quote: Original post by tortoise
Would ya mind sliding that crystal ball this way? I''d like to get a peak at my future wife, maybe next week''s lotto numbers...

They are irrelevant because what you commented on was not that Linux and open-source makes more revenue than commercial/closed-source software. When I purchase a server that I intend to run Linux on, I do not buy it with Linux bundled, and I imagine most do as I. Therefor, the number of "Linux servers" sold by vendors doesn''t have anything to do with how many systems are out there.

quote: Any sources on that? Any sources at all? Hmmm? Anyone? I''ll even give you a freebie, Netcraft has rated Apache as the most used web server for many years now. Apache is one of the very few documented cases of open source being more succesful than commercial, it''s quite possibly the only case. You can spout stuff out your ass till your blue in the face, doesn''t somehow make it true.

Do you have any sources for the countrary? Apache is hardly the only one. The most important protocols on the Internet are DNS, SMTP and HTTP, for which BIND, sendmail and Apache, respectivly, are indeed *the standard* software for providing. Apache has the least market share of these; I think I read somewhere than BIND handles more than 98% of all DNS queries made on the Internet.

quote: My company''s servers are roughly 90% Linux. I''ve got Linux installed on two boxes in this room, and all my friends use Linux. Despite all that, I know my very small view of the industry does not somehow reflect the whole, you''ve not yet realized this. Like I said, take those blinders off. Linux has got a long way to go. It will only go that way if the money''s there. Period.

I''m not sure what you are arguing about.

I''m wondering, how does a corporation like Red Hat exist if there''s practically no money going around?

What the hell is funding development?

quote: Original post by Promit
I''m wondering, how does a corporation like Red Hat exist if there''s practically no money going around?

What the hell is funding development?

They sell support. It works quite well. For example, currently I develop web applications, but our revenue is not primarily the software itself, but what we charge for implementation/customization and support contracts for the the implemented software. When we put down time into development, it enables us to sell customization and support.

Michalson,

I already answered this weak FUD spreading argument of yours in a previous posting on this same topic. Maybe you missed it:

FSF & Communism FUD

Besides the fact that Stallman is a freak, it is a poor rhetorical trick to debase an idea or group based an offtopic fact/opinion.

Once again, if you choose to give donate your code to the FSF and provide it via the GPL you may (probably should) retain the copyright and you retain the extended privileges that go with originating the software, namely, IP rights to the techologies you''ve developed. This includes the ability to sell a wholely original product that you''ve GPL''ed without providing the source. In fact you can even sell the source and product to another company to distribute at their own discretion as well, outside the terms of the GPL. (reference Trolltech''s qtopia product). Additionally if a company wants to implement your IP outside of a GPL product then they must purchase the IP from you.

The GPL simply prevents another company from profiting from the complete or subset of source you have chosen to provide to the community free of charge without providing their additions back to you and the community. In this case you are entitled to the source code for their additions free of charge (hence undermining their ability to really profit from your ideas). In the case of accepting additions back to the source you are prevented from incorporating the additions into your own proprietary dealings.

The BSD license makes no such requirments and provides no such freedoms as far as I know.

RandomTask

quote: Original post by tortoise

quote: Original post by Oluseyi
Distros that depend on unit sales ought to die as commercial entities, perhaps to be reborn as purely community-driven initiatives a la Debian.

So who gets the awful of job of writing new OS code for zero bucks an hour, in their free time at that? I sure wouldn''t take that from my boss, why should the distros?

Well, maybe we''ve hit the core issue: if an enterprise is not commercially sustainable, then don''t build a company around it. Building and selling OSes will not remain viable for much longer, as applications are divvied up by the hardware peripherals best suited to them. Hell, even traditional applications are becoming an annoyance: consider that every time you purchase a digital music player the vendor expects you to install a proprietary application on your computer for "managing" the device, creating a situation where you might have several redundant (because they do the same thing, to varying levels of competence) applications on your computer. It''s one thing if these applications are Open Source; you can remove them, and they don''t usually have proprietary data formats associated with them (replacing Konqueror, lynx, links, and Galeon with Mozilla, for example).

My point is that Open Source applications are commodities. It is in the provision of complementary products and services that there is money to be made. That''s how Red Hat, SuSe, Penguin Computing and many other Linux companies survive, selling support, customizations, training, consulting and hardware tuned for Linux-based solutions. For them, sponsoring/contributing to Linux software development is a no-brainer cost of operation, because doing so aids the growth of their potential customer base.

quote: Why is it so acccepted that the open source community will just create new software with no other incentive than to merely create it, and people then take and eat it up with barely even a thank you?

Perhaps because they do? PErhaps because that''s their entire philosophy? Have you ever read ESR''s Cathedral and the Bazaar and other essays? Not all software should be Open Source, but software that should/is, is because it''s effectively a commodity with few margins. Windows in and of itsef isn''t a compelling product; it''s the access to exclusive applications and intrinsic compatibility with the products and data of millions of other people worldwide that drives it.

quote: Volunteering is great and volunteer driven software is often great as well, but it has its limits. It''s no coincedence that commercial software, hell commercial everything, rules the land.

a.) I don''t give a rat''s ass about volunteering. b.) Commercial software is killing itself, because the producers don''t know which applications to focus on. Browsers used to be commercial products; today Opera is the only company still selling a browser (AOL retired Netscape) for PCs. AvantGo can sell its PDA browser because the market is still fairly exclusive; once it becomes a commodity like the PC, expect to see that go the way of the dodo, too. Commercia utility software is faced with two diametrically opposed impetuses: to provide compatibility with the products of other vendors; and to maintain exclusvity, thus justifying the price tag. In most categories, it can''t last. Windows (or whatever) may be the dominant server platform right now, but far enough into the future that market will be swallowed up - again - by commodities, at which point there will be virtually no margins and thus Open Source will dominate.

The money is in services and unique applications. It always has been, but with the explosion of computing and programming availability, the number of unique domains is constantly shrinking.

quote: Original post by tortoise
So who gets the awful of job of writing new OS code for zero bucks an hour, in their free time at that? I sure wouldn't take that from my boss, why should the distros?

Instead of buying an OS with your dollars, you take an OS with 99% of what you want and you buy the extra 1% with your time (your numbers may vary). It's not a difficult decision: pay a small amount with your time, or pay an amount of money sufficient to compensate all the developers for theirs, plus a profit sufficient to make the work attractive in favor of any other potential projects developers could do.

quote: Original post by Michalson

quote: Why is it so acccepted that the open source community will just create new software with no other incentive than to merely create it, and people then take and eat it up with barely even a thank you? Volunteering is great and volunteer driven software is often great as well, but it has its limits. It's no coincedence that commercial software, hell commercial everything, rules the land.

The Free Software Foundation (GNU/GPL) and the Open Source Software movement are two different things. The OSS movement is about having the sourcecode for software open so that people can see it and improve it (the difference say, between selling cars where the hood is welded shut and cars where you can easily take a look at the engine). While the fact remains that source code is a lot easier to pirate than some security through obscurity binary, there is nothing stopping people from selling OSS in exactly the same way that CSS is sold (that is, you can't download the source from any official site, you only get the source and binaries by buying them).

Actually, the schism between OSS and FS is about open source/free software being resold as a commercial product. (likely with modifications, or integrated with another system)

quote: The Free Software Foundation (created by Richard Stallman, self described "sexual pervert" who spent his years before computer communism trying to promote sexual communism) is about taking rights away from developers and giving them to inanimate software. The FSF is the organization that wants programmers to be slaves of the community, producing code and then being dictated that their hard work must be given freely to others (read the GPL licence for more on that).

Wow. Blast from the past. I thought those arguments went out of style in the '70s. First you think you're maligning his character with vague allegations of sexual misconduct (witness the election of Schwarzenegger -- nobody cares about that anymore). Then you throw out unsubstantiated claims about communism or totalitarianism. I see the Red Scare is alive and well.

I'm a libertarian; here's how I characterize Free Software:
Does your "right" to own your software trump my right to own my computer? No? Then I should be able to do with my computer, and the software that is an abstraction of the state of my computer, what I choose. I should be able to observe the state of my computer, and inform others of this state (i.e. share the software on my computer) I should be able to modify the state of my computer, and therefore the software on it. This accounts for the "what is free software" definition.

Now for the contentious parts of free software. It's not meaningful liberty if it requires superhuman abilities to exercise (i.e. read disassembled code, write assembly). Therefore, in order to secure these rights, developers should be required to distribute the source code to software whenever they distribute binaries, in order that the users can exercise their rights over their computers. That a legal regime exists that does not require this only demonstrates that the laws need to be changed.

The free software foundation has taken a reasonable and practical course by founding and furthering a microcosm where software is free, instead of lobbying for laws that were already entrenched. In so doing, some members found that the distributed development scheme that emerged was desirable in and of itself, and decided to push for the source-code availablity that is a prerequisite for that system, hence the existence of the OSI.

Copyright exists in America because the founding fathers desired a balance between the interests of authors and readers. I think the current regime (commercial software can exist, and using contract and copyright law, free software can exist) with perhaps slight modifications strikes that balance in software. In the future, either companies will exploit software patents to kill free software projects (killing also the balance between freedom and progress), or free software will gradually edge commerical projects out of the public eye and into niche markets (maintaining the balance. Wherever commercial software makes sense for buyers, sellers, and the public, it will exist. Where it does not, free software will exist).

[edit]Corrected who said what in my quotes of previous posts.

[edited by - Flarelocke on October 25, 2003 5:56:20 PM]

Because the GPL allows all free software projects to benefit from each other, free software will eventually overwhelm commercial software. OSS is based around the next big thing, service-based programming, where a programmer combines uses existing libraries plus a bit of customization code to meet a customer''s needs. As programs become more and more complex, the commercial software approach to programming, each company reinventing the wheel, will become wasteful and stupid.

Imagine, for a moment, if every computer game company in the world sat down and designed a single, cross-platform window class...it would have to be generalized as hell, have tons of options so that it could compile to the exact specifications of the programmer...but the end user could be guaranteed a well thought out, consistent interface across all games.

It''s just a matter of time before the wisdom of that approach becomes apparent.

quote: Original post by Ishan
Because the GPL allows all free software projects to benefit from each other, free software will eventually overwhelm commercial software. OSS is based around the next big thing, service-based programming, where a programmer combines uses existing libraries plus a bit of customization code to meet a customer's needs.

The vast majority of Open Source programming is dissipated effort: code rewritten, projects duplicated, etc. It's fine since no one is paying for it, but it's not exactly an ideal. Free software won't "overwhelm" commercial software; it will supplant it when there are no longer any margins to derive from it, as with DNS, web browsers and many other infrastructural systems.

quote: As programs become more and more complex, the commercial software approach to programming, each company reinventing the wheel, will become wasteful and stupid.

Companies don't reinvent the wheel today (as much as possible). They license components and write glue code. And more and more companies are being allowed to incorporate Open Source components, so long as the ownership is crystal clear (no SCO-like IP FUD).

quote: Imagine, for a moment, if every computer game company in the world sat down and designed a single, cross-platform window class...it would have to be generalized as hell, have tons of options so that it could compile to the exact specifications of the programmer...but the end user could be guaranteed a well thought out, consistent interface across all games.

Which wouldn't work. Games are highly specialized, and only very similar games really benefit from high level abstractions (ever noticed that all commercial products, and nearly all hobbyist projects, built on Quake engines are FPSes?) You can't even begin to build a coherent argument for these kinds of issues based on the microcosm that is game development.

As an aside, I now fervently hope for the abolition of software patents. We're coming to the point where they do more harm than good.

[edited by - Oluseyi on October 27, 2003 12:55:41 PM]

quote: The vast majority of Open Source programming is dissipated effort: code rewritten, projects duplicated, etc. It''s fine since no one is paying for it, but it''s not exactly an ideal. Free software won''t "overwhelm" commercial software; it will supplant it when there are no longer any margins to derive from it, as with DNS, web browsers and many other infrastructural systems.

I wonder if the reason apple and microsoft have primarily application-centric user interfaces is because a document-centric user interface discourages third-party software development by commoditizing software. (Linux has the same user-interface, pretty much, because linux''s users are developers, and the documents developers deal with are applications) In other words, there are only margins because a change in the user-interface to decrease user power means that applications which could be the integration of orthogonal components need instead to be constructed by the developer, who has skills no user can reasonably expect to learn.

quote: Which wouldn''t work. Games are highly specialized, and only very similar games really benefit from high level abstractions (ever noticed that all commercial products, and nearly all hobbyist projects, built on Quake engines are FPSes?) You can''t even begin to build a coherent argument for these kinds of issues based on the microcosm that is game development.

As one other user of these forums put it, profile before you post. Granted, not everything could be abstracted, but how far could you abstract various components of games? Lots of games have in and out of game cutscenes. Could one have a generic method of manipulating the in-game objects for in-game cutscenes, and a generic method for loading and displaying prerendered cutscenes? I think so, although it''s a bit beyond the reasonable abilities of C++, I think, though I could be wrong. That games need to be written from scratch or around a preexisting graphics engine seems to be an article of faith for game developers (not that I don''t see the difficulty involved; I just don''t see the impossibility).

quote: As an aside, I now fervently hope for the abolition of software patents. We''re coming to the point where they do more harm than good.

The economic foundation of patents is that in a market in which there is a high fixed cost of research and development, patents increase the return and decrease the risk. Practically speaking, they only increase the barrier to entry. There stopped being a high fixed cost of research when the personal computer was first introduced. It increased a bit when people shipped computers without compilers. It decreased a *lot* with gcc. Now any monkey living in his parents basement can do computer research. The cost of entry is about $500 these days. The increased return of a patented invention is often thousands of dollars, and the barrier to entry it raises can be insurmountable (companies simply refusing cross-licensing agreements for any amount). Yeah, if software patents ever had a time, it''s over now.

quote: Original post by Flarelocke
I wonder if the reason apple and microsoft have primarily application-centric user interfaces is because a document-centric user interface discourages third-party software development by commoditizing software.

Note that Apple has been making quite a bit of hoopla about its UIs being more document-centric (warning: it''s a long read, and I couldn''t be bothered to find the specific page reference).

quote: In other words, there are only margins because a change in the user-interface to decrease user power means that applications which could be the integration of orthogonal components need instead to be constructed by the developer, who has skills no user can reasonably expect to learn.

Absolutely. However, I believe that the next evolution of document editing applications will be a generic tablet into which specific functionality will be integrated by several vendors. Because the base application exists and each vendor only needs to write what is effectively a plugin, the overhead of development is lower and the returns are higher for both developer and consumer. All consumers become your market, not just, say, Microsoft Office version XYZ users, and consumers can purchase plugins from all developers.

But I''m thinking wishfully again.

quote: As one other user of these forums put it, profile before you post. Granted, not everything could be abstracted, but how far could you abstract various components of games? Lots of games have in and out of game cutscenes. Could one have a generic method of manipulating the in-game objects for in-game cutscenes, and a generic method for loading and displaying prerendered cutscenes?

The former is called a script or scripting language, the latter is called playing a movie.

I''m as much an advocate of reuse as the next guy, and I actually think that more reuse can be achieved through the creation and publication of software routine libraries which each developer can cobble together into an "engine" specific to his/her project. I just didn''t buy your analogy.

quote: The economic foundation of patents is that in a market in which there is a high fixed cost of research and development, patents increase the return and decrease the risk. Practically speaking, they only increase the barrier to entry. There stopped being a high fixed cost of research when the personal computer was first introduced. It increased a bit when people shipped computers without compilers. It decreased a *lot* with gcc. Now any monkey living in his parents basement can do computer research. The cost of entry is about $500 these days. The increased return of a patented invention is often thousands of dollars, and the barrier to entry it raises can be insurmountable (companies simply refusing cross-licensing agreements for any amount). Yeah, if software patents ever had a time, it''s over now.

What bothers me the most is the patenting of nebulous ideas - so-called "business processes". These aren''t even applications that run, prototypes or executable code. No, they''re simply ideas like "a method for conducting a complete merchandise and credit transaction within a single step via a web-based interface" which, once patented, allows you to sue Amazon over One-Click.

Software patents are evil.