (Thu) by bronson (subscriber, #4806) [Link] (1 responses)

Jon, I'm sure you didn't intend this, but I also came away from the article feeling like it was trying to circle the wagons. You bring a number of examples of problematic behavior from outside the kernel community but no examples from within. You even say that it would (probably) be a mistake to assume the attacks could even originate from within.

I can't say I agree with that.

Maybe now is not the time but I hope at some point you'll give more coverage to the more negative side of the LKML discussion. Rather than acknowledging that we could do better (obvious and always true), it would be nice to hear patterns of what's wrong, and what could be done to make them better.

We're not all saints here.

On the sickness of our community

Posted Oct 9, 2014 18:14 UTC (Thu) by josh (subscriber, #17465) [Link]

This is the best summary I've seen of the fundamental problem with this article; thank you, bronson. "circle the wagons" is exactly right; this article comes across as defensive and reassuring.

On the sickness of our community

Posted Oct 9, 2014 18:36 UTC (Thu) by daney (guest, #24551) [Link]

I am not disappointed, I rather liked the article.

Certainly most of your articles are of a technical nature, and as such are not open to criticism of the type seen here. But when you include what is essentially an editorial comment, I don't think it is unreasonable to expect heated discussions and disagreements with what you are saying.

The people that are mostly in agreement with what you said, probably weren't irked enough to post any responses.

As somewhat of a side note: It seems that some treat victimhood as a binary state. You are either a victim, or you are not. Thus, Blaiming The Victim becomes a black and white issue. In the real world there is sometimes a grey area, and this may produce differences of opinion.

On the sickness of our community

Posted Oct 9, 2014 23:59 UTC (Thu) by samth (guest, #1290) [Link] (6 responses)

You could have said what other people such as Lennart and Sarah Sharp have said, which is the the repeated abusive behavior by leading members of the Linux kernel community, starting with Linus, is a problem. That you didn't do this, and instead defended Linus, is what disappointed me and others.

More generally, I think the attitude embodied in: "A common response to such a person would be to flame them to a crisp in the hope that they simply go away." is a serious issue. The idea that "flam[ing] to a crisp" is an acceptable and normal way to treat people is the root of the problem here. I realize that this is very much a part of the heritage that the free-software community comes from. But that doesn't make it not a problem, and one that we need to fix. And it's a problem that can't be addressed as long as people like Linus (or many others) continue to behave the way they do.

On the sickness of our community

Posted Oct 10, 2014 3:56 UTC (Fri) by raven667 (subscriber, #5198) [Link] (5 responses)

> The idea that "flam[ing] to a crisp" is an acceptable and normal way to treat people is the root of the problem here.

This may just be a cultural and generational shift as the younger generation have different social norms than the older one, norms which don't include flaming people on mailing lists.

On the sickness of our community

Posted Oct 10, 2014 4:20 UTC (Fri) by rahulsundaram (subscriber, #21946) [Link]

There is also the fact the internet is a lot less U.S centric and male dominated than it used to be and part of being inclusive is to treat people with more courtesy and assume goodwill. Flaming people in public in LKML can cause serious repercussions in one's job as a kernel contributor in some places for instance. Onlookers who might be thinking about contributing are less likely to do in such an environment as well.

On the sickness of our community

Posted Oct 10, 2014 4:32 UTC (Fri) by samth (guest, #1290) [Link] (3 responses)

I used to be the sort of person who enjoyed flaming people on the internet -- it can be both cathartic and fun. But it's not a good thing to do, and it actively excludes people from our communities. So I don't do it anymore. People can change their behavior.

On the sickness of our community

Posted Oct 10, 2014 8:51 UTC (Fri) by tao (subscriber, #17563) [Link] (2 responses)

I love writing flames, but I (almost) never post them. It's rather cathartic.

Whenever I see something I find really stupid, I write a long rant about it, then proof-read it, then finally discard my reply.

After that I write (or not, depending on whether I have any input that actually hasn't been repeated 100x already) my real reply and submit that instead.

Works surprisingly well.

Imagine how much calmer mailing list discussion and online forums would be if there was a "Please wait 5 minutes, re-read your post (and the post you reply to), wait 5 minutes more to have the send button activate" policy...

On the sickness of our community

Posted Oct 10, 2014 18:30 UTC (Fri) by k8to (guest, #15413) [Link]

I tend to find those "wait to send your reply" controls make me annoyed, and if already annoyed make me more annoyed. So I suspect the results would be mixed.

On the sickness of our community

Posted Oct 11, 2014 5:26 UTC (Sat) by apoelstra (subscriber, #75205) [Link]

I do this too. Even LWN's "Preview Comment" requirement with no timer at all is sufficient for me to reconsider what I wrote. Probably 90% of my posts here I decide aren't sufficiently original or non-obvious (or polite :P) to be worth sending.

And I always wonder whether the editors can see posts that were previewed but not posted..

pretty nice babies, and awful, vile bathwater

Posted Oct 10, 2014 6:17 UTC (Fri) by louie (guest, #3285) [Link]

So, Jon, I hear what you're trying to say, and as I said in a comment above, there is an almost miraculous amount of good in what we, as a community, have done. It can be, I think, positive to talk about what we get right even as we engage in self-criticism about what we get wrong. So you're right to say "hey, let's not throw the baby out with the bathwater" - the baby is pretty great!

But you have to do the self-criticism. Or to put it another way, the piece seems to be in denial about the state of the bathwater. Let's be clear-the bathwater is very, very dirty, to the point it risks killing the baby.

Your comment here doesn't help. As I said above, it is great that we have a community of thousands of people which is often very helpful and constructive, but there is no plausible way you can say those people are "from an incredible range of backgrounds". The vast, vast majority of us are from a very narrow range of backgrounds, hitting most or all of the privileges listed in this article- grew up with computers, great fluency with language, schools with nerdy friends, etc., etc. Not coincidentally, we're also overwhelmingly white and male. This is partially related to various "pipeline" problems, but given how much worse it is in open source than programming at large, it is also almost certainly related to the problems identified by Lennart and others of late.

So, I hear you about celebrating what we have done; I do think many (probably including Lennart) have gone too far in implying that there is no value in it. But please, step back and take a more serious, thoughtful look at the shortcomings. The first step towards solving these problems is for leaders like you to be honest with themselves and with others about what the problems are. They run much deeper than this article seems to admit. :/

Sick indeed

Posted Oct 14, 2014 8:15 UTC (Tue) by man_ls (guest, #15091) [Link] (10 responses)

This is very hard to articulate for me and I hope it comes across. The problem is that our esteemed editor, a well-mannered, educated, white person finds no problems in the LKML community while engaging in uncontroversial projects, and 99% of the time there are no problems. But what drives people away is the 1% of jerky reactions, and they are more visible if they come from the top (i.e. Linus). I've seen healthy dev communities and this is not one of them.

It is not surprising that the cadre of LKML regulars (or Linus lieutenants) don't find issue in the community, but I think they are not looking hard enough. Until they have the guts to tell Linus: "Hey boss, this is not a nice way of dealing with people", or do the same with any other devs that are not polite to people, it will continue to be a community of bros with a dominant male and a cadre of lieutenants (which I fear is what they want it to be). There, I said it.

Sick indeed

Posted Oct 14, 2014 9:26 UTC (Tue) by anselm (subscriber, #2796) [Link] (1 responses)

It is the leaders of a community who very often set the tone for the rest. If a community leader comes across as an arrogant asshole, this tells other people in the community that (a) it is OK to be an arrogant asshole in that community and (b) this is what the leader does so if you want to be more like him or her then try to emulate that. Conversely, if the leaders emphatically aren't arrogant assholes then that gives the community the required leverage to tell any other arrogant assholes to cut it out.

The problem is that, especially in free-software projects, if the project leaders foster a culture of arrogant assholery it is very difficult for others in the community to tell them to tone it down, since (a) they're usually the ones who first came up with the project and do most of the work, and (b) since they're arrogant assholes they generally don't like to be told how to behave, especially by incompetent wimps who are too stupid to see how the leaders walk on water, and ought to get out of the kitchen if they can't stand the heat. Often the community contains a crowd of sub-assholes who, simply to show off their own prowess at being arrogant assholes, will gang up on and incinerate anyone putting forward the concept that perhaps it might be bad to be an arrogant asshole. This drives away the more reasonable people and thus makes the problem self-reinforcing.

Sick indeed

Posted Oct 14, 2014 10:09 UTC (Tue) by man_ls (guest, #15091) [Link]

Yes, I suppose that it could be said as simply as "leaders lead", but your analysis is much more interesting.

Also, people with status in a community tend to like it that way. In turn this means that they don't see any glaring problems (at least glaring to the rest of the world), even people who emphatically are _not_ assholes, such as our esteemed editor. They tend to rationalize abuse, sometimes as in this case as being something uncommon and due to extraordinary circumstances. But again, this behavior should _not_ be acceptable, even once a month.

Sick indeed

Posted Oct 14, 2014 21:33 UTC (Tue) by neilbrown (subscriber, #359) [Link] (4 responses)

> Until they have the guts to tell Linus: "Hey boss, this is not a nice way of dealing with people"

Quite a number of people have said that to him in a variety of ways. He appears to be impervious.

Just don't invite him to conferences and don't send your patches to him.

Sick indeed

Posted Oct 17, 2014 0:20 UTC (Fri) by Wol (subscriber, #4433) [Link] (3 responses)

From what I gather, when Linus does behave like an a....., two things are almost invariably true.

1) He's dealing with one of his own lieutenants, who he knows well.
2) His lieutenant is being somewhat of an idiot.

If they aren't true, then he's dealing with someone who has a massively puffed up sense of his own importance. It was put quite well somewhere - if you are a manager in a company, you will have about 10 direct reports. Any more and you're overloaded. Linus has THOUSANDS (okay I exaggerate a little :-) of people who would *like* to be direct reports. And I believe he is on record as saying that he has found in the past it is often useless to ask these people politely to p*** o**. Blunt, cruel language is often his tool of choice because it's the only tool that works. The alternative is the kill-file. And if the guy is a good guy naively trying to short-circuit channels, then the kill-file is the wrong choice ...

Cheers,
Wol

Sick indeed

Posted Oct 17, 2014 4:02 UTC (Fri) by raven667 (subscriber, #5198) [Link]

I think that you are probably right about the assertion that most of the time when Linus is incandescent it is with someone he knows personally, maybe any problem is the public nature of the performance. If he dressed down his lieutenants in private maybe the mailing list would have more civility and there wouldn't be the concerns that are being expressed now.

Sick indeed

Posted Oct 17, 2014 23:53 UTC (Fri) by neilbrown (subscriber, #359) [Link] (1 responses)

> From what I gather, when Linus does behave like an a....., two things are almost invariably true.
>
> 1) He's dealing with one of his own lieutenants, who he knows well.
> 2) His lieutenant is being somewhat of an idiot.

Can you provide a single example when this is the case?

Also "being ... an idiot" is a somewhat pejorative term and quite unlikely for so-called "lieutenants". "Mistaken", "Misinformed", "Careless" are all quite likely. Whether such behaviour deserves such treatment is, I guess, a matter of opinion.

The most recent example cringe-worthy behaviour from Linus that I have seen was his first line in

https://lkml.org/lkml/2014/9/29/824

That is certainly not a aggressive as some, but it made be cringe, was not aimed at a "Lieutenant" (as I would use the word) and was not a case of idiocy on the developer's part.

Sick indeed

Posted Oct 18, 2014 0:18 UTC (Sat) by anselm (subscriber, #2796) [Link]

That is certainly not a aggressive as some, but it made be cringe, was not aimed at a "Lieutenant" (as I would use the word) and was not a case of idiocy on the developer's part.

It's also an example of the “I'm right, you're wrong, go away“ attitude so often ascribed to Lennart Poettering.

Sick indeed

Posted Nov 2, 2014 1:52 UTC (Sun) by blujay (guest, #39961) [Link] (2 responses)

> The problem is that our esteemed editor, a well-mannered, educated, white person

What in the world does his skin color have to do with anything? Why is it in vogue to make everything an issue of skin color? Skin color was less of an issue 20 years ago than it is today! Racism won't die because people won't let it!

Sick indeed

Posted Nov 2, 2014 12:16 UTC (Sun) by mathstuf (subscriber, #69389) [Link] (1 responses)

The skin color here matters because while Corbet may see little to no issue (according to man_ls at least), those who have grown up with society already having the deck stacked against them through casual racism, sexism, etc. (nothing usually so overt as the KKK, but, for example, arrest statistics, salary statistics, etc., say *something* is skewed here even though we may not agree on the root causes) may not accept abuse in a venue such as LKML.

Sick indeed

Posted Nov 3, 2014 23:23 UTC (Mon) by blujay (guest, #39961) [Link]

... What? What you just said makes no sense and has nothing to do with Jon Corbet's skin color. You're just furthering racism. Skin color is completely irrelevant here. Stop trying to make everything an issue of skin color!

On the sickness of our community

Posted Oct 9, 2014 12:25 UTC (Thu) by timtas (guest, #2815) [Link] (18 responses)

And as usual, he completely fails to take any responsibility himself, just blaming Linus Torvalds. All we get from him is the usual "We are creating something really special and are moving forward in a pace that maybe is too fast for some lame old bags" bla bla bla.

On the sickness of our community

Posted Oct 10, 2014 2:08 UTC (Fri) by misc (subscriber, #73730) [Link] (17 responses)

Saying that he doesn't take responsibility imply that Lennart should take responsibility for something, but you do not say why, neither explain why he would be responsible for anything.

Are you trying to just convince people there is a problem with Lennart just by repeating baseless accusations, hoping that since everybody say it, it must be true ?

As that's one of the method discussed on the quoted article of Kathy Sierra, that would be a bit unfortunate to not realize that maybe you are playing the game that a potential sociopath want you to play. If you pride yourself to be technical, maybe you could then give facts rather than just repeating rumors, because this make you feel good.

For example, showing someone having a bad effect on community by pointing how that person is mimicing Lennart behavior, since that's the whole point of his post, and you think he should take responsibility for that. Also, pointing where he do that, and since everybody repeat it, it shouldn't be hard to find at least 5 or 6 occurrences, as you are for sure not basing your perception on this one bad day where someone was tired and said something a bit more harsh than usual.

On the sickness of our community

Posted Oct 13, 2014 12:24 UTC (Mon) by nim-nim (subscriber, #34454) [Link] (16 responses)

Yesterday on a radio show for kids one of them asked why it was necessary to be polite. The guest scientist then took pains to explain politeness was not hypocrisy, it was a way to show others your regard, and when people attempted to get by without those archaïc useless norms (at a guess in the 60's or 70's) it didn't end well.

Now Lennart is the archetypal example of coder prima dona who does not bother to be polite and will ruthlessly trample over other people feelings if he thinks he is technically right. That is why he generates such feelings even in "normal" people, nothing more, nothing less.

On the sickness of our community

Posted Oct 13, 2014 13:43 UTC (Mon) by zdzichu (subscriber, #17118) [Link] (13 responses)

Now Lennart is the archetypal example of coder prima dona who does not bother to be polite and will ruthlessly trample over other people feelings if he thinks he is technically right.

Please provide links to back this, it's an empty accusation.

On the sickness of our community

Posted Oct 13, 2014 16:12 UTC (Mon) by nim-nim (subscriber, #34454) [Link] (8 responses)

journald, json as a configuration language in pulseaudio, etc

But, you already knew all that, and I have better ways to spend my time than arguing with a fan club. misc wrote a reasonable message, I answered it, that's enough as far as I am concerned.

On the sickness of our community

Posted Oct 13, 2014 17:48 UTC (Mon) by rahulsundaram (subscriber, #21946) [Link] (7 responses)

I am not sure how your reply addresses your claim. Technical choices for a configuration format has nothing to do with not being polite or trampling over other people's feelings.

On the sickness of our community

Posted Oct 13, 2014 18:00 UTC (Mon) by nim-nim (subscriber, #34454) [Link] (6 responses)

It was treated the usual I am RIGHT everyone else is WRONG LP way (with "look how cool LP is" bystanders). json is gone in systemd without any apology (but more abrasiveness on other format alternatives)

On the sickness of our community

Posted Oct 13, 2014 18:56 UTC (Mon) by rahulsundaram (subscriber, #21946) [Link] (2 responses)

Well, if anyone didn't think their choices were right, they wouldn't be making the choices they do. I don't see why a change in configuration format in an entirely different software requires an apology or to whom. Sorry but I still don't have much of an idea what you consider a problem.

On the sickness of our community

Posted Oct 13, 2014 19:01 UTC (Mon) by dlang (guest, #313) [Link] (1 responses)

> Well, if anyone didn't think their choices were right, they wouldn't be making the choices they do.

There's a difference between thinking that their choices were right and being unwilling to consider that they may be wrong.

good programmers, and especially good project leaders need to be open to the possibility that they may be wrong and listen to feedback.

On the sickness of our community

Posted Oct 13, 2014 19:05 UTC (Mon) by rahulsundaram (subscriber, #21946) [Link]

> There's a difference between thinking that their choices were right and being unwilling to consider that they may be wrong.

Sure and systemd has changed a lot based on external input. Abstract conversations OTOH don't accomplish much.

On the sickness of our community

Posted Oct 13, 2014 19:52 UTC (Mon) by johannbg (guest, #65743) [Link] (2 responses)

The polypaudio/pulseaudio community needed a way to describe some digital audio properties in a fairly standard and extensible way that wouldn't need protocol updates in the future and they settled on JSON.

As far as I know it was not a sole decision made Lennart so it would be good if you could actually reference that discussion when he declares that.

You need to be a bit more spesific what you mean by "json is gone in systemd without any apology"

On the sickness of our community

Posted Oct 14, 2014 12:53 UTC (Tue) by cortana (subscriber, #24596) [Link] (1 responses)

I don't even see any JSON in /etc/pulse, ~/.config/pulse. Nor has, to my memory, systemd ever used JSON except as one of the export formats in journalctl.

On the sickness of our community

Posted Oct 14, 2014 13:33 UTC (Tue) by johannbg (guest, #65743) [Link]

You should be directing these comments at nim-nim where I in the comment you responded to am asking nim-nim to clarify and reference to what he meant by Lennart solely deciding upon JSON in pulseaudio as well as what he meant by JSON being gone in systemd without any apology.

On the sickness of our community

Posted Oct 16, 2014 20:11 UTC (Thu) by flussence (guest, #85566) [Link] (3 responses)

I recall seeing a Gentoo package maintainer getting some quite nasty verbal abuse from LP and his lackeys in an IRC channel in response to legitimate technical questions, and I'd be happy to link the page if not for the fact said developer's blog archives seem to cut off at 2012.

There's evidently been no attempt to fix the underlying bad attitude in the few years that've passed. What surprises me is that it's taken this long for a major backlash to happen.

On the sickness of our community

Posted Oct 16, 2014 20:16 UTC (Thu) by rahulsundaram (subscriber, #21946) [Link] (2 responses)

Yet another unsubstantiated and unverifiable claim.

On the sickness of our community

Posted Oct 17, 2014 17:06 UTC (Fri) by flussence (guest, #85566) [Link] (1 responses)

I retract that claim; after several hours of digging through archive sites it wasn't LP who made that specific attack after all, but De Icaza and the Mono crowd. Their similarity must've confused me. Apologies go to those whom it's due.

On the sickness of our community

Posted Oct 17, 2014 18:16 UTC (Fri) by jspaleta (subscriber, #50639) [Link]

The retraction is a good first step. But I require more from you.

I require that you also affirm that you will defend "LP and his lackeys" as you call them in the future if you see anyone making a similar unsubstantiated claim of abuse and actively deflate the meme that "LP and his lackeys" are extraordinarily abusive. If you can't affirm to do that in the future and if you choose to stand silently by while the meme continues to spread you are still complicit in helping spread it with your mistake here. It's like lighting a fire and saying you're sorry, admitting you set the fire but still standing there watching it burn a house down. Time for you to pick up a bucket of water and do help put the fire out.

Oh and stop calling other developers lackeys. That'd be pretty good too.

On the sickness of our community

Posted Oct 13, 2014 13:48 UTC (Mon) by anselm (subscriber, #2796) [Link] (1 responses)

Now Lennart is the archetypal example of coder prima dona who does not bother to be polite and will ruthlessly trample over other people feelings if he thinks he is technically right. That is why he generates such feelings even in "normal" people, nothing more, nothing less.

I don't think that this is actually the case. People don't seem to mind being treated the same way by Linus Torvalds and other icons of the community who are just as unwilling to suffer fools gladly. It also seems to be difficult to actually come up with concrete examples where Lennart Poettering was really rude (ruder than, e.g., someone like Linus would be) to somebody on a systemd mailing list. Which does not detract from the fact that it is a convenient meme to circle the wagons around even if there is no actual proof, and so it gets trotted out over and over again. Now as a matter of principle it would be great if on the whole there was more politeness around but if you think Lennart Poettering is rude then I would recommend you stay very far away from the likes of DJB or Theo de Raadt. In addition, there are people in systemd development who are way less abrasive, and while Lennart P is getting his attitude adjusted (or not) it might make sense to talk to them instead.

The main problem many people seem to have with Lennart Poettering is that he had the chutzpah to write PulseAudio and then not to bend over backward in order to clean up or work around other people's shit (such as bugs in other people's ALSA drivers). This attitude together with the fact that some distributions shipped PulseAudio while it was quite new and the bugs in the ALSA drivers in question had not yet been fixed gave him a bad reputation that was, if anything, only partially deserved, but is still good enough for “Lennart broke our systems with PulseAudio and is now going to break them again with systemd”.

On the sickness of our community

Posted Oct 13, 2014 14:41 UTC (Mon) by johannbg (guest, #65743) [Link]

What is even worse is that the Alsa community has had the opportunity to stand up and correct this misconception towards pulseaudio and Lennart but instead have chosen to bury their head in the sand and turn a blind eye when individuals are cyberbullying Lennart in that regard which is quite frankly disgusting and arguably one of the reason this misconception about pulseaudio and Lennart has escalated this far in the first place...

One thing that those cyberbully's need to start understanding is that Lennart is and never has been acting alone but I guess it's easier for them to single out and attack one individual from within group of developers or a community regardless if he had any saying or doing or involvement in whatever they are trying to blame/frame him for, than the entire group or the community itself.

On the sickness of our community

Posted Oct 10, 2014 18:04 UTC (Fri) by Tet (subscriber, #5433) [Link] (1 responses)

a true friend has to be willing to admit when there's a serious problem

Agreed. But in this instance, I think Jonathan's got it right. There isn't a serious problem. A minor one? Maybe. But there really isn't anything major here.

On the sickness of our community

Posted Oct 10, 2014 18:10 UTC (Fri) by rahulsundaram (subscriber, #21946) [Link]

Are you sure about that? This is hardly the first time people have complained about such behavior

https://lwn.net/Articles/417952/

If you look further, you can find a lot of examples.