If you look at the software that people actually use to do things -- compose music, design an airplane, edit a video, design a logo, lay out a brochure, etc., virtually anything besides computer programming -- the best-of-breed tools are proprietary and exist only on proprietary platforms (typically Windows or Mac).
This is because to just about everybody except computer programmers, the interface is the application, and considerable time and money have been invested in things which are more important than beautiful code -- things like usability, "looking pretty", and functioning in a way that mirrors how the intended audience things. This is something that programmers are notoriously poor at; for example programmers-who-are-also-musicians are likely to come up with music tools that make sense to them but utterly baffle nonprogrammer musicians. (Think Csound, Rosegarden, etc.)
Things like AutoCAD are innovative not because they contain fancy new algorithms but because they enable large chunks of the population to do things that they couldn't do before. And AutoCAD is proprietary, expensive, and runs only on Microsoft Windows.
In short, it's looking more and more like proprietary software is what has brought us real freedom.
Is this a joke? PageRank (academic paper, right?)... Flash (what kind of ace in the hole for closed source is that??)... And closed development is not necessarily closed source! And the iPhone running Linux would not make it a worse (or necessarily better) product. They would have to release their modified kernel code but that is it. How would this effect the final product? This has to be a joke article, or this guy is truly ignorant. Yet, he claims to support open source at times...
An honest empiricist must conclude that while the open approach has been able to create lovely, polished copies, it hasn't been so good at creating notable originals.
Hmm. http://www.well.com/~jaron seems to be running Apache, which I think has always been notable open source software.
As for being "original", its only predecessors are
NCSA's httpd and Tim Berners-Lee's original httpd.
Apache is much more "original" than the iPhone, which seems to me a "polished copy".
From what I can tell, this guy is a stoner who never actually does anything.
His webpage is ugly as shit (he has no taste, which you have to have if you're a do-nothing spectating whiner), too.
His arguments are bullshit. Nearly all implementations of advanced programming languages are open source and Google's core Internet-scale infrastructure is built on customized open source software.
"Google's core Internet-scale infrastructure is built on customized open source software."
Yes, but Google infrastructure isn't considered as innovative as Pagerank, which is proprietary.
I personally disagree that ranking pages by the amount of sites that link to them is properly patentable in terms of being non-obvious to an engineer skilled in the area, but that's off topic.
PS. Your whole post should be:
"Nearly all implementations of advanced programming languages are open source and Google's core Internet-scale infrastructure is built on customized open source software."
Everything else is playing the man.
PageRank is indeed proprietary, but it's proprietary more like a trade secret rather than proprietary like proprietary software that gets put on your computer and you have no idea whether, e.g., it's calling home and violating your privacy.
The guy is pretty annoying and I can't resist taking cheap shots at him, sorry.
"The guy is pretty annoying and I can't resist taking cheap shots at him, sorry."
Especially when he introduces terms like "encapsulation" and "testing" that don't seem to be connected to anything other than his bad analogy. I can't figure out what he's on about and I'm not especially motivated to, since I'm pretty sure he's looking for a pattern where none exists.
It looks to me as if inspiration and invention have happened throughout history independently of the circumstances of being employed by a company seeking to make patented and copyrighted products out of them, or being a hobbyist committed to sharing software.
PageRank is a ridiculous example. Isn't it proprietary specifically to try to gain an advantage over Google spammers? And why would a hobbyist start a PageRank-like project in the first place? I'm not convinced Google's decision to keep it secret is optimal, given the success of free and open email spam filters, but they have enough smart people working there to come up with a reasonable solution to the problem anyway.
Speaking of spam filters, I'd say those represent a far better example of innovation than anything Lanier could think of, and virtually all of that innovation has happened in free software.
A running joke between me and a friend of mine is "signals on wires!" referring to Jaron Lanier's criticism that all software is brittle because it operates on the "metaphor" of a signal traveling down a wire, instead of his preferred metaphor of interacting surfaces or something.
Like a nerd-issue Naomi Klein, he's acting as if metaphors are valid criteria on which to assess the technical merits of software. It's really quite ridiculous and indicative of being a bit too fond of the wacky tobacky.
Many have assumed that he is a drug user probably because of his dreadlocks, but Jaron Lanier has declared that he has never used controlled substances.
The author doesn't really seem to get what Linux does:
My guess is that a poorly encapsulated, communal gloop of organisms lost out to closely guarded species for the same reason that the Linux community didn't come up with the iPhone: Encapsulation serves a purpose.
Since when did Linus and friends come up with hardware? Does he mean the "community of people who rely on Linux as the core of their product offering?" -- would he then include Apple in the BSD community?
By his own criterion, Apple would be the perfect shop for OS innovation -- but I guess they just weren't feeling innovative that day and switched to FreeBSD instead. By doing so, they got the benefit of evolution -- not as snazzy as innovation, but something open source has excelled at for its long life.
Wasn't Apple the company that had nearly a decade of abandoned "next generation" operating systems (Pink, StarTrek, Raptor, NuKernel, TalOS, Copland, Gershwin) before finally buying NeXT (based on open-source Mach from CMU) and later establishing a cooperative coexistence with FreeBSD.
Innovation is often understood to mean one of two different things -- pushing boundaries and changing the world, and filling a niche that no one else is filling. They are different things. I'm sad to say it, but I think the first kind is often rooted in dreamy hubris, and thus fails; while the second one -- which is all about a willingness to serve -- does not appeal to the pride of entrepreneurs and will always be in demand in consequence.
The pace of innovation in computing will never again match that of the sixties or seventies -- lots of stuff has been built, has proven its value and been recorded in the Book of Wisdom. Whenever you think, "I know, let's build a such-and-such...", nowadays, you can be pretty sure it's been done at least half a dozen times. We really should emphasize looking back, to learn from history -- now that there is some.
As far as I can tell, the only computing field that does a good job of that is system administration -- experience, mentors and an appreciation for the "old ways" make up an integral part of that discipline.
Yeah, I've been wondering about Richard Stallman lately too. His searing experience was seeing the LISP world going commercial and closed... why did he choose to copy Unix instead? Was it just to piggyback on Unix's incumbency? That wasn't so strong at the time anyway. I should re-read Steven Levy's book "Hackers", the answer is probably in there.
I thought the thing that pushed Stallman into open source nirvana was that he couldn't get code to printer drivers to make the printers work in his lab.
He used to be a big deal when Virtual Reality was a big deal. Nowadays, I guess he writes articles where he name-drops. Having said that, I enjoyed reading his article.
Another way to think about whether closed-source sometimes has a place:
What if, by law, all software had to be open-source? Do you expect there would be more, or less, innovation? More, or less, diversity of implementation?
(I love open source but think the prospect of outsized returns from secret/proprietary codebases attracts some innovation and diversity we wouldn't otherwise have, especially in places where software creates a core competitive advantage -- investing, Google-scale cluster operations, domain-specific expert systems, etc.)
Some software is very generic and has a 1:100,000,000 ratio between itself and potential deployments. The same linux distro can work for billions of people. This kind of wild efficiency allows free software to be viable in a bottom-line sort of way--selling books, consultations, support, etc.
Most software has a ratio like 1:1 (custom software) or 1:10000 (industry-specific stuff). Open source is just completely impossible here. People need to make money, like it or not.
Most software is not innovative, though -- and innovative software has rarely been incentivized correctly. (How much is Xerox making for inventing the Macintosh?)
The author is one of many criticizing Open Source for failing to kick UNIX out of the white room, the sort of thing that, in its scope, would definitely be one to millions.
This is because to just about everybody except computer programmers, the interface is the application, and considerable time and money have been invested in things which are more important than beautiful code -- things like usability, "looking pretty", and functioning in a way that mirrors how the intended audience things. This is something that programmers are notoriously poor at; for example programmers-who-are-also-musicians are likely to come up with music tools that make sense to them but utterly baffle nonprogrammer musicians. (Think Csound, Rosegarden, etc.)
Things like AutoCAD are innovative not because they contain fancy new algorithms but because they enable large chunks of the population to do things that they couldn't do before. And AutoCAD is proprietary, expensive, and runs only on Microsoft Windows.
In short, it's looking more and more like proprietary software is what has brought us real freedom.