perl, patch, configure -- developer-facing, highly derived from original UNIX work
www/browsers -- giving you this one, as noted above; gopher is part of that but I haven't checked the origin
muds -- nope, originally developed on an Essex University mainframe in 1978, later licensed to CompuServe commercially; see this history from the co-inventor
markup language/links -- invention of these concepts is somewhat disputed, but none of the major early candidates I know of (SGML, Xanadu, HyperCard) were open source software projects
the open source development process -- sorry, I'm just not impressed by what I've seen of the software in this space, so it's hard for me to sing the praises of the system
I'm not using an inferior platform; I'm using a Mac. But there's no Mac version, so I started trying the Windows box they make me use at work. Looks like it would take hours just to install and configure the thing -- talk about old thinking....
Would it be safe to say that you did no historical research in compiling your list?
Anything that was part of BSD was not developed as open source. BSD was treated as a proprietary operating system. So much for nethack as an example -- BTW, nethack is derived from rogue, another part of UNIX.
You have not named a particular chess client that you believe created the product category and was open source or free software.
I found two pages on X Window history (here and here). Both clearly frame X/Windows as a derivative product that came about after other GUI platforms. (It was also radically inferior to the Macintosh, even though it came afterwards.)
GCC is a clone of the Unix CC.
TeX may be an example of an innovative piece of software that originated under free/open conditions. I can't find anything to the contrary. I have never used TeX so I can't comment on whether it really deserves to be considered innovative; I will note that it came after troff, and that although it came after the GUI revolution it missed the boat, so there may be reason to consider it regressive rather than innovative.
And so on. I don't know all of your examples, but the ones I knew did not encourage me to bother checking out the rest. If you continue to defend that list, please give specific references.
Emacs did not begin as an open source project. It was part of a proprietary operating system, Honeywell's Multics. You can read more than you want to know about its history here.
Unix was not an open source project. It used to cost tens of thousands of dollars for a commercial source license.
Perl is a very poorly designed language made mostly by cobbling together features and syntax from other UNIX utilities. Awk, yacc and lex are much better designed string manipulation languages.
Distributed file systems are nothing new, and did not originate in the open source world. Andrew was originally created under high security conditions at Carnegie-Mellon's ITC (where Jim Gosling worked, and where I had a summer job.) Though it was funded by IBM, that was long before IBM jumped on the open source bandwagon.
NCSA Mosaic is derivative from the original CERN web browser by Tim Berners-Lee. However, the CERN browser was released in a free software/open source (FS/OS) fashion (see the license).
We could quibble about whether the first web browser was really an innovation or not. Berners-Lee has been quite modest, as well as very honest about his debts to SGML, HyperCard, and Xanadu.
Still, prior art notwithstanding, it does seem reasonable to accept web browsers as an FS/OS innovation, by which I mean a new user-facing product category. In fact it was the biggest new product category of the nineties. So this should be taken as a partial answer to concerns about FS/OS stifling innovation.
It's not a contest. The question is whether open source innovates, not whether Microsoft does. I agree that Microsoft has been an imitative company but that has nothing to do with the point at hand. Other commercial software vendors clearly have created new user-facing product categories.
The chief claim made by Allchin was that free software and open source (FS/OS) do not innovate. That issue was discussed in this/. thread. No thread participant could provide an example of a new user-facing product category that had originated in the FS/OS world. (In addition, absurd claims were made that open source projects had invented skinning, virtual memory, and so on; and that a command-line network software installer was somehow an innovation.) Stallman does not provide examples either, nor did the Red Hat CTO.
Although it pains me to agree with an M$ representative who is speaking in his company's commerical interests, I must agree with Allchin that free software and open source primarily clone existing user-facing commercial products rather than innovating in that space.
Jabber appears to be a me-too messaging client, at least from its web site. It looks like every other such client I've seen. "Innovative"? Just because it talks XML under the hood? Please.
Enhydra is indeed a fine project, and it's the only actually innovative software I've seen listed in this entire thread. I do think developer-facing products like this have the potential of delivering some of the claimed benefits of open source, but what about user-facing products?
Zope is a slashdot clone. Some innovation.
Postgresql is a me-too database product, yet another imitation of already existing products.
KDE and GNOME skinnability came years after Apple's theme switching announcements, and after a third party (Kaleidoscope) beat Apple to market with Mac OS skins.
So far, we have exactly one innovative product in this thread. For that reason, I don't see a reason to change my position: Open source remains a haven for imitations of commercial software. Open source has never created a new user-facing software product category. Open source does not innovate except for the occasional developer-facing product.
BSD brought us virtual memory among other technologies. Surely you think that is innovative?
Right. Ever hear of a VAX?
BSD is another clone operating system. It's hard to call it innovative, although in some ways it is technically excellent.
And lest we forget, BSD was not an open source program when it was being developed. It was tightly held by the Regents of UC and took years to get released in source form to the public.
LyX is a joke, apparently. It's supposed to be a Windows program but it has twenty pages of installation instructions. Needless to say, I didn't bother evaluating its user-friendliness any further than that. You guys just don't get it.
Well, there's no need to be so rude, but I have to agree with your content. Both of the well-known champions of free software or open source, Stallman and Raymond, have based their entire platforms on bizarre claims that didn't add up, and as a result they both have increasingly come to be seen as embarassments.
In Raymond's case, his Libertarian dogma led him to pronounce that open source would make tremendous amounts of money and be a fabulous success, although as far as anyone can tell, Raymond has never been involved with creating an actual business plan, successful or otherwise. There was no business model behind his pronouncement, only the zeal of the religious devotee. Because the open source movement's value claims were snake oil, we are now left looking at the fading remnants of the Linux bubble.
In the message you quoted, Raymond states,
This is a question that a lot of us will be facing as open source sweeps the technology landscape. Money follows where value leads, and the mainstream business and finance world is seeing increasing value in our tribe of scruffy hackers.
If that's not embarassing now, I don't know what is. Open source has turned out to be marginal and only rarely even slightly profitable. The revolution will not be televised, because there is no revolution.
Yes, I have used apt-get. It's a network installer. There's nothing innovative about it.
WYSIWYG is an innovative approach which has numerous advantages over forcing someone to learn a programming language like troff or tex just to write a text document.
Your message is a good example of basic ignorance of design values, which is the norm in open source or free software projects. Name me three acknowledged principles of good design without looking them up, and then perhaps you will have demonstrated that you have some actual background in these areas.
Open-box software is not a panacea -- it does not solve all the problems. It still requires all of the discipline in development and operation that we would like to see in proprietary closed-box software. But it has enormous potential, and needs to be pursued as a serious contender.
In other words, open source software could be more reliable than closed source software, if it were produced in a disciplined way. Anyone who imagines that it is created in that way ought to spend a little more time at Freshmeat.
I don't know what commercial projects you've had the pleasure to work on, but that's not been my experience, and I've worked for a lot of closed-source companies.
Well, one of my several commercial projects, Mac OS 8, was at the only mainstream company on your resume, Apple. I can assure you that we agaonized over the bug curve for the user interface software I worked on. Ditto for every other significant commercial project I've ever worked on.
You think your commercial vendor uses test suites? Guess again. It's so frustrating when I have a client who I cannot convince there's a reason to actually perform QA of any sort, let alone use test suites.
I sympathize, although since I know you have used RADAR at Apple, I wonder why you are so insistent that major developers don't track bugs or use test suites. At Apple User Experience, we had written test suites for all our components.
My answer to the problem of development managers not caring about bugginess has been to refuse to work with any software company that does not have a strong SQA organization or does not plan to create one. Fortunately I have never been short of work. All major software companies have already adopted quality methodologies and they do pay off. Unit testing, test suites, and bug tracking are the norm in serious commercial software development.
Then you have the seat-of-the-pants guys who just won't use a quality process. In other words, they're just as bad as the open source community. Fortunately, they are a minority among major software developers.
I remain unclear on how a methodology with no quality standards can be more reliable than a methodology which does. It seems rather like an oxymoron. Raymond claims that given enough eyeballs, all bugs become shallow. The Debian and Mozilla bug curves demonstrate beyond reasonable doubt that he is wrong. Open source software is extremely buggy and the only reason people aren't more aware of that is that most open source projects don't track bugs in any useful fashion.
I would be curious how Dr. Neumann responds to these facts. The two largest open source projects which have tracked bugs in a useful way, Mozilla and Debian, have both shown ever-expanding defect curves, and Mozilla now sits at upwards of ten thousand open bugs. How is this a demonstration of enhanced reliability by the open source method?
Commercial projects I've worked on have always shown a distinctly downwards bug trend during alpha and beta phases, while it appears that open source bugs just go up and up -- when anyone even bothers to chart them.
Your list of programs is made up almost entirely of knockoffs, copies, emulators, and similar cloneware. I'm not sure how you can argue that open source creates innovation based on this list:
implements similar functionality to windows ICQ
conforms to the MPEG audio encoder standard
arcade emulator
based on the NCSA httpd
another POSIX implementation
(The only thing even marginally innovative here is apt-get, a lame command line tool that works around the lack of software packaging standards in Linux. It's not present on other systems only because they have no need for it.)
The open source community has yet to produce anything as innovative as the WYSIWYG word processor or the spreadsheet metaphor. Open source does not innovate in end-user space, only in developer space. All popular end-user open source applications are knockoffs of better-known commercial applications, e.g., the GIMP as a wannabe Photoshop, and StarOffice as a wannabe M$ Office.
The problem with the Ramen worm was not with Red Hat, but with sysadmins that don't frequently update and maintain their systems.
Perhaps to look a little more into the root causes of the problem, the problem is with fragile software systems which require frequent updates and maintenance.
You got moderated up to "4, Insightful" for missing the entire point of his statement about the OS as application? And the people who corrected you didn't get a single moderation point? Wow. No wonder the Slashdot moderation system is the envy of the world.
Yes, there is the likelihood of increasing the percentage of infertile people in the population if cloning becomes a common way of reproducing. This issue was flagged by bioethicists years ago.
Slashdot Mirror
perl, patch, configure -- developer-facing, highly derived from original UNIX work
www/browsers -- giving you this one, as noted above; gopher is part of that but I haven't checked the origin
muds -- nope, originally developed on an Essex University mainframe in 1978, later licensed to CompuServe commercially; see this history from the co-inventor
markup language/links -- invention of these concepts is somewhat disputed, but none of the major early candidates I know of (SGML, Xanadu, HyperCard) were open source software projects
the open source development process -- sorry, I'm just not impressed by what I've seen of the software in this space, so it's hard for me to sing the praises of the system
Tim
Tim
Anything that was part of BSD was not developed as open source. BSD was treated as a proprietary operating system. So much for nethack as an example -- BTW, nethack is derived from rogue, another part of UNIX.
You have not named a particular chess client that you believe created the product category and was open source or free software.
I found two pages on X Window history (here and here). Both clearly frame X/Windows as a derivative product that came about after other GUI platforms. (It was also radically inferior to the Macintosh, even though it came afterwards.)
GCC is a clone of the Unix CC.
TeX may be an example of an innovative piece of software that originated under free/open conditions. I can't find anything to the contrary. I have never used TeX so I can't comment on whether it really deserves to be considered innovative; I will note that it came after troff, and that although it came after the GUI revolution it missed the boat, so there may be reason to consider it regressive rather than innovative.
And so on. I don't know all of your examples, but the ones I knew did not encourage me to bother checking out the rest. If you continue to defend that list, please give specific references.
Tim
Emacs did not begin as an open source project. It was part of a proprietary operating system, Honeywell's Multics. You can read more than you want to know about its history here.
Tim
Perl is a very poorly designed language made mostly by cobbling together features and syntax from other UNIX utilities. Awk, yacc and lex are much better designed string manipulation languages.
Distributed file systems are nothing new, and did not originate in the open source world. Andrew was originally created under high security conditions at Carnegie-Mellon's ITC (where Jim Gosling worked, and where I had a summer job.) Though it was funded by IBM, that was long before IBM jumped on the open source bandwagon.
Tim
We could quibble about whether the first web browser was really an innovation or not. Berners-Lee has been quite modest, as well as very honest about his debts to SGML, HyperCard, and Xanadu.
Still, prior art notwithstanding, it does seem reasonable to accept web browsers as an FS/OS innovation, by which I mean a new user-facing product category. In fact it was the biggest new product category of the nineties. So this should be taken as a partial answer to concerns about FS/OS stifling innovation.
Tim
Tim
(What the hell do I know about it? Well, I wrote my first TCP/IP stack in 1984....)
Tim
In addition, and as I pointed out in the previous thread, emulators are cloneware by definition.
Tim
Although it pains me to agree with an M$ representative who is speaking in his company's commerical interests, I must agree with Allchin that free software and open source primarily clone existing user-facing commercial products rather than innovating in that space.
Tim
Saying that open source innovated by creating itself seems rather tautological.
What has it created in software? Can you name one user-facing (not developer-facing) product category that was invented in the open source world?
Tim
Jabber appears to be a me-too messaging client, at least from its web site. It looks like every other such client I've seen. "Innovative"? Just because it talks XML under the hood? Please.
Enhydra is indeed a fine project, and it's the only actually innovative software I've seen listed in this entire thread. I do think developer-facing products like this have the potential of delivering some of the claimed benefits of open source, but what about user-facing products?
Zope is a slashdot clone. Some innovation.
Postgresql is a me-too database product, yet another imitation of already existing products.
KDE and GNOME skinnability came years after Apple's theme switching announcements, and after a third party (Kaleidoscope) beat Apple to market with Mac OS skins.
So far, we have exactly one innovative product in this thread. For that reason, I don't see a reason to change my position: Open source remains a haven for imitations of commercial software. Open source has never created a new user-facing software product category. Open source does not innovate except for the occasional developer-facing product.
Tim
Right. Ever hear of a VAX?
BSD is another clone operating system. It's hard to call it innovative, although in some ways it is technically excellent.
And lest we forget, BSD was not an open source program when it was being developed. It was tightly held by the Regents of UC and took years to get released in source form to the public.
Tim
Tim
Tim
In Raymond's case, his Libertarian dogma led him to pronounce that open source would make tremendous amounts of money and be a fabulous success, although as far as anyone can tell, Raymond has never been involved with creating an actual business plan, successful or otherwise. There was no business model behind his pronouncement, only the zeal of the religious devotee. Because the open source movement's value claims were snake oil, we are now left looking at the fading remnants of the Linux bubble.
In the message you quoted, Raymond states,
This is a question that a lot of us will be facing as open source sweeps the technology landscape. Money follows where value leads, and the mainstream business and finance world is seeing increasing value in our tribe of scruffy hackers.
If that's not embarassing now, I don't know what is. Open source has turned out to be marginal and only rarely even slightly profitable. The revolution will not be televised, because there is no revolution.
Tim
WYSIWYG is an innovative approach which has numerous advantages over forcing someone to learn a programming language like troff or tex just to write a text document.
Your message is a good example of basic ignorance of design values, which is the norm in open source or free software projects. Name me three acknowledged principles of good design without looking them up, and then perhaps you will have demonstrated that you have some actual background in these areas.
Tim
Open-box software is not a panacea -- it does not solve all the problems. It still requires all of the discipline in development and operation that we would like to see in proprietary closed-box software. But it has enormous potential, and needs to be pursued as a serious contender.
In other words, open source software could be more reliable than closed source software, if it were produced in a disciplined way. Anyone who imagines that it is created in that way ought to spend a little more time at Freshmeat.
Tim
Well, one of my several commercial projects, Mac OS 8, was at the only mainstream company on your resume, Apple. I can assure you that we agaonized over the bug curve for the user interface software I worked on. Ditto for every other significant commercial project I've ever worked on.
You think your commercial vendor uses test suites? Guess again. It's so frustrating when I have a client who I cannot convince there's a reason to actually perform QA of any sort, let alone use test suites.
I sympathize, although since I know you have used RADAR at Apple, I wonder why you are so insistent that major developers don't track bugs or use test suites. At Apple User Experience, we had written test suites for all our components.
My answer to the problem of development managers not caring about bugginess has been to refuse to work with any software company that does not have a strong SQA organization or does not plan to create one. Fortunately I have never been short of work. All major software companies have already adopted quality methodologies and they do pay off. Unit testing, test suites, and bug tracking are the norm in serious commercial software development.
Then you have the seat-of-the-pants guys who just won't use a quality process. In other words, they're just as bad as the open source community. Fortunately, they are a minority among major software developers.
I remain unclear on how a methodology with no quality standards can be more reliable than a methodology which does. It seems rather like an oxymoron. Raymond claims that given enough eyeballs, all bugs become shallow. The Debian and Mozilla bug curves demonstrate beyond reasonable doubt that he is wrong. Open source software is extremely buggy and the only reason people aren't more aware of that is that most open source projects don't track bugs in any useful fashion.
Tim
Commercial projects I've worked on have always shown a distinctly downwards bug trend during alpha and beta phases, while it appears that open source bugs just go up and up -- when anyone even bothers to chart them.
Tim
(The only thing even marginally innovative here is apt-get, a lame command line tool that works around the lack of software packaging standards in Linux. It's not present on other systems only because they have no need for it.)
The open source community has yet to produce anything as innovative as the WYSIWYG word processor or the spreadsheet metaphor. Open source does not innovate in end-user space, only in developer space. All popular end-user open source applications are knockoffs of better-known commercial applications, e.g., the GIMP as a wannabe Photoshop, and StarOffice as a wannabe M$ Office.
Tim
Perhaps to look a little more into the root causes of the problem, the problem is with fragile software systems which require frequent updates and maintenance.
Tim Maroney
Tim
Correct me if I'm wrong, but that is a tautology, yes?
"Many products are well liked by people who don't return them."
Tim
Tim