MATHEMATICS

Sabtu, 15 April 2006

Software Needs Philosophers

Software needs philosophers.

This thought has been nagging at me for a year now, and recently it's been growing like a tumor. One that plenty of folks on the 'net would love to see kill me.

People don't put much stock in philosophers these days. The popular impression of philosophy is that it's just rhetoric, just frivolous debating about stuff that can never properly be answered. "Spare me the philosophy; let's stick to the facts!"

The funny thing is, it's philosophers who gave us the ability to think rationally, to stick to the facts. If it weren't for the work of countless philosophers, facts would still be getting people tortured and killed for discovering and sharing them.

Does it ever strike you as just a teeny bit odd that after a brief period where philosophy flourished, from maybe 400 B.C.E. to ~100 C.E., we went through a follow-on period of well over one thousand five hundred years during which the Roman Catholic Church enslaved everyone's minds and killed anyone who dared think differently?

What's weirder is that we tend to pretend it didn't really happen. We like to just skip right over the dominance of religion over our minds for a hundred generations, and think of religion today as a kindly old grandpa who's just looking out for us kids. No harm, no foul. Let bygones be bygones. Sure, there were massacres and crusades and genocides and torture chambers with teeth grinding and eyes bleeding and intestines torn out in the name of God. But we were all just kids then, right? Nobody does that kind of thing today, at least not in civilized countries.

We try not to think about the uncivilized ones.

It was philosophers that got us out of that Dark Ages mess, and no small number of them lost their lives in doing so. And today, the philosophy majors are the butts of the most jokes, because after the philosophers succeeded in opening our minds, we forgot why we needed them.

And if we stop to think about it at all, we think that it was other people, people who are very unlike us, who committed those atrocities in the name of Faith (regardless of whether it's faith in a god, or in a political party, or any other form of mind control carried out by force).

We like to think we live in an enlightened age, but we don't. Humans haven't changed significantly in 10,000 years. We're still killing and torturing each other. It's apparently incredibly easy to decide to kill someone and then do it. Happens every day, all around the world. Torture, too.

But those people are just people. If they had been born down the street from you, they'd have gone to school with you, been friends with you, learned to program with you, written blogs and comments, never tortured or killed anyone in the name of an idea. They'd have been you. Which means they are you; you just got lucky in where you were born.

One of the commenters on my last blog entry expressed the fervent wish that I drop dead. To be sure, they qualified it with "on the internet". But if they really feel that way, especially about something as hilariously and absurdly unimportant in the Grand Scheme as whether the Lisp programming language has any acceptable implementations, then what does it say about us?

Everyone who commented angrily on that blog entry was caught. I caught you, anonymous or not, being a religious fanatic. The only "negative" commenter who doesn't appear to be a religious zombie was Paul Costanza (ironic, since he claims to be the opinionated one), who relegated his comments to pedantic technical corrections. They're welcome, of course; I'm always looking to correct any technical misconceptions I harbor. But they're moot, since even if I was wrong about every single technical point I brought up in that entry, my overall point — Lisp is not an acceptable Lisp — remains largely uncontested by the commenters.

Some of them just don't get it, which is fine; no harm in that. If you've been using Lisp for years and years, and you've written books and articles and zillions of lines of Lisp code, then you're unlikely to remember anything about what it's like coming to Lisp for the first time. They're religious because they've forgotten what it's like to be a skeptic.

But make no mistake; a substantial percentage of people who take a side in any programming language discussion that devolves into a flamewar know exactly what the other side means, and they want to invoke the Ultimate Censorship: drop dead! Killing someone, after all, is one of the best ways to silence them. You also have to burn all their writings, which is getting harder these days; hence the increased vehemence on the 'net.

Those of you who've followed what I've written over the past year or so know where I'm going. I'm taking a stand, all right, and it's a very definite one. I'm finding myself drawn inexorably towards a single goal: stamping out technological religion, because I'm frigging tired of not being able to stick to the facts.

FACT: Java has no first-class functions and no macros. This results in warped code that hacks around the problem, and as the code base grows, it takes on a definite, ugly shape, one that's utterly unique to Java. Lisp people can see this clear as day. So can Python folks, so can Ruby folks. Java people flip out, and say "macros are too much power", or "what do u mean i dont understand u" or "fuck you, you jerk, Lisp will NEVER win".

You think I don't hear ALL that, and much more, in the hate mail I get every day?

I sure wouldn't want to be alone with a Java fanatic in a medieval torture chamber, because God only knows what they're capable of.

Turn the mirror towards Python, and what happens? Funny, but the Java folks will mail me saying: "yeah, I've always known I detested Python, and you really nailed exactly why. Thanks!" Meanwhile, Python folks are literally frothing at the mouth, looking for the "Kill That Bastard" key on their 101-key keyboards.

I turned the mirror towards Lisp yesterday. Had to go to the bathroom like nobody's business, and my wife was expecting me home any minute, so I rushed it out: just a few thoughts here and there. So the Gorgon only caught the tiniest glimpse of itself, but hell evidently hath no fury like that of a Lisper scorned, and all that.

It doesn't matter that I rushed it out. I'm glad I did; spending any more time on it, trying to get it "right" by looking up useless factoids like how you can override length's non-polymorphicness with some weird setting (when it plainly should just be the default), would have had the exact same net effect: Lisp zealots would have found some way to turn it into a flamewar. And I'd have been out 2 or 3 more hours.

Let's call it a troll, then, because it was poorly researched; it was just some months-old recollections of pain I'd gone through last year trying to commit to Common Lisp, after another year of trying the same with various flavors of Scheme and finding them all wanting. As far as I'm concerned, Lisp is unacceptable today; it's my opinion and just that, but I'll stick with it.

I still need Lisp; after you learn enough of it, it becomes part of your soul. I get my fix hacking elisp, and I do a lot of it. The commenters are quite right; I've never written anything substantial in Common Lisp, because in each of my serious attempts, there was too much friction. Risk/reward wasn't high enough, and believe me, I wanted it.

But after many attempts, I've given up on Common Lisp. They won't let me use it where I work, and there are probably more Lispers per capita where I work, including some famous ones, than at any other big company in the world. If we can't use it where I work, then it's frigging unacceptable; that's the shortest proof I can offer.

What I'm far more interested today is the situation that arises if you consider my post a troll. I'm far more interested in the social consequences of working in a world filled with religious fanatics of different religious persuasions. Especially given that it's a world in which "natural religion" has, by and large, been marginalized through the work of philosophers.

Let's look at this world in a little more detail, starting with Peter Siebel's comment, which I believe is the most interesting. Peter said:

I was trying to figure out why on earth you spent so much time writing about something that you apparently don't like. Then it hit me: HCGS. So thanks for your help.

His first sentence speaks volumes about the sociology. His viewpoint is exactly what they teach us all as kids: If you don't have anything nice to say, don't say anything at all. We like to think people have a right to believe whatever they want, and that it's not nice to say mean things about other people's beliefs, especially when their livelihoods are at stake.

That's where philosophers come in, folks. They pick your beliefs apart and show you in unforgettable ways the consequences of what you believe in. I'm no philosopher; I know basically nothing about it, but I can tell you I wish fervently that some great philosophers would come along and effect change in our technical society.

Because if nothing else, I can see the consequences of the way we're thinking about things. One of many such consequences is that languages aren't getting any better, and the worst offenders are Lisp and Scheme, which by rights should be racing along the innovation curve faster than their supposedly less capable peers. But they've stagnated worse than any other non-dead language I can think of.[1]

Programming languages are religions. For a long while now I've been mildly uncomfortable calling it "religion", but I don't feel bad about it anymore. They're similar enough. At the top of the language religion is the language itself; it serves as the deity and the object of worship.

Like any other organized religion, there's always a Pope (or a politburo chairman, in countries where the government has brutally set itself up as what is for all intents the religion of choice): a spiritual leader that gives the religion the human touch. This person is almost always the language designer, of course. In Lisp's case it's complicated, because McCarthy, Sussman and Steele aren't very active as spiritual leaders for their languages anymore.

Every major organized religion is a heirarchical government, and programming languages are no exception. You'll find equivalents of cardinals, bishops, priests and laity in programming language camps: the closer you are to the fire, to the spiritual center, the higher your rank. It's a great way to quantify your perceived self-importance: a high-score list, in effect. Great for the ego, but it makes you a piss-poor debater, because you're so emotionally invested in your status.

You'd think your rank would be accrued by virtue of your technical and/or documentation contributions, but in practice it's usually more of a function of how many converts you've gained, how many followers you have, how much you've been spreading the Word.

That's why Paul Graham isn't the Pope of Lisp. He's eminently qualified, but unfortunately he's a heretic. Notice that almost none of the commenters on my last blog mentioned the PG argument I made. The only one who did (as of this writing) tried to make it an argument for Common Lisp. Let's face it: you can't give those heretics too much press; people might start listening to them!

Peter, are you beginning to understand why I write so much about something I apparently don't like? It's because I wanted to like it but found it fatally flawed, technically and culturally. It's as if I were a would-be convert to Roman Catholicism, but I can't bring myself to commit because I've seen too much of their role in creating a history that ironically we all wish we could rewrite.

I was born and raised a Roman Catholic, and I renounced it when I was thirteen years old, after my Uncle Frank (a devout terrorist Catholic if there ever was one) told me to stop reading the Bible, that it would "really screw a person up" to do that, that you needed someone to interpret it for you. That wasn't the only reason I renounced it, but it'll suffice for our purposes.

Technologically I was born and raised an assembly-language programmer; at least that's what my first real job was, for 5 years after I got my CS degree. Assembly is just flagellation, though, and damned uncomfortable at that, so I joined the Church of Java for fully seven years. And practically at the very moment I'd finally tired of chafing at Java's limitations, Paul Graham came along and through his early essays, showed me Lisp. What a great new religion!

Problem is, each time you switch religions, the next one has less impact on you. Once a Catholic, always a Catholic, they say. I don't know what that means for me, since I was raised by the assembly-language wolf, but it appears to mean that I'm never going to be enthralled with another programming language. And now that I've swallowed the red pill, what choice do I have? I need to try to show people what's out there.

Interestingly, it was Peter Siebel's most excellent book, Practical Common Lisp, that played the role of Uncle Frank and killed my desired to continue with Common Lisp. Peter was the first person to show me beast's underbelly. Every other Lisp book had pretended it was pure and beautiful and uncorrupted, because they left all the nastiness out as "implementation-defined". Once I saw what you really need to do in order to build something resembling a portable Lisp code base, and then had a few runs at it myself, I threw in the towel.

I much prefer Lisp the idea to Lisp the implementation.[2]

I can tell you this: I've tried writing this essay for a year. I've tried fully a dozen times. I've tackled it from a dozen angles. I've wanted to say it — software needs philosophers! — so many times, in so many ways. We need great thinkers — the Fyodor Dostoyevskys and David Humes and Aristotles and Jean-Paul Sartres and Ben Franklins and Galileo Galileis and Bertrand Russells and Albert Einsteins to show us the way through the Software Dark Ages we're in today: a time that will doubtless be remembered as every bit as mired in darkness and ignorance as the Dark Ages themselves.

But I've failed. This isn't the essay I wanted to write, because I'm neither a great thinker nor a great writer. However, you might be: if not now, then perhaps someday. So I think it's better to get the idea out now than to hoard it in the hopes of someday writing a world-changing essay.

For those of you who were surprised at the suddenness and vehemence of the Lisp community's backlash to my little rant, I hope I've helped shed a little light, helped you see its inevitability. Basically they've had a lot of practice. Lisp is one of the oldest technology religions, and they've both experienced and doled out their share of religious persecution.

But that's not the lesson you should take away. The lesson is that they are you. Whenever you hear someone ranting about something you take for granted as wonderful and praiseworthy, and you're wondering why they don't leave well enough alone so we can all get back to our incestuous cheerleading, just remember: we went from the Dark Ages to our reeeeasonably enlightened society today by questioning our most cherished beliefs.

So keep questioning them.




[1] Yes, I've read all of R6RS. It's a lukewarm compromise that punts on most of the important issues. It's not going to make Scheme any more successful than it is today, which to me feels practically criminal; it was their one big chance to break out of the rut they're in. But it doesn't matter. Let's pretend this footnote is just a troll. If your hackles went up, then you're a techno-religious zombie, and I hope in my lifetime to find you a cure. Try your best to think about that long and hard before responding.

[2] For the record, the commenter I agree the most with is the one who said the problem basically boils down to an IDE issue. SLIME doesn't cut it, either, as beautiful as SLIME is. Can't use it on Windows to save your life, for instance. But that's one of a thousand problems with the Lisp IDE situation; it's pointless to try to discuss them all in blogger. It's probably pointless to discuss them at all, because it's just going to make me more miserable that no decent IDE exists for Lisp, except for Emacs-as-Elisp-IDE. Which is why I get my Lisp fix by hacking elisp these days.

Tidak ada komentar:

Posting Komentar