It's because we [developers] are highly opinionated people and [most of us] love to be "right". Some people have a hard time understanding other people might not think the same - and in cases when they don't know how to express their feelings - you get name calling.
Maybe there is a correlation between developers and math? Since I'm guessing many developers might have a CS background that involves a lot of Math, Statistics, Physics, and of course Programming. These subjects require people to answer things correctly hence subconsciously create a habit of being "right".
You are probably right, and it's odd because if I've learnt anything from my time in science research, it's that science isn't like your school homework or undergraduate exams where there is a right answer. I find the notion that there is some absolute Truth which science expounds not to be useful. A useful thing is to be less wrong and to have a good handle for all the ways in which one might be through assumptions, approximations, systematic uncertainties and so on. The wider relevance of all this is the recognition that technical decisions involve some sort of trade-off and are not necessarily the end of the story. In science it's a healthy thing to have different approaches or formalisms because it gives some confidence in the results if they agree. I don't know if that has any wider relevance.
The need to be "right" applies to most people. It's why we have fashion, MTV (or whatever the kids are watching these days!) and so on. It's also why it's important to reach that critical mass where you no longer need to convince people to use your stuff: once you're there the rest will convince themselves.
Instead of talking about why you are better than the other guy, let’s focus purely on why your system of choice rocks. That’s it.
Nice diagnosis; useless solution. Tool quality is relative. Fast, easy, safe, these are relative terms. Take me back to 1995, and CVS would be my best friend. Repositories rarely get corrupted, and when they do, they're easy to fix! For 1995 values of "rarely" and "easy", at least. Standards are different now because better competitors have emerged.
I like to think that promoting what you currently use gains more traction for what you use in terms of; community, documentation, features, demand, etc.
Applies to all languages and software.
But yeah, the immaturity with using fallacies (especially ad hominem) to argue puts me off and it doesn't help with the general community votes for it.
Some links for those who wish to know how to debate like an intellectual:
"""
After he has said, "I understand but I disagree," he can make the following remarks to the author: (1) "You are uninformed"; (2) "You Are misinformed"; (3) "You are illogical - your reasoning is not cogent"; (4) "Your analysis is incomplete."
"""
With this in mind, I think saying "I love mercurial" without at least referencing git, the more popular DCVS, will be less convincing since anyone can say, "Your analysis is incomplete - haven't you tried the most popular DCVS?"
I agree: the issue with "I love X" as a pure statement is that it has nothing to do with why we argue on the Internet.
Debating a topic is supposed to enrich us on that subject. I have changed my opinion on things before, based on Slashdot comments (this was way back), and now based on HN comments.
Of course, there is also the need to portray ourselves as clever. That's the real driver for flamewars: sophistry.
But saying "I love X" does nothing for either cause. People interested in rational decision-making are not engaged; trolls are not discouraged; and newbies are not educated on how to argue better.
I am utterly offended, DOS is awesome. Esp for embedded devices where the OS and the application are the same thing. Less complexity leads to more stability.
Flame wars are the most virulent when the enemy is extremely close to your position but doesn't take it. Emacs and VIM, essentially the same. Mercurial and GIT, again, the same. If the differences are too far apart the whole argument becomes, meh.
Fair enough, so let me remind anyone who wants a very thin OS for embedded x86 applications that FreeDOS (http://www.freedos.org/) is a great DOS for the purpose.
Relevant, focus on the positive outcome instead of a stupid futile private little war that only distracts people from doing great work with whichever freaking tool they needed to use in order to achieve it.
Apparently there are many good arguments for Mercurial, and many bad arguments for mercurial. The author says:
"These questions matter. And they’ve been answered, very eloquently, many times."
I've used CVS, svn, bzr, and git - I'm still a newbie on those last two and therefore persuadable - so where do I read these Hg answers? Seriously, I'm not doubting that they exist, but I'm getting the impression that google results will be polluted, and I'm not sure even the most beatific tweet stream is going to pull me in; but I will peruse the tweet recommended http://hginit.com/
Seriously, where are the issues carefully explained and addressed?
Well, saying "I love mercurial" without some mention of why you aren't using git, which is more popular, will seem like an incomplete argument. Not using what most others are using has drawbacks, so some justification is needed. But it is true that spending too much time promoting something in terms of its competition can weaken the argument. A tough road for the underdog; try explaining why you are excited about android tablets without mentioning the ipad 2 :)
saying "I love mercurial" without some mention of why you aren't using git, which is more popular, will seem like an incomplete argument.
That's the point. When you say why you love Hg without saying why you aren't using git, you aren't arguing. The author isn't explaining how to argue more civilly, he's explaining how to make a perfectly valid statement without getting into an argument.
Example: I love riding a hard tail MTB. Most riders will ask why I don't want full suspension. But I'm not arguing about why a hard tail is better. I'm explaining that I love riding a hard tail.
Not using what most others are using has drawbacks, so some justification is needed.
This is only true in the open source world and then only in one corner of it. Subversion is still, by far, the most used version control system in the enterprise.
And why should someone have to argue why they aren't using git? If he loves Mercurial, great! He doesn't have to explain why he doesn't use every other version control system in existence.
Good question. I think git is so relevant to the mercurial argument since it is distributed, so if you are saying why you love mercurial, at least acknowledging git will strengthen your argument, as it is the most popular DCVS. Without that mention, readers might falsely think you are uninformed and haven't heard of or tried git.
That won't make a lick of difference. Any statement of the form 'X is great, because ...' or 'I like X, because ...' will be interpreted by a bunch of people as 'He means Y, which I love, is not great and is lesser than X. Otherwise, he would be using Y. He must obviously think it sucks, which shows what an idiot he is'. And the stage is set.
People can't help themselves. They want to defend their choices, even when their choices aren't under scrutiny. They will follow up with a fallacy or two and they start arguing from positions that are so ludicrous that even with the best initial intentions you forget to point out the fallacies they must obviously have committed to arrive at their starting point and just start refuting them. That way, they suck you in.
Only if there are sufficiently many people in a community resistant to this kind of trolling, pointing it out to each other and refraining from feeding the trolls, only then can sensible discussion come about. But it's hard not to get sucked in. It reminds me of David Foster Wallace's speech about being aware of the water that you are in: it's hard work and requires an amount of self-awareness that can be exhausting.
That won't make a lick of difference. Any statement of the form 'X is great, because ...' or 'I like X, because ...' will be interpreted by a bunch of people as 'He means Y, which I love, is not great and is lesser than X. Otherwise, he would be using Y. He must obviously think it sucks, which shows what an idiot he is'. And the stage is set.
Nobody forces you to pander to these people's neuroses. You are not the troll whisperer. This over here is my favourite interview question. That over there is my favourite line of Ruby code. This is my favourite book. I wrote blog posts about all of these subjects and attracted exactly the responses you describe. But I was happiest when I ignored the people who had an agenda of arguing about their favourite interview question or their favoruite line of code or their favourite book.
I find it difficult not to respond for at least two reasons:
- The information is often factually wrong. Letting wrong information go uncorrected will lead to that information spreading and that doesn't seem like a good thing to me.
- It often doesn't seem to be the intention of people to troll and it is hard to suppress the urge to teach. It would be awesome if you could get more people to intentionally not-troll.
The first may be an instance of http://xkcd.com/386/ and the second may be a case of vanity (thinking you can change people), but just ignoring people doesn't improve things either, so shouldn't you at least try?
Unfortunately, attempting to address both these points simultaneously seems a recipe for disaster and just results in 'feeding the troll'.
By choosing a favorite, you cannot help but choose an infinitude of not-favorites. With a high-traffic blog such as yours, it's near certain that some people will believe you've shown yourself to be an idiot by missing a superior choice. Neuroses may be at work in some cases, but it's arrogant to assume that all who disagree with you are neurotic.
Choosing to believe something about what I believe based on my expressing my like for something is not neurotic. ASking me to explain myself or seeking further information about my choices is certainly not neurotic. Choosing to believe I'm an idiot because you like something else and because I didn't attempt to intellectually bludgeon you into agreeing with me is _____, but it's not neurotic.
But someone getting so invested in what they think that a stranger believes, such that they have to start a flame war? And expecting that I have some sort of obligation to their egocentricity such that I must respond to their questions?
Well, I'm not a psychologist, so the odds are strongly against this being clinically neurotic. But it isn't for me, and I don't feel a compulsion to play along.
p.s. Again, this isn't about people's right to believe what they want to believe. I'm ok with people believing I'm an idiot. They may be right! But what I'm talking about is whether I have to respond. I don't. I also don't have to worry about it when I write such that I try to deflect criticism or protect my reputation.
I find it depressing that your comment was down-voted by at least two people — two people who, by virtue of their karma, are supposed to somehow have an understanding of what a productive comment is and what it isn't. The response your comment has received from these Gaurdian[1]-like members of our supposed nerd virtual polis[1] is evidence that you have a point.
Several weeks ago, Daring Fireball featured a link to an exchange between Ian Fleming and a man who wrote him offering suggestions (and a critique of James Bond's current) pistol selection. In a comment here at HN, I believe, someone noted the civility, the maturity of the exchange.
I thought about that comment, and I decided that I was going to try to restrain myself and make comments that a hypothetical adult might make.
HN is part of a problem, and that problem is what I'll call the cable news-ification of everything. No, not the Fox News-ification, but the cable news-ification, because the premise of all these channels is that when something happens, the most important thing is to have an opinion about what happened, and to broadcast and debate that opinion as vigorously as possible.
Wasn't me! I have a strict policy of either down voting or responding but never both. If a comment doesn't contribute to the discussion, I down vote it and move along. If it merits a response, it clearly is doing something of value to engage me, and the back-and-forth may uncover something else interesting.
"I find it depressing that your comment was down-voted by at least two people — two people who, by virtue of their karma, are supposed to somehow have an understanding of what a productive comment is and what it isn't."
Am I missing something, or how can one possibly see a list of downvoters for a given comment?
I don't, unfortunately, have any special insight into the voting system. I knew that it was downvoted more than once because my re-upvote didn't turn the comment black, meaning that its score was less than zero.
I've (mistakenly) assumed that you need to see the list of downvoters to make a statement such as edw's, whereas I now see how, having information you and edw provided, you can just infer this statement.
When I read "X is great" while using Y, I reflexively put my choice under scrutiny. I see that as the main reason to talk about technologies X or Y at all. And if X has some terrible drawback, network effects can make it worthwhile for me to dissuade other people. Hg vs. git isn't a compelling example, but if I want to never face another svn-induced merge conflict again, I myself am actually better off if everyone I may collaborate with already knows how troublesome that tool is (I would have said "nightmarish" in the svnmerge.py days).
I think it does.
I for example are somewhat attached to my decision of using git and would probably dismiss someone telling me how much it sucks. On the other hand I am not a power user, it intimidates me at times.
If I read a post explaining a mercurial workflow that looked easier then what I do now I might check it out. Yeah you read correctly, I didn't even install mercurial.
I just moved from svn to git once rails and most gems did. I guess I am not the only one so there are people youcould convince.
