LWN.net Logo

Digital Maoism: The Hazards of the New Online Collectivism (Edge.org)

Edge.org is running Jaron Lanier's criticism of the "hive mind,", which includes a couple of paragraphs on free software. "These movements are at their most efficient while building hidden information plumbing layers, such as Web servers. They are hopeless when it comes to producing fine user interfaces or user experiences. If the code that ran the Wikipedia user interface were as open as the contents of the entries, it would churn itself into impenetrable muck almost immediately. The collective is good at solving problems which demand results that can be evaluated by uncontroversial performance parameters, but it is bad when taste and judgment matter."
(Log in to post comments)

Digital Maoism: The Hazards of the New Online Collectivism (Edge.org)

Posted Jun 13, 2006 13:11 UTC (Tue) by malefic (subscriber, #37306) [Link]

Total BS. Standard FUD from people who don't understand the principles of FOSS. I've seen many proprietary applications with crappy UI (well, the majority). OTOH there's lots of nice and usable open-source applications. BTW, MediaWiki that runs Wikipedia is open-source.

Digital Maoism: The Hazards of the New Online Collectivism (Edge.org)

Posted Jun 13, 2006 15:51 UTC (Tue) by dmarti (subscriber, #11625) [Link]

Good software takes ten years. Get used to it. -- Joel Spolsky

GNOME project founded: August 1997.

Digital Maoism: The Hazards of the New Online Collectivism (Edge.org)

Posted Jun 13, 2006 19:33 UTC (Tue) by holstein (subscriber, #6122) [Link]

Only a few months to wait to get a good and nice KDE then!

KDE was founded in October 1996. Its announcement is dated 14th October 1996.

;)

Digital Maoism: The Hazards of the New Online Collectivism (Edge.org)

Posted Jun 14, 2006 4:45 UTC (Wed) by khim (subscriber, #9252) [Link]

Not yet: In fact when Notes 1.0 finally shipped it had been under development for five years. You should not start from KDE's initial annonce but from KDE 1.0 release or GNOME 1.0 release. Few more years, sorry...

Digital Maoism: The Hazards of the New Online Collectivism (Edge.org)

Posted Jun 14, 2006 8:33 UTC (Wed) by nix (subscriber, #2304) [Link]

But the version number is nearly arbitrary (at best it indicates library stability, and for non-libraries it indicates much less).

There's a lot of very stable software out there with version numbers less than 1 after years: clamav, GNU ed, gstreamer, PAM... hell, even etherwiresharkereal is still <1.0.

Sure, some of this is still evolving (gstreamer, say), but PAM, not released and still under development? Just because of its version number?

Digital Maoism: The Hazards of the New Online Collectivism (Edge.org)

Posted Jun 13, 2006 16:44 UTC (Tue) by NapalmLlama (guest, #26327) [Link]

Agreed. This man can never have used a modern Linux (or *BSD) distro if he think that FOSS is "hopeless when it comes to producing fine user interfaces or user experiences".

I just minimised Firefox (itself increasing in popularity by the day) and saw a Konqueror window, containg SVG icons and handy panes transparently listing FTP folders and storage volumes. Behind that is a desktop with yet more pretty icons, and embedded applets forecasting my local weather and detailing my current system performance.

Come back when you know what you're talking about, Mr. Lanier.

He didn't say WP is closed source

Posted Jun 13, 2006 17:53 UTC (Tue) by stevenj (guest, #421) [Link]

To be fair, he didn't write this. What he wrote is correct: unlike the text of Wikipedia, the code is not editable by any anonymous Internet user (at least, not the MediaWiki code WP actually runs). If it were, it would be a broken.

Of course, the same is true for every FLOSS project. (Lanier seems to understand this, as he writes, "Linux programmers are not anonymous and in fact personal glory is part of the motivational engine that keeps such enterprises in motion.")

Take a deep breath, people; try not to get too excited about one throwaway paragraph on free software in a long essay mainly on Wikipedia.

He didn't say WP is closed source

Posted Jun 14, 2006 4:55 UTC (Wed) by khim (subscriber, #9252) [Link]

What he wrote is correct: unlike the text of Wikipedia, the code is not editable by any anonymous Internet user (at least, not the MediaWiki code WP actually runs). If it were, it would be a broken.

And so would be Wikipedia. When have you where able to freely edit this page for example ? If you'll think about it Wikipedia and Linux distribution are written in exactly the same manner: some parts are written by "trusted highly visible writers" (Linux kernel and some important parts of Wikipedia) while some pages and just random throw-away code is written by god knows who (and of course the whole range in between the two opposites). And last category is often broken - both in Wikipedia and in free software. Differences are quantitative, not qualitative.

Digital Maoism: The Hazards of the New Online Collectivism (Edge.org)

Posted Jun 13, 2006 13:27 UTC (Tue) by cventers (subscriber, #31465) [Link]

What would make for better user experiences then? Layers and layers of
management on top of the programmers? Total development disasters that
leave your operating system _way_ behind schedule, even after you rip out
nearly every useful enhancement its new version might have?

Some people just don't get it.

Digital Maoism: The Hazards of the New Online Collectivism (Edge.org)

Posted Jun 13, 2006 13:43 UTC (Tue) by kirkengaard (subscriber, #15022) [Link]

Mmmm... kitten-herding... :)

Digital Maoism: The Hazards of the New Online Collectivism (Edge.org)

Posted Jun 13, 2006 14:08 UTC (Tue) by kirkengaard (subscriber, #15022) [Link]

" Democracy is based on the assumption that a million men are wiser than one man. How's that again? I missed something. Autocracy is based on the assumption that one man is wiser than a million men. Let's play that over again, too. Who decides?" -- Lazarus Long

What he calls collectivism, in several cases I see as relying on a larger pool of individual expertise. It is not that the group is smarter, it is that in the large sample, we may find individuals with the desired knowledge on a given subject area. This is the why of groupwork: collaboration to produce a greater knowledge base and a better total output, while reducing the weaknesses inherent in unquestioned single perspective.

The collective, as such, is never smarter. However, an aggregated knowledge base contains data beyond the possession of any individual member. We're not talking about intelligence; wiki is about skimming the good data as a group exercise. Yes, perhaps it works beter in a controlled group; OTOH, perhaps that merely pre-selects data limitations.

Digital Maoism: The Hazards of the New Online Collectivism (Edge.org)

Posted Jun 13, 2006 15:08 UTC (Tue) by smitty_one_each (subscriber, #28989) [Link]

I took the piece as a well-written caution against becoming too enthusiastic about a shiny new thing.
Wikipedia might be a great place to start a bit or research, but the old warning about nails and hammers applies.

Digital Maoism: The Hazards of the New Online Collectivism (Edge.org)

Posted Jun 13, 2006 16:07 UTC (Tue) by forthy (guest, #1525) [Link]

Actually, Democracy isn't based on the assumption that a million men are
wiser than one men. It's mainly about power and selfishness. If you allow
few people come to power and stay there, they'll just be selfish, and use
the power for their own good. That's where democracy came in: To kick out
the Tryrant that came to too much power. The basic test of "too much
power" is: Create too much enemies. That's how the Attic democracy
worked: You can't vote for someone, but you can certainly vote against
someone. So you can overthrow too selfish regimes.

Only direct democracy assumes that people can make real factual
decisions, and direct democracy is not very common. The people in power
fear that even more than the election ;-).

Digital Maoism: The Hazards of the New Online Collectivism (Edge.org)

Posted Jun 13, 2006 17:06 UTC (Tue) by kirkengaard (subscriber, #15022) [Link]

See, there again is the assumption that the group is intelligent. "Throw the bums out" only happens when enough small groups recognize the problem to affect the opinion of the whole. The aggregate population of the republic is generally satisfied as long as they get their bread and circuses, and don't have to look away from the tube. They will only overthrow regimes that are too stupid to keep them pacified.

Digital Maoism: The Hazards of the New Online Collectivism (Edge.org)

Posted Jun 13, 2006 17:15 UTC (Tue) by Ross (subscriber, #4065) [Link]

It's based on the same idea as a market economy: people may not be smart about everything, but they generally know what will benefit them.

Digital Maoism: The Hazards of the New Online Collectivism (Edge.org)

Posted Jun 15, 2006 7:26 UTC (Thu) by Wol (guest, #4433) [Link]

Actually, you don't need to throw out a Tyrant.

In the days when Rome was a republic, they instituted the role of Tyrant. An *elected* absolute dictator with a short term of office (six months, I believe).

A tyrant was elected at times of crisis, with a mandate to solve the problem by any means possible. It was actually quite a dangerous job - if you upset a fellow senator and he became the next tyrant after you ... well, one of the tyrant's powers was summary execution ...

Many Open Source projects are genuine tyrannies, in that sense. Linux is, for example.

Cheers,
Wol

Digital Maoism: The Hazards of the New Online Collectivism (Edge.org)

Posted Jun 15, 2006 11:47 UTC (Thu) by nix (subscriber, #2304) [Link]

You're thinking of 'dictator'. 'Tyrant' is of Greek origin. (Amusingly, Wikipedia has good articles on both subjects. :) )

Digital Maoism: The Hazards of the New Online Collectivism (Edge.org)

Posted Jun 13, 2006 15:29 UTC (Tue) by rknop (guest, #66) [Link]

Err... MediaWiki *is* free software (GPL and all). That's the code that runs the Wikipedia user interface.

Unless he's talking about the web browser, and Firefox too is free software.

In any event, sure, Wikipedia is homoegenized collective. The great thing about the web, though, is that it's not all that's there. You can also find blogs, which are usually flagrantly individualized.

Encyclopediae have always been bland homogenized base information. And, when I was a kid in school and tempted to use an encyclopedia for all of the research for my report, they told us that we had to have at least three (or five, or whatever) sources, and *only one* could be an encyclopedia. Encyclopediae are great because they summarize stuff all in one place, and make it easy to find. Their main use for research, though, isn't as a primary source, but as a way of getting you started, of giving you enough background in a subject so that you can *find* the primary sources and effectively place them in some kind of context. More than that, though, encyclopediae are great for people who are curious and want to know something about a topic, but aren't doing scholarship with it. Wikipedia works very well for this sort of thing.

If Wikipedia were the whole world, then, yeah, I'd be worried about what it is. But, that's true of *everything*. If all we had in the world were well-researched scholarly texts that are getting people tenure, then it would also be a very sad world, for such texts are usually inpenetrable to people who aren't PhD researchers in the discipline. If all we had were highly individualized blogs, it would be a sad world, for as great as blogs (like, er, say, mine) are, they're mostly for ranting. If all we had were newspapers, it would be a sad world, because there's the gatekeeper deciding what you get and what you don't, and there is limited space, and there is no coherent reference of topics over time. If all we had was FOX News it would be a sad world, but that goes without saying.

-Rob

Digital Maoism: The Hazards of the New Online Collectivism (Edge.org)

Posted Jun 13, 2006 16:03 UTC (Tue) by dps (subscriber, #5725) [Link]

While I do not buy the argument as stated, it is half right. If you leave it to developers stratching their own itches then many things will stop when the interface is "good enough", which may be less visually appealing than commercial standard.

Sometimes free software has higher standards of "good enough" than the commercial alternatices. MS word was writen to make a buck and poor quality typesetting was good enough. TeX's version of "good enough" was the best quality typesetting and the output is obviously better.

If taste and judgement is the issue then I personally reckons M$ windows is awful. No only is there no virtual desktop but it also unfailingly implements all of my pet hates. Need I say more?

Digital Maoism: The Hazards of the New Online Collectivism (Edge.org)

Posted Jun 13, 2006 19:07 UTC (Tue) by zgoda (guest, #36215) [Link]

User interface is for users, if you don't take the ergonomy into account, you will quick end up with a candidate for "UI Hall of Shame" (there was a webpage with this title at some time, I cann't find it now). So "user experience" is of really high value, unless you don't plan anyone to use your software. Many of us will happily accept usual "unix-style" way of doing things (by piping or redirecting outputs and chaining commands), but this is not an usual way of getting things done by people. Unfortunately, most of users of our programs are computer educated by using Windows and Windows-based programs, so if we neglect their UI expectations, we are on straight way to hell. First lesson of new-era linux programmer: get yourself used to Windows-based user interfaces. As a side note I can add, it's not that hard.

Digital Maoism: The Hazards of the New Online Collectivism (Edge.org)

Posted Jun 13, 2006 19:27 UTC (Tue) by bersl2 (subscriber, #34928) [Link]

Many of us will happily accept usual "unix-style" way of doing things (by piping or redirecting outputs and chaining commands), but this is not an usual way of getting things done by people. Unfortunately, most of users of our programs are computer educated by using Windows and Windows-based programs, so if we neglect their UI expectations, we are on straight way to hell.
By separating the Windows-style frontend from the Unix-style backend, we can preserve the Unix Way.

Digital Maoism: The Hazards of the New Online Collectivism (Edge.org)

Posted Jun 13, 2006 20:28 UTC (Tue) by arcticwolf (guest, #8341) [Link]

[...] you will quick end up with a candidate for "UI Hall of Shame" (there was a webpage with this title at some time, I cann't find it now).

Right here:

http://www.userinterfacehallofshame.com

First hit on Google for "UI Hall of Shame", too. ;)

Digital Maoism: The Hazards of the New Online Collectivism (Edge.org)

Posted Jun 14, 2006 9:56 UTC (Wed) by erwin (subscriber, #4707) [Link]

But TeX wasn't produced by a "collective", either -- it was produced by a single, dedicated individual
(Donald Knuth). And while it's very, very good for some publishing purposes -- typesetting plain
text and mathematics -- it's terrible for others, specifically for graphic- and design-rich things like
magazines and catalogs. Of course, the proper comparison isn't with MS Word; it's with programs
like Quark Xpress and Adobe InDesign.

Digital Maoism: The Hazards of the New Online Collectivism (Edge.org)

Posted Jun 13, 2006 19:55 UTC (Tue) by allesfresser (subscriber, #216) [Link]

My problem with this article's comments on free software development is that he doesn't realize that code doesn't magically appear out of nowhere, and no one expects it to. The point is not that we free the code and somehow little collectivist gremlins appear and make it shiny, happy software. The development community is comprised of *individuals*, who contribute their time and effort to making the software better. We know who they are--a few get to be modestly famous in their own way--and the community is indebted to them for their work. We (the community) shape the software to fit our needs--it's very deliberate and intentional, not some idealistic gas about how things improve themselves somehow.

But the big point is why we do it this way. We develop software this way because it preserves the ability of anyone to join and contribute, to become a member of the community as well as because none of us has all the answers and all of us will disappear from the community eventually, for one reason or another. In contrast to the closed model of development, software released freely will likely outlive its authors, and the code will hopefully provide inspiration and benefit to people not yet born when the software was written. (I am speaking here as a second-generation **ix user; I was alive when Unix was first written, but not yet sentient, so I consider myself to be the continuing and indebted beneficiary of that first generation's groundbreaking work.) The free and open release of methods, concepts and source code could perhaps be compared to plowing one's fields under to replenish the soil, instead of intensively planting until the soil is exhausted, then moving on to other land to do the same thing. Maybe we should call it 'sustainable development', since it's self-renewing in a way when new people come in and pick up the baton. :-)

p.s.: Of course, the FUDdish technique of using words that evoke the old Red Scare Bogeyman ("collectivist", etc.) gets a big roll of the eyes from me...one of these days hopefully trotting out the Communist scarecrow will fall out of fashion.

Digital Maoism: The Hazards of the New Online Collectivism (Edge.org)

Posted Jun 15, 2006 21:34 UTC (Thu) by kingdon (subscriber, #4526) [Link]

Well, I guess he is willing to tiptoe into Red Scare territory (most obviously in the title), but on the whole I thought it was a much better essay than you'd think from that. His examples don't come from Stalin or Mao or even primarily from left groups - he also cites the Liberatarian faith in the wisdom of the market or the faith that digg has in the wisdom of the links which people digg.

As for open source software being written by individuals, seems like he is mainly asking how those individual work products assemble themselves into the entire product. Is there some kind of averaging (well, lots of people are asking for or sending in this patch, so I guess I better merge it)? Is there a discussion where we try to come up with some kind of code which makes sense to everyone? Or is it something where that crazy Van Jacobson guy came up with a way to rewrite the network stack and others can say "wow, that's great" or "no, that stinks". The question is whether there is thought going into things. Or whether the evaluation of an idea is more of a popularity contest.

Ironically, when I think of software projects run according to the style which I think he is advocating - where strong personalities present ideas - I think of the Linux kernel. And when I think of the style which he seems to condemn - where there is a lot of discussion and attempts to define a group direction - I think of GNOME (and maybe KDE, although I don't know as much about that one). So I'm not sure whether he is right about user interface code requiring "authorship" (or whatever the right word is for the non-anonymous-collective way) and backend code not.

P.S. Did anyone make it to the end of the article? The bio they gave for him there is quite amusing if you read the whole article.

Copyright © 2006, Eklektix, Inc.
Comments and public postings are copyrighted by their creators.
Linux is a registered trademark of Linus Torvalds