On kernel mailing list behavior [LWN.net]

On kernel mailing list behavior

Posted Jul 17, 2013 19:12 UTC (Wed) by Tobu (subscriber, #24111)
Parent article: On kernel mailing list behavior

I have had rather positive experiences interacting with kernel maintainers; my only want is for more of it, which must be the common case for contributions at the periphery. And whenever Linus pops into some discussion I'm reading, he communicates clearly and well and brings something to the discussion. So I'm trusting the kernel community, and Linus particularly.

Since Sarah posted her thread, which broke off from the recent stable discussion, the opinion I've made is that she's working really hard at getting offended. The “What was said” paragraph in this article is clear enough, but I've noticed another striking example on her blog comments.

As far as diplomacy and communication skills are concerned, I'd rather interact with a Linus than a Sarah, and I'll look up to the former and not the latter.

Maybe the flip side of a passionate advocate is someone who's at least a little bit self-righteous and prone to black and white thinking. Anyway, I'd prefer good examples to opinion leaders, from what I've seen so far.

to post comments

On kernel mailing list behavior

Posted Jul 17, 2013 21:21 UTC (Wed) by khim (subscriber, #9252) [Link] (11 responses)

The “What was said” paragraph in this article is clear enough, but I've noticed another striking example on her blog comments.

I believe this is the proper link.

And yes, it really sad that people like Sarah who tend to thwart honest discussion (because noone is allowed to talk honestly near them because they can easily be offended) are valued more in today's world then people who actually do honest work.

Note that both examples which blog post points to (this one and that one) are so awful I don't know how people can blame Linus for strong words.

Try an experiment:
1. Pick any busy interstate highway.
2. Drive to place near police booth.
3. Turn the rear speed and go back for awhile.
4. Hear polite speech from policeman with nice explanation for why what you are doing is wrong.
5. Compare it to what Linus have said.

I bet you'll hear a lot of similar words - and for similar reason.

On kernel mailing list behavior

Posted Jul 17, 2013 21:34 UTC (Wed) by corbet (editor, #1) [Link] (1 responses)

A lot of folks will think it's not relevant, but it's still amusing to consider that the second of Linus's rants (the one toward Rafael) was a complete mistake on his part. The regression he was cursing about did not exist...

On kernel mailing list behavior

Posted Jul 17, 2013 21:51 UTC (Wed) by khim (subscriber, #9252) [Link]

Sure - and both sides apologized in the end. Rafael apologized for his "but it's a fix" explanation (which is quite obviously is not a justification for regressions) and Linus apologized for his rant.

But Linus reacted on the facts he had at the time and it surely looked to him like a totally crazy idea: we don't tolerate regressions is pretty well-established rule of kernel development and only few very serious things can affect it. "But there is ready-to-use-exploit in the wild" may qualify while "but it's a fix" does not - and Rafael should know that, he's not a newbie at this point!

On kernel mailing list behavior

Posted Jul 17, 2013 23:24 UTC (Wed) by tbird20d (subscriber, #1901) [Link] (8 responses)

And yes, it really sad that people like Sarah who tend to thwart honest discussion (because noone is allowed to talk honestly near them because they can easily be offended) are valued more in today's world then people who actually do honest work.

Careful there. You're implying that Sarah thwarts honest discussion and doesn't do honest work. Are you familiar with Sarah's work in the kernel?

Frankly, if it were an the option I'd much rather submit code to Sarah than to Linus - not because I think she'd be easier on my code, but because I think she'd be more constructive with her criticism.

On kernel mailing list behavior

Posted Jul 18, 2013 1:46 UTC (Thu) by Tara_Li (guest, #26706) [Link] (5 responses)

*shrugs* Then encourage Sarah to start her own fork of the Linux Kernel.

Heck, we've had plenty of cases in the past where it became obvious that Linus doesn't scale well, and where it showed that his interest in the Kernel per se have faded. If Sarah asked, Linus would probably be *glad* to hand ultimate control over the kernel community to her - except that he can't.

Linus controls, in a sense, everything about the community *except* one detail - the fact that he's the one in control. That is decided by the community itself, based on his track record. And complaints about his civility are hardly new. But when all is said and done, people trust Linus, most likely because they know what he says is what he considers the unvarnished truth. He has proven able to admit when he's wrong, but he's also been proven right an insane amount of the time.

Perhaps Sarah should be discussing her opinion of Linus' civility with Ms. Amanda McPherson, Ms. Angela Brown, Ms. Jennifer Cloer, or Ms Karen Copenhaver. They're part of the Management Team that actually pays Linus at the Linux Foundation - so you could say they're his enablers.

On kernel mailing list behavior

Posted Jul 18, 2013 11:09 UTC (Thu) by simosx (guest, #24338) [Link] (1 responses)

It is actually Jim Zemlin who is at fault here.

Have a look at his mystifying talk at http://www.linux.com/news/featured-blogs/185-jennifer-clo...

He looks surprised that even if things are done /wrongly/ at the Linux development, the Linux kernel appears to come out OK.
His last point was *Don't be nice*, and he was referring to this exact issue we are discussing here. Jim, as an executive director, seems unable to reign or influence the core kernel developers.

What Jim is missing is that the Linux kernel is used in almost all devices. No matter what deficiencies exist in the development, at the moment there will be people around to replace the disaffected.
The Linux kernel development sets the tone to other projects, and you can see how it works with other projects were they lack the contributors.
This immature attitude in the Linux kernel development makes most core development to happen by paid developers and the volunteers are doing peripheral work.

On kernel mailing list behavior

Posted Jul 18, 2013 22:12 UTC (Thu) by nix (subscriber, #2304) [Link]

You appear to have no understanding of how kernel development works. Jim Zemlin is not a king. He does not 'reign'. He is not even Linus's boss in any real sense: he pays Linus, but if he tries to tell Linus what to do, and Linus disagrees with it, Linus could jump ship somewhere else in a microsecond. Linus has so much bargaining power in this relationship that in a very real sense he is *Jim's* boss -- Jim's job, like the organization he heads, wouldn't exist without the kernel Linus has spent the last 22 years on, after all.

(This is true of major contributors to other projects too. You think someone could order the GCC developers to do something seriously stupid? No, they'd jump ship if it came down to it, and they'd even know where to go: one of the other companies employing GCC hackers. To know what those are, they only need to be able to read the bit after the @ on the development mailing list they follow every day.)

On kernel mailing list behavior

Posted Jul 18, 2013 11:10 UTC (Thu) by simosx (guest, #24338) [Link] (2 responses)

> *shrugs* Then encourage Sarah to start her own fork of the Linux Kernel.

I think this is called 'being patronizing'.

On kernel mailing list behavior

Posted Jul 18, 2013 16:48 UTC (Thu) by Tara_Li (guest, #26706) [Link] (1 responses)

No, it's the whole point of open-source licensing. If you don't like *ANYTHING* about the code, the process, whatever - you are free to take it and do something else. In the case of the GPL2, you can fork the project, keep on building from that base, and cherry pick things you find useful from the other guy's tree. Of course, they're cherry picking things they find useful in return - again, go back to RMS who started this whole mess. If, as Sarah says, civility in the environment is so important - her branch will take off and fairly quickly overshadow the Linus-lead branch.

Though in truth, I don't think that would happen.

I'm of two minds on the civility bit...

It eases things along, lubricating where people's egos rub together, but it also seems to reduce bandwidth, increase latency, and all too often fail integrity tests, requiring resending of the message and all too often slipping around integrity tests, contaminating the communications with falsehoods.

On kernel mailing list behavior

Posted Jul 19, 2013 20:13 UTC (Fri) by daniel (guest, #3181) [Link]

No, it's "being patronizing" just as GP said.

On kernel mailing list behavior

Posted Jul 18, 2013 3:28 UTC (Thu) by nevets (subscriber, #11875) [Link] (1 responses)

> Frankly, if it were an the option I'd much rather submit code to Sarah than to Linus - not because I think she'd be easier on my code, but because I think she'd be more constructive with her criticism.

I've never worked with Sarah because I don't do USB work, so I can't comment on what it's like to work with/for her. But I've have worked with Linus, and I have been on the receiving end of his rants. Most of that was actually in private emails and it was more misunderstanding between the two of us.

That said, when Linus has commented on my work, I have found it to be actually very constructive. More constructive than pretty much any other developer I worked with. There has been times that I have disagreed with him (on minor things), but he definitely knows how to run the Linux community, and I really believe he's the main reason for most of the Linux kernel success.

On kernel mailing list behavior

Posted Jul 19, 2013 10:17 UTC (Fri) by daniel (guest, #3181) [Link]

"he definitely knows how to run the Linux community"

Well... except Linus was a bit of a git with the Bitkeeper thing, which was hugely and unnecessarily divisive for the community. Then there was the IDE thing. And a few other things, just a few isolated smudges on an otherwise legendary record.

On kernel mailing list behavior

Posted Jul 19, 2013 14:28 UTC (Fri) by kh (guest, #19413) [Link] (5 responses)

> As far as diplomacy and communication skills are concerned, I'd rather interact with a Linus than a Sarah

I really agree. I am not even sure what Sarah is complaining about, at first I thought it was salty language, but there are many examples of her using the same. Then I thought she was complaining about threats of violence, but I don't see how anything Linus said could be taken as a serious threat. On the other hand, I definitely would be afraid to say /anything/ to Sarah after this, between her equating Linus's emails to real threats and misogyny on her personal blog (where there is no real connection), and the threatening behavior of her supporters who seem to wish to shout anyone down who says anything positive about Linus, or public calls to employers to muzzle Linus and his supporters.

It seems to me like this was something Sarah had planned in her head for quite some time, and she got impatient waiting for Linus to really explode, so decided to jump on any little statement, blowing it out of proportion as necessary. (I also think equating this to sexual violence in the military really belittles the terrible things that some have gone through while serving their country.)

On kernel mailing list behavior

Posted Jul 19, 2013 18:49 UTC (Fri) by raven667 (subscriber, #5198) [Link] (4 responses)

> I really agree. I am not even sure what Sarah is complaining about,

I believe you that you are coming by your confusion honestly as many people are confused the same way.

The complaint by Sarah is abusive behavior by Linus against others, not herself, abusive behavior including being personally insulting, calling people names and other public berating and shaming of contributors.

> On the other hand, I definitely would be afraid to say /anything/ to Sarah

Sarah is not a shrinking violet and is certainly capable of speaking her mind and interacting with the LKML community, however she would seem to be a poor target for abusive comments or threats. Were you intending to make abusive comments to others online? Even on LWN that really isn't permitted, Editor #1 will step in for people who get too far out of line.

> threatening behavior of her supporters

I haven't seen any threatening behavior by supporters so I don't know what you are talking about, or maybe we have different ideas of what we find to be a threat, but I should point out that supporters are independent actors, such as yourself or myself and their behavior is their own responsibility.

> wish to shout anyone down who says anything positive about Linus, or public calls to employers to muzzle Linus and his supporters

I would ascribe that behavior to those who argue that Sarah shouldn't be allowed to express her opinion and should keep quiet, not those who argue for professional, adult discourse when speaking in a public mailing list professionally, to other professionals. The only shouting and calls for muzzling I have seen are towards Sarah, not Linus.

> It seems to me like this was something Sarah had planned in her head for quite some time

I wouldn't presume some Machivellian malevolent intent, I would start by presuming that Sarah is a human being and is motivated by much the same things that any other person. She saw some behavior that she thought was BS and instead of letting it slide this time, called Linus out on it.

This is a worthwhile discussion to have and we are now having it.

On kernel mailing list behavior

Posted Jul 19, 2013 19:03 UTC (Fri) by kh (guest, #19413) [Link] (1 responses)

No need to look far:

https://lwn.net/Articles/559434/

On kernel mailing list behavior

Posted Jul 19, 2013 19:30 UTC (Fri) by raven667 (subscriber, #5198) [Link]

I'm sorry, what are you expecting me to see here?

On kernel mailing list behavior

Posted Jul 19, 2013 19:31 UTC (Fri) by Cyberax (✭ supporter ✭, #52523) [Link] (1 responses)

> Were you intending to make abusive comments to others online? Even on LWN that really isn't permitted, Editor #1 will step in for people who get too far out of line.
Our dear leader #1 mostly steps in to kick sexist trolls. And whatever Linus is, he's definitely not a sexist idiot.

On kernel mailing list behavior

Posted Jul 19, 2013 20:23 UTC (Fri) by raven667 (subscriber, #5198) [Link]

While there is one example of this in recent memory, #1 has stepped in and handed yellow cards to several different instances of unacceptable behavior, not just for gross sexism.