Welcome, Guest. Please login or register.
Did you miss your activation email?

Login with username, password and session length

When bottom-up systems are superior to top-down ones

Quote
I had been preaching the Unix gospel of small tools, rapid prototyping and evolutionary programming for years. But I also believed there was a certain critical complexity above which a more centralized, a priori approach was required. I believed that the most important software (operating systems and really large tools like the Emacs programming editor) needed to be built like cathedrals, carefully crafted by individual wizards or small bands of mages working in splendid isolation, with no beta to be released before its time.

Linus Torvalds's style of developmentŚrelease early and often, delegate everything you can, be open to the point of promiscuityŚcame as a surprise. No quiet, reverent cathedral-building hereŚrather, the Linux community seemed to resemble a great babbling bazaar of differing agendas and approaches (aptly symbolized by the Linux archive sites, who'd take submissions from anyone) out of which a coherent and stable system could seemingly emerge only by a succession of miracles.

The fact that this bazaar style seemed to work, and work well, came as a distinct shock. As I learned my way around, I worked hard not just at individual projects, but also at trying to understand why the Linux world not only didn't fly apart in confusion but seemed to go from strength to strength at a speed barely imaginable to cathedral-builders.

http://www.catb.org/~esr/writings/homesteading/cathedral-bazaar/

This document is popular with The Crowd because they like to think that authority is not necessary, and somehow "people power" will get us through.

History suggests that of course that's bullshit, and that people power rapidly leads to anarchy which leads immediately to tyranny because anarchy paralyzes all societies. Anarchy and Libertarianism are basically the same thing, except Libertarians believe in free markets as the means to their anarchy.

Proof of this is the shameful fact that for the most part, "open source software" has only succeeded when it has made clones of existing commercial softwares -- Linux is a UNIX clone, Open Office is a Microsoft Office clone, and so on. The few notable exceptions like Apache and Emacs grew from academic projects.

Although I accept 100% that centralized, top-down, planned organization is the vital core of a society, I am interested by how bottom-up (bazaar) organization systems are often more efficient than top-down (cathedral) ones. Sometimes, free markets are better -- for example, in administering public works. Sometimes, socialism is better -- government should own and dispense some means of production, like roads and fissionable materials. The default is that if you don't manage something, it declines to the lowest common denominator, so the exceptions need to be carefully vetted. Even more, they work best in emerging markets where having everyone rush at the problem produces a few recognizable innovators who then become cathedrals on their own (see Apache, again).

Still, there's a lot of emphasis out there on fusing seeming opposites into singular systems:

  • Religion and science
  • Top-down and bottom-up
  • Capitalism and socialism
  • Death metal and black metal


It seems to me some common intellectual heritage -- a goal, a culture, a philosophy, a narrative -- must do this work. That and strong leaders unafraid to make unpopular but probably correct decisions,

I think bottom-up systems are useful for dealing with areas of total ignorance (i.e. when we have no idea where something is going to lead), and conversely top-down systems are best when we have a grasp of what we are doing.  Therefore, most of the time some combination of the two is best.  A top-down overarching plan, with bottom-up aspects for specific areas (e.g. economics).

Conservationist: "Although I accept 100% that centralized, top-down, planned organization is the vital core of a society, I am interested by how bottom-up (bazaar) organization systems are often more efficient than top-down (cathedral) ones. Sometimes, free markets are better -- for example, in administering public works. Sometimes, socialism is better -- government should own and dispense some means of production, like roads and fissionable materials. The default is that if you don't manage something, it declines to the lowest common denominator, so the exceptions need to be carefully vetted. Even more, they work best in emerging markets where having everyone rush at the problem produces a few recognizable innovators who then become cathedrals on their own (see Apache, again)."

Sometimes, free markets are better -- for example, in administering public works. - free market is based on private property, a public work resolved by free market is first turning it into private property. Public domain is theft on basis of propaganda. If a community considers this square this road this forest etc as free to use for all it's members than why a government to tax them for this? as they arranged, by their local elites, the status of that place or good, they could as well work together or find someone to hire for the job... no need for some third party to be involved. that's how things worked once.

A centralized planned organization is one thing. A (centralized planned) monopolist of organization (by monopolizing protection, and by this any other domain of human subjects under it(...) ) is another thing. For the first may very well be a protection firm but the second is always a form of government.A community may use service of a protection firm that it choses, or other, and change it if they do not do their job. Of course that is if they think they cannot protect themselves. You may know already that a libertarian sees the individual in full right of self protection, meaning the natural right of guarding your life and property. Now, you can easily understand why any government would not be very happy to rule a population that can guard itself...

On "public domain" issue you should understand that any public good or domain etc is taken over by government by simply and in a very democratic (socialistic) way luting private property, using "national", "common" needs argument sustained by "scientific" figures, statistics etc All of them being proven mere lies and completely driven by anything but selfless interests, please consult Mises and Rothbard's works on this. Any move in economy made by the stat cannot be natural and useful, as it has no way to calculate profits, only "great goals" and some ideology (ALWAYS socialistic, be it a nazi state or a democracy or anything in between) to back it up. I agree that there are governments that fail more often then the others and, funny (?) thing,  hitler knew maybe more about basic things in economy than many of the chiefs of states nowadays, but these are degrees of failure, couse neither nazi germany nor soviet russia nor democrat usa are exemples of great societies.

The natural way as it once was is based on homesteading, natural elites, uninvented laws (private property), armed population. Any state, any government is against this. Politicians are in no way natural elites, as democracy is in no way natural. Private property is hated by any form of government, no matter as liberal etc it claims to be. Any government sees itself as the final judge and legislator, hence unnatural invention of rules and laws. The far goal of the natural vision on society is a society where all known land and resources (or that we now consider as resources) is privately own. No artificial hierarchies, no "public" domains, no multicultural,democratic,socialistic "society". (no "free" immigration)

Society as organism, as cathedral, and other romantic (oh, pardon, scientific!) metaphors are those that inflict deviance in the views upon society and human life.If society is something that it is to be this when free, not when will of some individual, be it a great greek philosopher, an afro american president, some crazy russian or austrian and so on.

I'm not certain that Linux really is a great example of a truly "decentralized" system.  Yes, there's no overriding authority, but it is, relatively speaking, a very small, highly self-selecting and basically like-minded group of people - much closer in spirit to the "small group of mages working in splendid isolation" than to a true bazaar.  It's not exactly surprising that groups the consist mostly of self-selecting, self-motivating individuals of the higher type don't need the same level of...supervision and central direction to achieve great things. 

I'm not certain that Linux really is a great example of a truly "decentralized" system.  Yes, there's no overriding authority, but it is, relatively speaking, a very small, highly self-selecting and basically like-minded group of people - much closer in spirit to the "small group of mages working in splendid isolation" than to a true bazaar.  It's not exactly surprising that groups the consist mostly of self-selecting, self-motivating individuals of the higher type don't need the same level of...supervision and central direction to achieve great things. 

This is all true, as is the truth that Linux wasn't an invention -- it was a clone. I think we'll find that most really useful things came from small groups of smart, motivated people. Like ANUS.

I'm not certain that Linux really is a great example of a truly "decentralized" system.  Yes, there's no overriding authority, but it is, relatively speaking, a very small, highly self-selecting and basically like-minded group of people - much closer in spirit to the "small group of mages working in splendid isolation" than to a true bazaar.  It's not exactly surprising that groups the consist mostly of self-selecting, self-motivating individuals of the higher type don't need the same level of...supervision and central direction to achieve great things. 

This is all true, as is the truth that Linux wasn't an invention -- it was a clone. I think we'll find that most really useful things came from small groups of smart, motivated people. Like ANUS.

I would add that even in organizations that otherwise operate in a rigid, top-down fashion, elites are often given much more latitude and room for initiative because they are elites.  Special Forces types in the US military operate under a much more informal chain of command and with substantially reduced levels of institutional chicken shit because they can be trusted to do so, and the operational flexibility that confers on them is a serious "force multiplier," but no one is rushing to apply Special Forces style informality to the rest of the military, either.


Quote
I had been preaching the Unix gospel of small tools, rapid prototyping and evolutionary programming for years. But I also believed there was a certain critical complexity above which a more centralized, a priori approach was required. I believed that the most important software (operating systems and really large tools like the Emacs programming editor) needed to be built like cathedrals, carefully crafted by individual wizards or small bands of mages working in splendid isolation, with no beta to be released before its time.

Linus Torvalds's style of developmentŚrelease early and often, delegate everything you can, be open to the point of promiscuityŚcame as a surprise. No quiet, reverent cathedral-building hereŚrather, the Linux community seemed to resemble a great babbling bazaar of differing agendas and approaches (aptly symbolized by the Linux archive sites, who'd take submissions from anyone) out of which a coherent and stable system could seemingly emerge only by a succession of miracles.

The fact that this bazaar style seemed to work, and work well, came as a distinct shock. As I learned my way around, I worked hard not just at individual projects, but also at trying to understand why the Linux world not only didn't fly apart in confusion but seemed to go from strength to strength at a speed barely imaginable to cathedral-builders.

http://www.catb.org/~esr/writings/homesteading/cathedral-bazaar/

This document is popular with The Crowd because they like to think that authority is not necessary, and somehow "people power" will get us through.

History suggests that of course that's bullshit, and that people power rapidly leads to anarchy which leads immediately to tyranny because anarchy paralyzes all societies. Anarchy and Libertarianism are basically the same thing, except Libertarians believe in free markets as the means to their anarchy.

Proof of this is the shameful fact that for the most part, "open source software" has only succeeded when it has made clones of existing commercial softwares -- Linux is a UNIX clone, Open Office is a Microsoft Office clone, and so on. The few notable exceptions like Apache and Emacs grew from academic projects.
You're completely missing the point. ESR's cathedral and bazaar metaphor is about top-down and bottom-up design, not  systems. For example:

The Cathedral model, in which source code is available with each software release, but code developed between releases is restricted to an exclusive group of software developers. GNU Emacs and GCC are presented as examples.

The Bazaar model, in which the code is developed over the Internet in view of the public. Raymond credits Linus Torvalds, leader of the Linux kernel project, as the inventor of this process. Raymond also provides anecdotal accounts of his own implementation of this model for the Fetchmail project.

Both methods of design are pretty much exclusively top-down, as these projects are maintained by a small group of volunteers.

Emacs didn't grow from an academic project, it's based on the primitive TECO editor (which was non-WYSIWYG). Linux started out as a Minux clone, not a UNIX clone. Calling something a "UNIX clone" is as ridiculous as calling modern FPS games "Doom clones". UNIX is a specification and has a long history of succesful clones, like BSD. Four modern variants of BSD are OpenBSD, FreeBSD, NetBSD and OSX. The first 3 being very succesful open-source projects. OpenOffice is based on StarOffice, which was a project started in the mid-80s, a few years before MS Office. Note that Windows' & Office's GUI was ripped off from research projects at Xerox PARC (just like Apple ripped off the GUI stuff).

Your argument that open-source software has only succeded when it made clones of succesful commercial software is just ridiculous. You could say the same thing about commercial software. You just cherry picked some open-source software to proof your point, as there are hundreds of thousands (if not millions) of open-source projects. Here's some more succesful open-source software that aren't clones of commercial ones: Bash, XCFE, Wine, Blender3D, VLC media player, X-Chat, GCC, Google Chrome, Konqueror (the rendering engine of Apple's Safari called WebKit is a fork of Konqueror's KHTML), Perl, OpenSSH, the GNU toolchain, ReiserFS (a true killer app!), ext4, OGG, FLAC, MySQL, SQLite, and there are hundreds of thousands more.

Ofcourse they're all based on early work, so quod erat demonstrandum they're all clones!

NHA

Quote
This document is popular with The Crowd because they like to think that authority is not necessary, and somehow "people power" will get us through.

History suggests that of course that's bullshit, and that people power rapidly leads to anarchy which leads immediately to tyranny because anarchy paralyzes all societies. Anarchy and Libertarianism are basically the same thing, except Libertarians believe in free markets as the means to their anarchy.

The difference is that bad decisions and idiots are strongly selected against. If someone sees something that works terribly in open source software its gonna get replaced.


Either way, research scientists are the ones driving things forward, not engineers and product managers (although its possible to be all 3 at once). It really doesn't make a whole lot of difference if they're being funded commercially or not, the money just needs to come from somewhere.

Lets look at photoshop and gimp for example:
http://chathuraw.wordpress.com/2010/03/29/photoshop-cs5-content-aware-fill-vs-gimp-resynthesizer/

- Photoshop just recently added this feature which gimp has already had for a while.
- The tool is based off the research in this paper: http://portal.acm.org/citation.cfm?id=1576246.1531330
- 2 of the 4 authors of that paper are at Adobe Research, while the others are at Princeton Univ