Community faults

Posted Sep 26, 2007 14:30 UTC (Wed) by mjg59 (subscriber, #23239)
In reply to: Community faults by esr
Parent article: My Fabulous Geek Career (O'ReillyNet)

Honesty isn't a problem. Hostility is, and hostility is often waved away as "Just being brutally honest". I know many talented coders who are reluctant to submit anything to LKML because of the perception that it'll just result in people calling them incompetent on the basis of coding style, failing to understand a specific aspect of the layering models or trivial bugs.

This isn't about wrapping things in fluffy pink PC bunting. It's about not discouraging competent coders from producing useful code. Right now we're successfully alienating vast swathes of the population. You're right that a change in standards of acceptable behaviour would probably drive some long standing members of the community away, but I think that's an entirely acceptable cost for opening our community to a much wider range of people.

to post comments

Community faults

Posted Sep 26, 2007 15:13 UTC (Wed) by peace (guest, #10016) [Link] (11 responses)

If your submitting to a project and can't get something as trivial as coding style right than you probably deserve to be ridiculed. If your so fearful of coding style problems that you can not submit to a project you should probably see a therapist. How hard is it to investigate how a project expects code to be formatted and than just conform your code to that standard. It is really hard to organize a massive project when every one has their own vanity braces.

Kind Regards

Community faults

Posted Sep 26, 2007 15:23 UTC (Wed) by mjg59 (subscriber, #23239) [Link]

Which is more likely to result in a correct version of the patch being posted - ridicule or a polite response with a link to a description of the coding style and a request that the patch be tidied up to conform to that?

It's pretty straightforward. Supply constructive criticism, not plain criticism. Tell people what's wrong with their code and show them the resources they'll need to fix it. Act like you want their code, not like you'd be happier if they never submitted anything again. It's basic politeness, not political correctness gone mad.

Community faults

Posted Sep 26, 2007 15:38 UTC (Wed) by mjg59 (subscriber, #23239) [Link] (8 responses)

To clarify that a bit further - if my response to your comment had been more like "If you're contributing to a disucssion and can't spell something as trivial as 'you're' right then you probably deserve to be ridiculed", then you'd probably think I was something of an asshole. If you got a complaint about every single misspelled word you posted here, you might end up thinking that LWN was sufficiently full of assholes that posting was more effort than it was worth. That wouldn't really benefit anybody.

Community faults

Posted Sep 26, 2007 17:37 UTC (Wed) by peace (guest, #10016) [Link] (7 responses)

If LWN was a site about spelling and grammar I would agree. As it is, I am dyslexic you insensitive clod! No, really, I am. My fragile ego has been shattered into many small bits which I will now pathetically pick up shard by shard as I retreat to a more civil oasis somewhere on the Internet. Be happy firefox has a spell checker now or I doubt you'd be able to endure my posts at all.

"Which is more likely to result in a correct version of the patch being posted - ridicule or a polite response with a link to a description of the coding style and a request that the patch be tidied up to conform to that?"

As long as the devs themselves can point to a pedant or a clear coding style in their code base you are deserving of whatever you get. This kind of ridicule would happen quickly and early in order to align contributions. Who has the time to school everyone who needs it. There is a reason for the "F" in RTFM. Use your brain and figure out whats going on and how best you can contribute.

Trying to control how other people should behave to you is self centered.

Kind Regards

Community faults

Posted Sep 26, 2007 21:25 UTC (Wed) by man_ls (guest, #15091) [Link] (6 responses)

LWN is a site about people writing things that others read. There are published conventions for this activity, just as there are published code conventions for the thousands of software projects out there; but these ones have been available for centuries. I don't think it is enough to say "I am dyslexic" or "I am a foreigner", since you are obviously not trying hard enough and just hiding behind the "dyslexic" label.

And yet most people here are polite enough to ignore such matters, and to keep a civil tone; this makes me quite proud to support it, actually. These people (who have taken the work to learn how to spell) might say "use your brain and RTFLexicon", but they choose to be constructive and listen to you anyway. This attitude would be very useful on some developer lists I have seen.

Community faults

Posted Sep 26, 2007 23:52 UTC (Wed) by peace (guest, #10016) [Link] (5 responses)

Yet, despite your comments I am not going to flee the site or the community nor am I going to demand you behave any differently. Besides, I have seen the conventions here and I think I fit right in.

I am not arguing *against* civility. But I don't demand it. As long as someone has a point, in the end, thats what matters. Not everyone is going to fit the mold you prefer.

-peace

Community faults

Posted Sep 27, 2007 4:01 UTC (Thu) by mjg59 (subscriber, #23239) [Link] (2 responses)

You may not be put off by this sort of thing, but to suggest that others shouldn't be is suggesting that they fit a mold if they want to participate. I see it in a more straightforward manner. By and large, people have a choice as to whether they want to be offensive. People have much less choice over whether something is going to upset or offend them. With no terribly good argument for why the offensive guys are going to be better programmers, I'd prefer that the people with the choice modify their behaviour so the people with less choice get to participate as well.

Of course, if you want to behave in a way that reduces the number of potential developers, feel free. You'll have to convince people as to why the community as a whole will benefit from that, of course.

Community faults

Posted Sep 27, 2007 17:59 UTC (Thu) by pizza (subscriber, #46) [Link] (1 responses)

> By and large, people have a choice as to whether they want to be offensive. People have much less choice over whether something is going to upset or offend them.

You left out the third part:

"People also have little control whether something is going to upset or offend someone else"

Some people are very easily offended.

Community faults

Posted Sep 27, 2007 18:21 UTC (Thu) by mjg59 (subscriber, #23239) [Link]

Indeed. Some people are excessively easily offended. That doesn't imply that everyone who is offended by something is over-sensitive.

In any case, the same basic argument still stands. Your first interaction with someone may offend them. That gives you a better understanding of where their thresholds are, and you can choose to modify your behaviour to reduce the probability of offending them in future. It may be that the compromises you'd have to make are excessive (if "You're failing to acquire this lock before modifying this data structure, which could lead to unexpected behaviour" ends up offending them, for instance), in which case it's probably better to just tell them that and avoid interacting with them in future.

Not everyone is able to accept constructive criticism, and those who aren't are unlikely to make especially useful contributions to the free software world. But that's not the same set of people as the ones who can accept constructive criticism but are put off by hostility. Losing the latter because you don't think we should deal with the former isn't a sensible tradeoff.

Community faults

Posted Sep 27, 2007 8:53 UTC (Thu) by man_ls (guest, #15091) [Link] (1 responses)

Besides, I have seen the conventions here and I think I fit right in.

The conventions I referred to, in case it is not obvious, are called "grammar" and "orthography" (or "spelling"). Some people stick to them more strictly than you do (not me, though).

As long as someone has a point, in the end, thats what matters.

Well said. Contrast this with what you said before about code contributions:

"If your submitting to a project and can't get something as trivial as coding style right than you probably deserve to be ridiculed".

People might paraphrase your latter contribution: "As long as someone has good ideas, in the end, that is what matters. Not everyone is going to fit the code conventions you prefer", and it would still be reasonable. Code is the expression of an idea, just as common language.

What we are requesting is the same level of civility in development mailing lists as on LWN. There is no reason to behave differently.

Community faults

Posted Sep 27, 2007 14:49 UTC (Thu) by peace (guest, #10016) [Link]

What I'm getting at is this: is Free software going to be able to include the grungy dirty smelly realities of humanity or is it going to require a certain esoteric social etiquette defined largely by the dominant main stream culture, likely all white and docile. Technology was largely driven by social out casts of all flavors. Now that it's main stream are we only going to accept bathed vanilla?

man_ls, I've been responding to several of your posts here, please don't think I'm picking on your handle :). You may be referring to very specific instances of abusive behavior or even a certain class of abuse and I would likely agree that those cases were unfortunate.

I *am* responding more generally to a certain trend of gentrification of the tech ghetto you might say. I don't really have it fully worked out but I do know that geek, hacker and nerd do not mean what they used to.

If you have any foll up thoughts I'd appreciate reading them. I'm going to give this thread a rest, though... I'm sure we'll meet again, muhaha!

Kind Regards

Community faults

Posted Sep 28, 2007 0:57 UTC (Fri) by chromatic (guest, #26207) [Link]

If your submitting to a project and can't get something as trivial as coding style right than you probably deserve to be ridiculed.

No, you don't. Why do you suggest that ridiculing someone trying to help you in good faith is ever a good idea?

I'm one of the first people, when someone starts trying to heap abuse on contributors, to stand up and say "Go away!" but abusing potential contributors who've already made the effort to contribute is unacceptable, even mean.

Community faults

Posted Sep 26, 2007 16:46 UTC (Wed) by gdt (subscriber, #6284) [Link]

...reluctant to submit anything to LKML because of the perception that it'll just result in people calling them incompetent ...

Been there, done that. I updated the Remote-Serial-Console-HOWTO and collected bug reports from its readers across two years. I submitted a set of patches to LKML to fix those bugs. I'll never do that again :-)

[And no, my code is not that bad.]

Interestingly, my code to set the TCP congestion control algorithm from Netfilter was also rejected. But this was done after a pleasant discussion and for a good technical reason. I attribute the pleasantness of that encounter to not CC-ing the code to LKML.

I've written a NAT module for Cisco's Skinny protocol for their IP phones. Not sure if I can stand the hassle of submitting it.

Community faults

Posted Sep 27, 2007 10:52 UTC (Thu) by alankila (guest, #47141) [Link]

The old "tact filter theory" might be worth a rehash in this exponentially ballooning thread.

http://www.mit.edu/~jcb/tact.html

You know, I now expect to see some theorizing about which side males' and females' tact filters are on in further posts. :-p

Seriously though, the impression that people are _rude_ in the free software word might be nothing more than simply result of how geeks grew up to communicate. You say it's "waved away as being honest" but what if it really _is_ honest?

On to your point. I remain unconvinced that getting people who can't take flaming to contribute is a good thing. Why? Because I'm not convinced that their contributions are necessarily better than those people's who you drive off. (Some guys like Joel Spolsky are fond of stating things like "good coder is worth 10 bad ones".)

The reason why I think some flame resistance should come for granted is because lots of code is not really an expression of your personality or artistic talents, but is rather grunt work that should flow from understanding the relevant theories and using appropriate tools to work towards your goal.

If you make bad choices and get criticized for them, and your evaluation of the criticism shows that it's relevant, then someone might actually know better than you. (A golden opportunity to learn from a master!) If you find the criticism highly juvenile or irrelevant, then you are in the position to say so, too. Discussions being a public archive means that meritocracy of its sort should run its course and decide who is right.

Of course, valid criticism (maybe "failing to understand a specific aspect of the layering models") also means that you really might not know enough about the system to properly contribute to it. (In other words, accepting the patch to kernel might introduce bugs.) I find this concern strangely valid. I mean, even a trivial bug can serve as a signal that this guy is not careful enough to get even simple stuff right: how could he possibly contribute something of real value?

Not to mention that code isn't everything. The person behind the code matters because future maintenance is likely to fall to him. And how well he interfaces with others matters because others may want to add stuff to his code. These matters are decided by the existing community: you have to integrate to it. If it seems hostile, then remember the tact filter stuff, give others the benefit the doubt and don the flamesuit.