On kernel mailing list behavior

Posted Jul 22, 2013 17:32 UTC (Mon) by Baylink (guest, #755)
Parent article: On kernel mailing list behavior

> Greg, the reason you get a lot of stable patches seems to be that you make it easy to act as a door-mat. Clearly at least some people say "I know this patch isn't important enough to send to Linus, but I know Greg will silently accept it after the fact, so I'll just wait and mark it for stable".

> You may need to learn to shout at people.

Really?

*THIS* is what set Sarah Sharp off? Then the problem is even worse than I'd imagined from the fact-free Wired article I read earlier.

The kernel is big. Really big. I mean, you might think it's a long way from one end of the PHP sources to the other, but that's just peanuts to the kernel.

And there are *lots* of people working on it.

I'm going to assume Sarah has never been in management. If she had, she would realize that there is an absolute limit to the size of a team you can manage without having to become somewhat -- or more than somewhat -- of a hardass about people and their habits and what you will allow and what exceptions you will make... and that's all we're seeing here.

Linus, having been at this for 25 years, simply doesn't sugar coat things, and since the management's all being done in public, you *see* that. I guarantee you, 2nd and 3rd tier management in traditional corporations, overseeing hundreds of professionals, is this raw, or worse, it's just that its in private, and you don't get to see it.

But no, this *particular* incident is an overreaction.

Perhaps it's a straw, and Sarah is a camel, but -- while I don't follow LKML in realtime -- I do look in on it occasionally for various reasons, and I can't say I've ever seen anything that qualifies as *abuse*.

I've myself gotten into it with Linus on that list once, and his reply didn't constitute abuse, either.

On kernel mailing list behavior

Posted Jul 22, 2013 17:52 UTC (Mon) by raven667 (subscriber, #5198) [Link] (5 responses)

> I'm going to assume Sarah has never been in management.

I don't know Sarah but I wouldn't assume that.

> ... hardass about people and their habits and what you will allow and what exceptions you will make

I don't believe anyone, including Sarah, is advocating for reducing the quality standards of the kernel.

> ... management's all being done in public ...
> ... management in traditional corporations ... is this raw, or worse, ...

While this surely happens I don't think this is as common, required or ideal. Companies with management who abuse and berate employees often end up with high turnover and stories on thedailywtf.com, not with high-quailty output. As Sarah pointed out, over 80% of kernel developers are employed in a professional capacity to do kernel development, so while everyone wants to retain the frank exchange of ideas that fosters quality, we can probably have some minimum standard of behavior that most corporations are able to maintain.

Acting decent toward your fellow human beings and co-workers shouldn't be a significant burden.

On kernel mailing list behavior

Posted Jul 22, 2013 18:08 UTC (Mon) by Baylink (guest, #755) [Link] (4 responses)

Certainly, but you, like Sarah, are assuming the consequent: *the sort of abuse she implies simply isn't happening*. Or at least, no one's demonstrated it to *me*.

Extraordinary claims require extraordinary proof, and all workplaces have standards and practices. If you walk in cold, and say "y'all are doing it wrong", you need a) a much better sales pitch and b) much better evidence than I have seen so far.

This is nothing more than the standards that kept Usenet useful for over 2 decades: be a grownup, and if you're gonna say dumb stuff, make sure your Nomex undies are zipped up first.

On kernel mailing list behavior

Posted Jul 22, 2013 18:36 UTC (Mon) by raven667 (subscriber, #5198) [Link] (3 responses)

> *the sort of abuse she implies simply isn't happening*. Or at least, no one's demonstrated it to *me*.

We can certainly have an honest difference of opinion on that subject, I don't follow LKML so take my opinion for what it's worth, I'm just going by what I've read, but she did provide some examples of what she thought was unnecessary behavior, calling other kernel developers incompetent and stupid, telling them to SHUT THE F**K UP, etc. The tenor used between close friends may not be appropriate when used publicly with what are effectively subordinates, co-workers and strangers. The community on LKML are not just people shooting the breeze, they are largely paid professionals whose livelihood depends on their successful participation, which changes the standard a little bit. Since she did make a statement about this behavior I would extend her the respect of taking her statement at face value unless it is shown to be false.

On kernel mailing list behavior

Posted Jul 22, 2013 19:22 UTC (Mon) by Baylink (guest, #755) [Link] (2 responses)

Nope; extraordinary claims require extraordinary proof.

"Most of the people on LKML are abusive assholes", which is the substance of her complaint as I -- and a lot of other people, apparently -- are interpreting it, is an extraordinary claim.

In other news: Linus and the, oh, top 10 or 12 guys on that list *are*, plus or minus 10%, the close friends between whom you believe such a tone might be justified.

There is, as you suggest, room for difference of opinion on whether the behaviour perceived on LMKL a) is common, b) actually is abusive, and c) is thereby *inherently* inclined to scare off people with useful contributions to make at that level (which is *really* far up in the stratosphere; think about the literally billions of devices the Linux kernel presently runs).

The opinion seems to be (from this thread, anyway) close enough to 50/50 that I'd have to sit down and actually tally it. But it also seems to me that the people alleging it's not abuse are backing up their position with more data and less emotion.

I *am* pleased to hear that, with one or two small exceptions, no one seems to be reacting to Sarah a) on the grounds of her sex, or b) on the grounds of her technical competence -- only, like my response, on the grounds of her management competence, which is *clearly* (to me, anyway) in-scope for this discussion.

On kernel mailing list behavior

Posted Jul 22, 2013 20:27 UTC (Mon) by raven667 (subscriber, #5198) [Link] (1 responses)

> "Most of the people on LKML are abusive assholes", which is the substance of her complaint as I -- and a lot of other people, apparently -- are interpreting it, is an extraordinary claim.

That would be an extraordinary claim but I don't think that is not how I interpreted it and so I don't think that is the claim which was made. I understood the claim to be that Linus will sometimes, not often, go over a line with publicly dressing down contributions that he doesn't like. It is not clear to me what the relationship between Linus and the people being publicly berated is but it can set a tone on the list where that kind of behavior is considered acceptable. There are many maintainers on LKML and elsewhere in the professional sphere who are able to manage people, provide constructive feedback and maintain high quality standards without resorting to name calling and rude behavior.

I dunno, I think that the act of just talking about it probably will make the various actors think a little bit more about what they say and what the audience is when they say it which might move the conversation a few notches over in civility which would be mission accomplished from Sarah's perspective.

On kernel mailing list behavior

Posted Jul 22, 2013 20:32 UTC (Mon) by Baylink (guest, #755) [Link]

Definitely, the Hawthorne Effect will likely kick in here, as much as people will say that it won't.

> I understood the claim to be that Linus will sometimes, not often, go over a line with publicly dressing down contributions that he doesn't like.

You understood her complaint to be specific to Linus? Cause I didn't. It sounded to *me* like she was asserting that the LKML in general was a "hostile work environment".