Internet old-timers are not necessarily more polite than newcomers. They do, however, tend to know the rules, sometimes as a result of having been slapped down more than once by even-older-timers. Beginners in online communication often don’t even realise there are any rules at all.
Hardcore open source geeks, for example, that is people who collaborate globally to write software, software you are using right now, but you don’t realise you are using because it just works, do have flame wars—indeed, they invented the term—but they tend to start them for good reasons. The sort of people who scrawl on Comment Is Free are abusive because they can be. Toddlers want to throw their tantrums in front of the largest audience available.
For those who remember when access cost dear (except perhaps Bill Gates), one of the gravest online crimes is wasting bandwidth. For those who climb corporate hierarchies (both public and private), one of the greatest virtues is being seen to be doing things. Inevitably, clueless white-collar managers love sending email messages burdened with flabby Microsoft attachments to hundreds of users. Geeks hate them. They hate them not just because they are an ugly way to use a beautiful thing (the Net), but because they do not work.
In the eyes of the true geek, doing things that you know are not going to work and from which you will learn nothing is an even greater crime than being generally wasteful or simply being a jerk. Many people behave badly when they are behind the wheel of a car; almost as many of us change for the worse when we discuss things via bulletin boards, or blogs, or email. Ironically, some of the best geeks are in fact jerks in real life, but they make useful stuff online. To succeed in open source, being conspicuously active is not enough, you must get things done. In the corporate world, by contrast, those who talk about “delivery” have rarely delivered anything of any value at all in their lives, those who call themselves “executives” rarely execute.
This is a prelude to my suggesting that you follow the links over at Never Trust a Hippy on the subject of clever collaborative technologies and the stupid things people do with them. There’s some sensible, concise advice at the foot of the main article he links to, but before that, as if by way of warning, there’s a link to an unpretty, counter-productive Microsoft Powerpoint slideshow.
“To succeed in open source, being conspicuously active is not enough, you must get things done.”
Eric S. Raymond? All he does is write wildly overgeneralized nonsense about how preternaturally virtuous hardcore open source geeks are.
Paul Graham’s another.
“Hardcore open source geeks … tend to start [flame wars] for good reasons.”
Richard Stallman? Randal Schwartz? Slashdot? Language and OS holy wars? I’d’ve had a longer list a few years ago, before I started ignoring hardcore open source geeks because of all the flame wars they start for no good reason.
Also, the last time I read a flame about wasting bandwidth, I read it on a VT100.
I’m a fan of yours, but I think you’ve gotten a lot wrong in this one.
There are plenty of examples of hardcore open source geeks starting flame wars. It’s a point I concede openly in this post—it begins with the sentence “Internet old-timers are not necessarily more polite than newcomers”—and it’s also a point made by the Wikipedia article I link to, but, on average, the sound-to-noise ratio on most developers’ mailing lists is orders of magnitude higher than on most general mailing lists. It’s the difference between a school chess club the youth club down the road.
This is partly because developers’ forums are there to co-ordinate a real project with real goals that ends up doing real work. I am typing this in a Firefox window onto a WordPress blog from a Linux machine running Xorg and you will be reading it broadcast from Apache. These aren’t just great pieces of software; they’ve changed the World; they’re free; and they’ve generated huge amounts of wealth. And at least three of them were products of schisms between rival groups of programmers. Can you imagine a bunch of blog trollers backing up their rants against the government by founding their own not-for-profit public service organisation? Apart from the appalled amusement of sane bystanders, what did any of the “arguments” on Comment Is Free ever contribute to the sum of human achievement?
OK, fair enough. A dozen or so BSD groupuscules are products of developer schisms, too (Netcraft confirms it!), and one of those ended up in Mac OS X.
But is it useful to compare blog comments to anything but other blog comments? CIF should be compared to Slashdot, not the Linux kernel mailing list.
Dunno about the uniqueness of developers on the S/N ratio issue; I was on a Thomas Pynchon mailing list for a while some years ago and the ratio was excellent. I mean, the traffic was almost all meaningful, at least within the definition of “meaning” common to people who talk about Thomas Pynchon a lot, which I’ll grant is a highly specialized sense of the term. But I mean they didn’t flame each other much. Certainly not as much as I thought they deserved.
I’m just not convinced that at tendency to communicate meaningfully on professional stuff is all that distinctively a developer thing. Managers tend to be full of crap, yeah, or at least people like us like to think so, but there’s more to the human race than managers and code monkeys.
I also think it’s unfair to make an issue of people wasting bandwidth when the people in question don’t even know what bandwidth is. It’s not strange that programmers are going to relate to communication via software (or anything else via software) differently than people who know nothing about software. Also, as bandwidth becomes cheaper, that’s less of an issue anyway: Sweating over a 500 kb attachment begins to look like premature optimization. Reminds me of an old-timer I ran into online some years ago, who observed that it’s perfectly possible to do OOP in assembly language (that’s what all the compilers are really doing anyway, right?!), and more efficient at runtime (though not as much more as he thought, I bet). Which kind of misses the point about what computing resources are for. Really, he was just an old-timer who didn’t want to learn anything new.
Incidentally, I think Firefox/Mozilla may a questionable example of open source; my understanding is that the vast majority of the work has been done by paid developers at Netscape (or whatever it is now). The most OSS thing about it was how many years it took to ship. *Cough*perl6*cough*. But Linux, yes, and also Perl 5, MySQL, gcc, GNU everything else, and a laundry list of others. No arguments there.
I don’t agree about flame wars.
They’re not helpful at all. They start from typically a defensive, negative mindset about things. Most of the people involved in them are basically idiots who don’t know a thing about the software, but instead feel threatened by an attack on their camp.
The people building the changes can’t be bothered with that. They’re too busy trying to produce software that does what they want.
M-F Y:
Fair question. I think it’s useful if we want to identify what makes for successful online debate and what doesn’t. By identifying the cultural characteristics of the kernel mailing list that result in greater light than heat then we might be able to find ways of cultivating them in forums where the temperature is presently higher.
It isn’t, but it happens to be something I am familiar with. In my previous job I used to contribute in a small way to a collaborative international open source software project—I did so almost against my will: my boss wanted to me to turn the research I was doing into production quality software, which I am sure you are aware is not the normal academic way. One thing that fascinated me was that a bunch of odd people of various nationalities and outlooks got on fine online most of the time, and only ever seemed to have personal spats when they were physically together in a real room.
This is a point I was trying to make by as part of my general admission that “[b]eginners in online communication often don’t even realise there are any rules at all.” So we could say that knowledge of historical constraints of the infrastructure is one of those factors that contributes to the more productive use of electronic resources for collaboration.
It is, but the problem today is that people actually send 500Mb-plus attachments to one another, as they ignorantly email each other the raw output of their digital cameras or scanners. I have seen such things bring down internal networks. Each time one of those is CC’d to everyone on site hundreds of copies of them are made on central servers, interfering with urgent traffic and stopping more sensible people from getting their job done.
It is, but the time-consuming process of the from-the-ground rewrite that caused the long delay in “shipping” that you refer to was a very “open source” decision to make. It was the opposite of what would have been expected of a corporate software project. The result was a far better product and, as you say, a far later one.
Tim:
Yes, I think the vast majority of flame wars aren’t helpful, but some of them are, as mine and Wikipedia’s and M-F Y’s examples show. I am in favour of:
trying to reduce the number of flame wars by identifying the behaviours that discourage them,
trying to obtain good from the general bad of flaming by identifying what it is about some flame wars that leads to constructive outcomes.
Sometimes a flame war is an manifestation of the underlying brokenness of a particular collaboration. If you are part of a couple and you bicker constantly then this might be a signal of a structural problem that you have not otherwise acknowledged. Similarly, two subsets of developers on one project might subconsciously have different aims and ongoing flame wars between them might make them conscious that those aims are incompatible and would be better achieved by running separate projects.
I have also noticed that, in the process of shouting one another down, supposed “idiots who don’t know a thing about the software” can help developers who do to identify which of various different approaches might be best for those who will “consume” the results of their deciding between them. One truth that developers often forget is that most of the software they write is going to be used by idiots who don’t know a thing about it.
The flame wars I’m thinking of are more along the lines of opposing factions – Apple vs Windows, PERL vs PHP. That sort of thing.
What happens in these battles is that people defend their position as though they have an emotional attachment. So, when I talk of “idiots who don’t know a thing about the software”, I’m not referring to users who turn up in a forum asking for some help. I’m thinking more of defenders of something who clearly aren’t using the software.
I’ve not come across flamewars amongst developers working on the same thing. Mostly, people suggesting changes/forking seem to deal with things quite amicably and positively. But I suppose some projects may be different.
It was the opposite of what would have been expected of a corporate software project. The result was a far better product and, as you say, a far later one.
This is very true. A lot of companies don’t properly assess refactoring. Over time, code ends up like spaghetti, and each subsequent change costs more. Sometimes it’s worth starting with a blank sheet on it.
The thing is; do those who race to be first to “scrawl on Comment Is Free” sit up until the early hours just to be the first to share their incisive ‘it’s all a Blair conspiracy, if I were PM this would never have happened’ type of view* against every article or are they perhaps in Another Time Zone and, if so, why do they bother posting on a British newspaper’s ‘nice try but it hasn’t really worked very well has it’ web site?
* ‘cept they very rarely use a subjunctive…
Beginners in online communication often don’t even realise there are any rules at all.
Very true. I came across a commentator on a favourite blog of mine who leaped in feet first with all manner or abuse masquerading as an argument, and the points he was making simply being a re-hash of the same arguments people have been making for years, i.e. Bush is stupid, etc. After a few of these he actually apologised and said he is new to blogging and is just finding his feet. I even exchanged a few emails with him, explaining how I am a blogger and forum member of some years experience, and I simply walk away from arguments which are pointless or abusive, hence I’d not replied to earlier posts of his.
Rather refreshingly, he seems to have taken all this on board and become a useful and interesting commentator, all in the space of a fortnight.
Anyway, regarding flame wars, few can be more entrenched and bitter than the Nikon-Canon battles which take place on the camera forums.
Nikon and Canon suck. Minolta rocks!!
yOR AL RONG!!!!LUNIX IS BETER THAN WINDWOS AND CANNON COS ALL TEH LEET HACKURS DO IT BUT I DONO BOUT MAC!!!IS KEWL I SUPOSE!!!!I HAYT FLAMURS COS TEHY ARE FICK AN DONO WOT THY TORK ABOT WEN I TEL THEM THEY AR RONG!!!!aN BIL GATE$$$$ IS HITLER!!!!!!!!!!!!!!!!1