For what it's worth, here's my take on the issue. It looks to me like the 2.2(c) reporting requirement issue is annoying, but it won't lead to the license disappearing. They've got that one satisfactorily covered.
The 9.1 point is still a problem. I don't care what Apple intended Affected Original Code to mean. It is not clearly defined in the license, and I can see all sorts of abuse that can come from a loose definition of the term.
I don't care whether it is technically Open Source or not, I don't want a license which can end my legal use of a program on a company's whim. I think there is a clear need for a "No Termination of Use" clause to be inserted into the Open Source guidelines. While there is not one, the Open Source definition becomes pretty useless to me.
Section 7 does not supercede the termination clause in Section 9.1. All it says is you get to pick which version of the APSL you use and distribute under.
On your other point, if you have a GPL'ed program that gets into a patent lawsuit, you lose the right to redistribute under the GPL. You don't lose the right to use it under the GPL, which means that you can wait for the patent to expire, or you can negotiate use of the patent directly with the patent holder.
If an APSL program gets into a patent lawsuit, Apple can pull your rights to use the software regardless of whether the patent later expires or if you negotiate your own patent license.
As for the APSL only terminating Affected Code, the license does not clearly define "Affected Original Code", basically the effect of this is it can mean whatever their lawyers say it means, and if you think it means otherwise, be prepared to pay for your own lawyers. If there is a lawsuit on a couple of lines in a library, they can easily say the "Affected Original Code" is the entire library, and anything which uses it. This is why Bruce asked that the APSL at minimum clarify terms like this.
As for you update and everything is hunky dory, that assumes that 1) Apple releases an update, and 2) Apple releases the update under the APSL. If Apple invokes this clause, third parties will not be legally allowed to fix the problem. Such terms are against both the spirit of Free Software and the legal definition of Open Source. We don't have to stand for it, and we shouldn't.
As far as what has he done personally, I think he did a little of the coding work in MS-BASIC (the program that started Microsoft). AFAIK, most of the first version of MS-BASIC was done by Paul Allen and then the employees that they hired after they sold some early copies.
As far as what has his company developed themselves, while Windows (I'm talking versions 1 and 2 here) was mostly other peoples ideas, I think it was coded by Microsoft. For Windows 3, 95 and 98 they contracted out a lot of it.
Windows NT 3.5 was just OS/2 and VMS run together really fast with a Windows 3 GUI put on it, in fact the developers, and much of the code, were grabbed from the OS/2 design team and Digital.
TrueType was a purchase from Apple (they developed it and were discarding it in favor of Adobe Type1). I think Word, PowerPoint and Xenix were purchased, I don't know from who.
I think Microsoft Bob was written by them, I'm not sure. I also think the eariler versions of Masm and Microsoft C were written by them. Microsoft BASIC was written by them, and I think its decendants: BASICA, QBASIC and Visual Basic.
You said "the Internet he did legitimately add to his OS, even if it was around for ages before". They didn't write their internet stuff. Winsock (i.e. "the Internet") was a kludged port of the same BSD networking stuff used in most Unixes. The same thing with most of the "standard" internet tools, Telnet, FTP, Ping, Traceroute. And, as you've said, Internet Explorer was purchased from SpyGlass.
Wordstar did not become Word. Wordstar was still around when Word started to become popular. They responded by dumping most of their original code (which I hear was some of the most incredible hand-optimized assembly ever written), and making Wordstar2000 (in 1985) which flopped so badly that the company was never heard from again.
As far as I know, the only actual code that Bill Gates worked on was MS-BASIC, and even that one was mostly done by Paul Allen.
Section 7 of the GPL terminates your right to distribute a program with contested code. Section 9.1 of the APSL terminates your right to use a program with contested code. There's a very big difference between the two.
Just think, if anyone makes a claim against Apple, they have the option to, on a whim, invalidate all of their userbase. If they feel they tire of being Open Source, they can release a new version under a solidly closed license, and convince someone to make a claim against the open version. Bingo, everyone goes from having a nice almost open-source system, to having to pay for a proprietary upgrade just to use their machine legally. All derivative works would likewise be killed.
This is not nitpicking, this is a clause that keeps the software from being Free. While I like Apple, they are a corporation. That means that their first responsibility is to their shareholders. That means if they are not contractually bound to do the right thing, there is always a risk that they may later decide to do the wrong thing. I don't trust them not to take advantage of the loophole they left for themselves here.
If they want to call themselves Open Source, they need to follow the rules.
If you can pick the terms of the benchmark, you can make anything outperform anything else. According to the article:
Mac OS X Server, when coupled with a new Macintosh Server G3, is the fastest platform for running Apache for under U.S. $5,000 -- outperforming Linux, Solaris and Windows NT Server...Based on WebBench benchmark testing performed by ZD Labs on a Dell PowerEdge 2300 Pentium II 450 MHz running Red Hat Linux, and a Sun Microsystems Enterprise Ultra 10S Server 333 MHz running Solaris; and NetBench benchmark testing performed by Apple on a Dell PowerEdge 2300 Pentium II 450 MHz running Windows NT Server, and a 400 MHz Macintosh Server G3 running Mac OS X Server.
Correct me if I'm wrong, but it looks like they are comparing the WebBench benchmark on RedHat on a Pentium II to the NetBench becnchmark on MacOS X Server on a G3. Not only aren't they running on the same processor, but they aren't even the same benchmark!!!
MacOS X may or may not be faster than Linux in some sense, but this test proves nothing.
Actually, for a troll it was better than average. The responses had at least some thought put into them.
And no, I don't consider myself a sucker for replying to a Troll. My posts are more to get my viewpoint out there, either directly or in response to an opposing viewpoint. The troll's viewpoint here was one that is common enough that I thought it should be rebutted regardless of whether or not it was the person's real opinion. It all goes to show that, even if you think you're playing petty games, there's a valid place for some trolls.
Look. Red Hat is a threat. Pure and simple. Sure, Gnome is great. Thanks. But right now Red Hat is VERY CLOSE to becoming the new MS of Linux. Look at what IBM did-- it didn't want to settle on one company and crown it "CZAR of the PCs" like it did with MS in the 1980s.
This is FUD pure and simple. Saying "RedHat is a threat" over and over does not make it so. If it is a threat, tell us how it is a threat. Vague references to IBM granting the PC-DOS contract to Microsoft in the early eighties doesn't cut it. Not only is IBM's investment in RedHat completely different from granting a monopoly contract, but IBM is a completely different company now than it was then. If RedHat is a threat than you should be able to come up with a plausible chain of events from the current situation to a bad one. RedHat has bent over backwards to make sure there is no such chain.
As far as I'm concerned, it's Debian all the way, until Debian gets too corporate. And then it's off to some other model.
More FUD. How can Debian possibly get "too corporate"?!!? They're a non-profit organization based on volunteers with almost no cash flow.
They're welcome to contribute to the free source initiative all they want. (Hey, Red Hat, FUCKING JOIN IN ON STANDARDS, why doncha?) But I don't trust them just because they're kewl and sold Linux back in the old days.
What standards aren't they joining in on that they should? Again, vague accusations just make for FUD. They have a delegate to the LSB, their distribution is POSIX compliant, they ship an ICCCM compliant X Server, they follow the FHS better than most distributions. What do you want from them, Unix trademark compliance?
Stick to core principles: Trust no one but yourselves.
Finally, some wisdom. I'm certainly not going to trust an anonymous coward with anger but no information.
It is so a word: irregardless == regardless, just like inflamable == flammable and English == confusing. Just because English has its troublesome bits doesn't mean you can unilaterally label them as wrong.
A program that contains no derivative of any portion of the Library, but is designed to work with the Library by being compiled or linked with it, is called a "work that uses the Library". Such a work, in isolation, is not a derivative work of the Library, and therefore falls outside the scope of this License...
This wording was designed so that proprietary or non-GPL-but-still-Free software could link statically to an LGPL library by distributing the object modules. An executable file already statically linked would not qualify for this term, and falls under Section 6.
My understanding is that a dynamic executable has not yet been linked by the terms of the license, but instead is linked at runtime (with ld.so or some other mechanism), so the dynamic executable still qualifies as a "work that uses the Library" per this section, and so it can be distributed directly without object modules.
All this assumes that the object module or dynamic executable doesn't still become a "derived work" by the vague explanation later in section 5.
The wording was carefully arranged so that you can update or modify the LGPL part of the work, while keeping the API intact, and still be able to have a working executable. Dynamic executables allow this without including the object modules.
Yes, Microstation is primarily Unix based. Unfortunately they limit themselves to the commercial Unixes. The only CAD program they offer for Linux is the crippleware Student Edition of Microstation. I have no idea how well or poorly it works.
While the original developer retains the right to use his own code in closed source software, I do not believe that he(or she) may use submitted code in that software--at least not under the GPL license.
Technically, you're right, but to a certain extent it depends on the project. If a person submits code to a software project that is distributed under a particular license, it is traditionally assumed that the submission is also under that license (unless explicitly otherwise). The only person who can change the license on code is the copyright holder. Small patches contributed to a larger work are often considered as being copyright assigned to the holder of the large work by default. Most big projects which care about copyright (eg. GNU or CygWin) ask for a formal assignment of copyright when you submit a significant amount of code. When you assign your copyrights to someone else, they can legally change the license at will.
Sendmail, of course, is not(to my knowledge) covered under GPL, so that probably explains why its makers can use publically submitted patches in a private product.
Until recently, Sendmail was under the BSD License. This license has always been interpreted that you can redistribute BSD source or binaries under any terms (i.e. license) you wish, provided the specific conditions of the BSD license are also met. As of version 8.9, Sendmail, Inc is supporting three versions of Sendmail, each under different license, each license includes the BSD conditions.
An interesting contrast can be drawn with closed source software, which increasingly is including time limits on usage in the fine print. While you can never lose the right to use OSS software, certain popular programs are legally limited to only twenty five to thirty years of usage.
I'm a radical spritualist mystic, actually. And I don't think John Katz talks to much. The majority of/.ers don't seem to think he talks too much. It's just a small group of Anonymous Cowards who talk too much about how they hate him.
PS: Slashdot poll question idea! I consider myself: * Atheist * Agnostic * Deist * Catholic * Christian (other than Catholic) * Muslim * Jewish * Buddist * Hindi * Wiccan * Other * Huh?
I suspect agnostic will be most common, but that's just a wild guess.
That's a problem with the press (particularly the US press), not RedHat. Every release from RedHat I've seen has been pretty clear that they are one of many distributions of Linux. The US media is scarily corporate-centric.
Yes, cars were invented before then, but cars as we know it really date back to the 1920's, with a few innovations tacked on in the 50's and 60's.
What's the point of griping that something is based on old technology? Old means that it has time to evolve, mature, and be tested. Linux is certainly many times more robust and powerful than the Unix of the 60's, just like even the cheap cars today are many times more safe and powerful than the cars of the 20's.
There has been competition in PC and peripherals since the mid-80's, only because there was a de facto standard platform on which to compare and compete. In the pre-DOS-ubiquity days, competition was less robust.
True, and Microsoft deserves some of the credit for that standardization. IBM deserves more credit for it, however. Also, one of the reasons the market was so standardized wasn't because the vendors wanted to, but the customers demanded it. In the 80's, my father bought a Mindset PC. It was an 8086 machine with souped up graphics and a snazzy design (they had one on display at the Museum of Modern Art). In spite of the fact that it was better than pretty much everything out there, it flopped. Why? They didn't use a normal ISA bus.
IT WAS VERY HARD TO ENTER THE DISPLAY CARD OR PRINTER MARKETS. Windows made that 1000% easier and made hardware competition much more intense.
No, it was easy to enter the markets, it was very hard to add a new standard. Windows 3.x allowed for hardware manufacturers to deviate from standards with abandon, but most didn't since they'd lose the non-windows market. Windows 95 removed that barrier, so we now have printers that are so non-standard that they only work with certain versions of Windows 95 and nothing else (not NT, not DOS, nothing).
Another example. CDROMs are ubiquitous and nearly free. You take them for granted, you stick em in the back of books. Who organized and sponsored all of the early CDROM conferences and drove standards in the PC space?
Um, Sony and Phillips if I recall.
I'm not saying that Microsoft didn't play a part in the whole computer business, and in the price reductions we have seen. I am saying that the original poster's argument that if it weren't for Windows 95 we wouldn't have Linux is ludicrous.
I'd also like to point out that while Microsoft has always encouraged strong growth and lower prices in the computer industry as a whole, they haven't followed suit. The Microsoft tax on a new Compaq machine ($75 for Windows 95) is more than the Microsoft tax on an old AT-clone was ($69 for MS-DOS if I recall). Yes, Windows 95 does more than MS-DOS 2.0, but everyone else in the business is charging less for things that do more.
Microsoft did reduce the cost of computing - there is no doubt about it. Linux would not exist if Microsoft hadn't help drive down the prices of x86 hardware.
They produced crappy products then and now, and have used snakelike tactics to increase and defend their market share, but there is no doubt that the millions of people buying win95 did in fact make the x86 platform cheap and viable.
First off, side effects of people buying Windows 95 had absolutely no effect on the creation of Linux. Linux was started in 1992, before Windows 95 was even hype. Linux was viable in 1993 or 1994 (depending on how you count viable).
Secondly, the price drops in the early nineties was due to a long price war amongst the various PC manufacturers. By most accounts, Compaq ended up the winner of the price war (they've been the biggest PC manufacturer since I think 94), but they did this with no help from Microsoft. Compaq had so much trouble with Microsoft products that they maintained their own version of DOS for many years.
But regardless, the fact that the x86 platform is so cheap is because of companies like Compaq, ALR, AMD, Hewlett Packard, etc trying to outpower and undercut each other while trying to keep survivable profit margins and fend off the Korean no-name clones. Since Microsoft products were pretty much a fixed cost, they didn't help or hurt one whit.
Slashdot Mirror
For what it's worth, here's my take on the issue. It looks to me like the 2.2(c) reporting requirement issue is annoying, but it won't lead to the license disappearing. They've got that one satisfactorily covered.
The 9.1 point is still a problem. I don't care what Apple intended Affected Original Code to mean. It is not clearly defined in the license, and I can see all sorts of abuse that can come from a loose definition of the term.
I don't care whether it is technically Open Source or not, I don't want a license which can end my legal use of a program on a company's whim. I think there is a clear need for a "No Termination of Use" clause to be inserted into the Open Source guidelines. While there is not one, the Open Source definition becomes pretty useless to me.
But it was!
/1246212.shtml
http://slashdot.org/books/99/01/31
Section 7 does not supercede the termination clause in Section 9.1. All it says is you get to pick which version of the APSL you use and distribute under.
On your other point, if you have a GPL'ed program that gets into a patent lawsuit, you lose the right to redistribute under the GPL. You don't lose the right to use it under the GPL, which means that you can wait for the patent to expire, or you can negotiate use of the patent directly with the patent holder.
If an APSL program gets into a patent lawsuit, Apple can pull your rights to use the software regardless of whether the patent later expires or if you negotiate your own patent license.
As for the APSL only terminating Affected Code, the license does not clearly define "Affected Original Code", basically the effect of this is it can mean whatever their lawyers say it means, and if you think it means otherwise, be prepared to pay for your own lawyers. If there is a lawsuit on a couple of lines in a library, they can easily say the "Affected Original Code" is the entire library, and anything which uses it. This is why Bruce asked that the APSL at minimum clarify terms like this.
As for you update and everything is hunky dory, that assumes that 1) Apple releases an update, and 2) Apple releases the update under the APSL. If Apple invokes this clause, third parties will not be legally allowed to fix the problem. Such terms are against both the spirit of Free Software and the legal definition of Open Source. We don't have to stand for it, and we shouldn't.
This
Mr P wrote:
:-)
Let's see his innovations:...the Internet
That's just not true! Everyone knows that Al Gore made the internet personally.
As far as what has he done personally, I think he did a little of the coding work in MS-BASIC (the program that started Microsoft). AFAIK, most of the first version of MS-BASIC was done by Paul Allen and then the employees that they hired after they sold some early copies.
As far as what has his company developed themselves, while Windows (I'm talking versions 1 and 2 here) was mostly other peoples ideas, I think it was coded by Microsoft. For Windows 3, 95 and 98 they contracted out a lot of it.
Windows NT 3.5 was just OS/2 and VMS run together really fast with a Windows 3 GUI put on it, in fact the developers, and much of the code, were grabbed from the OS/2 design team and Digital.
TrueType was a purchase from Apple (they developed it and were discarding it in favor of Adobe Type1). I think Word, PowerPoint and Xenix were purchased, I don't know from who.
I think Microsoft Bob was written by them, I'm not sure. I also think the eariler versions of Masm and Microsoft C were written by them. Microsoft BASIC was written by them, and I think its decendants: BASICA, QBASIC and Visual Basic.
You said "the Internet he did legitimately add to his OS, even if it was around for ages before". They didn't write their internet stuff. Winsock (i.e. "the Internet") was a kludged port of the same BSD networking stuff used in most Unixes. The same thing with most of the "standard" internet tools, Telnet, FTP, Ping, Traceroute. And, as you've said, Internet Explorer was purchased from SpyGlass.
Wordstar did not become Word. Wordstar was still around when Word started to become popular. They responded by dumping most of their original code (which I hear was some of the most incredible hand-optimized assembly ever written), and making Wordstar2000 (in 1985) which flopped so badly that the company was never heard from again.
As far as I know, the only actual code that Bill Gates worked on was MS-BASIC, and even that one was mostly done by Paul Allen.
Or to any of the responses to this :-)
Section 7 of the GPL terminates your right to distribute a program with contested code. Section 9.1 of the APSL terminates your right to use a program with contested code. There's a very big difference between the two.
Just think, if anyone makes a claim against Apple, they have the option to, on a whim, invalidate all of their userbase. If they feel they tire of being Open Source, they can release a new version under a solidly closed license, and convince someone to make a claim against the open version. Bingo, everyone goes from having a nice almost open-source system, to having to pay for a proprietary upgrade just to use their machine legally. All derivative works would likewise be killed.
This is not nitpicking, this is a clause that keeps the software from being Free. While I like Apple, they are a corporation. That means that their first responsibility is to their shareholders. That means if they are not contractually bound to do the right thing, there is always a risk that they may later decide to do the wrong thing. I don't trust them not to take advantage of the loophole they left for themselves here.
If they want to call themselves Open Source, they need to follow the rules.
If you can pick the terms of the benchmark, you can make anything outperform anything else. According to the article:
Mac OS X Server, when coupled with a new Macintosh Server G3, is the fastest platform for running Apache for under U.S. $5,000 -- outperforming Linux, Solaris and Windows NT Server...Based on WebBench benchmark testing performed by ZD Labs on a Dell PowerEdge 2300 Pentium II 450 MHz running Red Hat Linux, and a Sun Microsystems Enterprise Ultra 10S Server 333 MHz running Solaris; and NetBench benchmark testing performed by Apple on a Dell PowerEdge 2300 Pentium II 450 MHz running Windows NT Server, and a 400 MHz Macintosh Server G3 running Mac OS X Server.
Correct me if I'm wrong, but it looks like they are comparing the WebBench benchmark on RedHat on a Pentium II to the NetBench becnchmark on MacOS X Server on a G3. Not only aren't they running on the same processor, but they aren't even the same benchmark!!!
MacOS X may or may not be faster than Linux in some sense, but this test proves nothing.
One of the first personal webcams. Probably the first to be in a bedroom (well, a dormroom) 24/7. www.jennicam.org
Actually, for a troll it was better than average. The responses had at least some thought put into them.
And no, I don't consider myself a sucker for replying to a Troll. My posts are more to get my viewpoint out there, either directly or in response to an opposing viewpoint. The troll's viewpoint here was one that is common enough that I thought it should be rebutted regardless of whether or not it was the person's real opinion. It all goes to show that, even if you think you're playing petty games, there's a valid place for some trolls.
Check the header of your message, not the body.
Anonymous Coward (why am I not surprised) wrote:
Look. Red Hat is a threat. Pure and simple. Sure, Gnome is great. Thanks. But right now Red Hat is VERY CLOSE to becoming the new MS of Linux. Look at what IBM did-- it didn't want to settle on one company and crown it "CZAR of the PCs" like it did with MS in the 1980s.
This is FUD pure and simple. Saying "RedHat is a threat" over and over does not make it so. If it is a threat, tell us how it is a threat. Vague references to IBM granting the PC-DOS contract to Microsoft in the early eighties doesn't cut it. Not only is IBM's investment in RedHat completely different from granting a monopoly contract, but IBM is a completely different company now than it was then. If RedHat is a threat than you should be able to come up with a plausible chain of events from the current situation to a bad one. RedHat has bent over backwards to make sure there is no such chain.
As far as I'm concerned, it's Debian all the way, until Debian gets too corporate. And then it's off to some other model.
More FUD. How can Debian possibly get "too corporate"?!!? They're a non-profit organization based on volunteers with almost no cash flow.
They're welcome to contribute to the free source initiative all they want. (Hey, Red Hat, FUCKING JOIN IN ON STANDARDS, why doncha?) But I don't trust them just because they're kewl and sold Linux back in the old days.
What standards aren't they joining in on that they should? Again, vague accusations just make for FUD. They have a delegate to the LSB, their distribution is POSIX compliant, they ship an ICCCM compliant X Server, they follow the FHS better than most distributions. What do you want from them, Unix trademark compliance?
Stick to core principles: Trust no one but yourselves.
Finally, some wisdom. I'm certainly not going to trust an anonymous coward with anger but no information.
It is so a word: irregardless == regardless, just like inflamable == flammable and English == confusing. Just because English has its troublesome bits doesn't mean you can unilaterally label them as wrong.
To quote the LGPL:
A program that contains no derivative of any portion of the Library, but is designed to work with the Library by being compiled or linked with it, is called a "work that uses the Library". Such a work, in isolation, is not a derivative work of the Library, and therefore falls outside the scope of this License...
This wording was designed so that proprietary or non-GPL-but-still-Free software could link statically to an LGPL library by distributing the object modules. An executable file already statically linked would not qualify for this term, and falls under Section 6.
My understanding is that a dynamic executable has not yet been linked by the terms of the license, but instead is linked at runtime (with ld.so or some other mechanism), so the dynamic executable still qualifies as a "work that uses the Library" per this section, and so it can be distributed directly without object modules.
All this assumes that the object module or dynamic executable doesn't still become a "derived work" by the vague explanation later in section 5.
The wording was carefully arranged so that you can update or modify the LGPL part of the work, while keeping the API intact, and still be able to have a working executable. Dynamic executables allow this without including the object modules.
Yes, Microstation is primarily Unix based. Unfortunately they limit themselves to the commercial Unixes. The only CAD program they offer for Linux is the crippleware Student Edition of Microstation. I have no idea how well or poorly it works.
jnik wrote:
Federico works at RHAD, not Miguel (IIRC).
Miguel works as a Network Administrator at UNAM, a huge University in Mexico City.
Federico was hired by RHAD, but only after he had done huge amounts work on GNOME on his own time already.
Effugas wrote:
While the original developer retains the right to use his own code in closed source software, I do not believe that he(or she) may use submitted code in that software--at least not under the GPL license.
Technically, you're right, but to a certain extent it depends on the project. If a person submits code to a software project that is distributed under a particular license, it is traditionally assumed that the submission is also under that license (unless explicitly otherwise). The only person who can change the license on code is the copyright holder. Small patches contributed to a larger work are often considered as being copyright assigned to the holder of the large work by default. Most big projects which care about copyright (eg. GNU or CygWin) ask for a formal assignment of copyright when you submit a significant amount of code. When you assign your copyrights to someone else, they can legally change the license at will.
Sendmail, of course, is not(to my knowledge) covered under GPL, so that probably explains why its makers can use publically submitted patches in a private product.
Until recently, Sendmail was under the BSD License. This license has always been interpreted that you can redistribute BSD source or binaries under any terms (i.e. license) you wish, provided the specific conditions of the BSD license are also met. As of version 8.9, Sendmail, Inc is supporting three versions of Sendmail, each under different license, each license includes the BSD conditions.
An interesting contrast can be drawn with closed source software, which increasingly is including time limits on usage in the fine print. While you can never lose the right to use OSS software, certain popular programs are legally limited to only twenty five to thirty years of usage.
Yeek! Do you have an example of such a license?
I'm a radical spritualist mystic, actually. And I don't think John Katz talks to much. The majority of /.ers don't seem to think he talks too much. It's just a small group of Anonymous Cowards who talk too much about how they hate him.
PS: Slashdot poll question idea!
I consider myself:
* Atheist
* Agnostic
* Deist
* Catholic
* Christian (other than Catholic)
* Muslim
* Jewish
* Buddist
* Hindi
* Wiccan
* Other
* Huh?
I suspect agnostic will be most common, but that's just a wild guess.
That's a problem with the press (particularly the US press), not RedHat. Every release from RedHat I've seen has been pretty clear that they are one of many distributions of Linux. The US media is scarily corporate-centric.
Yes, cars were invented before then, but cars as we know it really date back to the 1920's, with a few innovations tacked on in the 50's and 60's.
What's the point of griping that something is based on old technology? Old means that it has time to evolve, mature, and be tested. Linux is certainly many times more robust and powerful than the Unix of the 60's, just like even the cheap cars today are many times more safe and powerful than the cars of the 20's.
Jobs is too hung up on the Mach kernel to adopt a Linux kernel. Thankfully, Mach is good too :-)
Anonymous Coward wrote:
There has been competition in PC and peripherals since the mid-80's, only because there was a de facto standard platform on which to compare and compete. In the pre-DOS-ubiquity days, competition was less robust.
True, and Microsoft deserves some of the credit for that standardization. IBM deserves more credit for it, however. Also, one of the reasons the market was so standardized wasn't because the vendors wanted to, but the customers demanded it. In the 80's, my father bought a Mindset PC. It was an 8086 machine with souped up graphics and a snazzy design (they had one on display at the Museum of Modern Art). In spite of the fact that it was better than pretty much everything out there, it flopped. Why? They didn't use a normal ISA bus.
IT WAS VERY HARD TO ENTER THE DISPLAY CARD OR PRINTER MARKETS. Windows made that 1000% easier and made hardware competition much more intense.
No, it was easy to enter the markets, it was very hard to add a new standard. Windows 3.x allowed for hardware manufacturers to deviate from standards with abandon, but most didn't since they'd lose the non-windows market. Windows 95 removed that barrier, so we now have printers that are so non-standard that they only work with certain versions of Windows 95 and nothing else (not NT, not DOS, nothing).
Another example. CDROMs are ubiquitous and nearly free. You take them for granted, you stick em in the back of books. Who organized and sponsored all of the early CDROM conferences and drove standards in the PC space?
Um, Sony and Phillips if I recall.
I'm not saying that Microsoft didn't play a part in the whole computer business, and in the price reductions we have seen. I am saying that the original poster's argument that if it weren't for Windows 95 we wouldn't have Linux is ludicrous.
I'd also like to point out that while Microsoft has always encouraged strong growth and lower prices in the computer industry as a whole, they haven't followed suit. The Microsoft tax on a new Compaq machine ($75 for Windows 95) is more than the Microsoft tax on an old AT-clone was ($69 for MS-DOS if I recall). Yes, Windows 95 does more than MS-DOS 2.0, but everyone else in the business is charging less for things that do more.
Cassius wrote:
Microsoft did reduce the cost of computing - there is no doubt about it. Linux would not exist if Microsoft hadn't help drive down the prices of x86 hardware.
They produced crappy products then and now, and have used snakelike tactics to increase and defend their market share, but there is no doubt that the millions of people buying win95 did in fact make the x86 platform cheap and viable.
First off, side effects of people buying Windows 95 had absolutely no effect on the creation of Linux. Linux was started in 1992, before Windows 95 was even hype. Linux was viable in 1993 or 1994 (depending on how you count viable).
Secondly, the price drops in the early nineties was due to a long price war amongst the various PC manufacturers. By most accounts, Compaq ended up the winner of the price war (they've been the biggest PC manufacturer since I think 94), but they did this with no help from Microsoft. Compaq had so much trouble with Microsoft products that they maintained their own version of DOS for many years.
But regardless, the fact that the x86 platform is so cheap is because of companies like Compaq, ALR, AMD, Hewlett Packard, etc trying to outpower and undercut each other while trying to keep survivable profit margins and fend off the Korean no-name clones. Since Microsoft products were pretty much a fixed cost, they didn't help or hurt one whit.