Nothing's wrong with subsidizing as a general rule. The problem is when you use profits earned from monopoly power (e.g., is Windows 98 *still* worth its $90 shelf price today?!?) and apply those profits to undermining competitors in another market using anti-competitive means (e.g., exclusionary deals with ISPs in Netscape's case, mandatory Windows/Office bundling in WP/Lotus's case).
Capitalism only works when there is *fair* competition, and monopolies can easily undermine any notion of fairness.
Small problem...what happens if someone does not install the source code or development tools on their box, and runs only the binaries provided for them in the distribution. Is the user still liable?
You seem to have a rosier recollection of the "facts" than I do, and you still haven't countered any of my points.
Re: WordPerfect. The DOS version wasn't bad. It took them too long to get onto Windows, but it should be noted that they were already unprofitable due to marketshare and price erosion.
Re: Lotus. Lotus halted development on their products because it wasn't profitable for them to continue development due to marketshare erosion. IBM bought them in the mid-90s largely because they wanted to bring Notes to all of IBM's server lines (especially AS/400).
Re: Netscape. And did Netscape 4.5 come out before or after Microsoft decided to release Internet Explorer to the market for free, "integrate" it with Windows, and sign ISPs to exclusionary contracts in exchange for "preferred" placement on the Windows desktop?
As I said, Microsoft didn't win on the merits of their products. They won because of the way they did business. It's hard to out-innovate Microsoft when they can use their hefty profit margins on their entrenched products to subsidize developments in other areas. Only three or four of their seven business units are currently profitable, with the others being subsidized so that they can gain marketshare.
But hey, don't let the big picture get in the way of your selective recall.
Yes, the software industry in the early 90s WAS pathetic. Microsoft had already established dominance in the OS market and used that OS dominance to subsidize Office.
Had the rivals been able to sell their software and make a profit margin, they would have continued to out-innovate Microsoft. The sad truth is that Microsoft is very good at using dominance in one market to subsidize entry into another, undercutting those rivals until they can't afford to fund innovation (WordPerfect, Lotus, Netscape, were all tremendous innovators until Microsoft undercut their price, reducing their profit margins and their ability to keep up with R&D). "Better than the competition" is easy when the competition can't afford R&D.
"better than the competition" didn't happen until the competition had been driven into the ground, or do you forget predatory pricing, mandatory bundling agreements, exclusionary licensing arrangements, and so forth?
When Microsoft has competition, it is never "better than the competition" that wins marketshare for them. It's "good enough and bundled with Windows".
It's things like this that make me believe that the only proper way for this to shake out is for the FSF to sue SCO for GPL infringement and seek a permanent injunction barring them from distributing any code distributed under the GNU General Public License. That should really kick them in the nuts, because they'd have to build their own stuff (compiler, SMB server, among others) for a change.
Still Seems Like Deceptive Trade Practices
on
SCO Execs Dumping Stock
·
· Score: 2, Insightful
Their case will be shown to be demonstrably weak in a court of law, so their trying of the case in the media is akin to running up the share price on lies and deceit. Of course, unless someone can prove they knew how weak their case was, they're not going to get punished for it.
If your basis for thinking that SCO will win is "evidence" spoon fed to you by SCO, then I've got some swampland in Florida to sell you.
Here's a consideration: what happens if the code in question didn't originate in SCO's codebase? What if it originated in Linux?
Here's another consideration: what happens if the code in question originated BEFORE its integration into UNIX or the "derivatives"? (RCU is an example of this. It was published as a research paper before it was ever productized. Read Cringeley for more information.)
Here's another consideration: what happens if part of SCO's purchase contract from Novell states that Novell has the right to order SCO to take certain actions vis-a-vis the UNIX licensees (such as their June 12 order to stop SCO from terminating IBM's license).
Never mind the fact that IBM CANNOT actually move away from the GPL license. My understanding is that works covered under the GPL cannot be switched to any other license...which makes sense, since part of the GPL's premise is to ensure that GPL'd code can never be privatized.
On the other hand, RedHat and everyone else may well be saved because SCO refuses to identify the infringement without an NDA. A legally cumbersome NDA makes it unreasonable for the "offender" to view the evidence...and this was a calculated move on the part of SCO to make sure that no one who could fix the infringement could view the evidence. Refusing to disclose the infringement to the infringer should be viewed negatively by the court.
I tried all of the different filesystems, under different kernel versions. They either weren't soup yet, caused stability problems, or performance bottlenecks.
Perhaps I'm oversimplifying but to me it looks like grid computing is a almagamation of web services and clustering. Shared resources, directory knows which machines are capable of which jobs, yadda yadda yadda.
At first, I was wondering why a 3d graphics chip company would not support for the platform that is growing fastest among 3d animation studios (note recent news stories about Pixar, Dreamworks, etc., moving from Sun or Windows to Linux/x86). Then I went to the site, and found that it was true, but only to a certain extent. They haven't dropped support - they never officially supported it. They do, however, support developers wanting to write the Linux drivers.
Reading the agreements on SCO's site that were mentioned earlier, SCO may have the right of termination of IBM's contract, but they're blowing smoke when they say they can make customer copies of AIX invalid. The licensing language only provides for copies of System V and AIX in IBM's possession.
The patent application is for a system allowing advertisers to bid for advertising space available on a website and the plumbing to make that all work without human intervention. It is not for advertisements in and of themselves.
However, I really question the non-obviousness of the implementation. Any practitioner of the art would be able to easily create this...there's no value in its disclosure to the IT community, and so the value of the "invention" as an advancement of the arts and sciences is pretty well worthless.
If your judgement is based on your experience at Microsoft, then I'd say that's not a representative sample. Really, considering the length of time its taken Microsoft to get Windows to five 9's on intel hardware, I wouldn't put stock in any claims made about any group.
Although, to be fair, the freedom that America has always had means it's easier for people just in it for the money to go to school and call themselves programmers...they don't really have to prove anything. Certifications aren't much better, because people cram for those and then forget everything 15 minutes after they've passed the text.
I tend to avoid making generalizations about any group of programmers, except when the groups are competent/incompetent, talented/untalented, motivated by technology/motivated by money.
It is apparent that you didn't read it. It clearly puts U.S. productivity near or at the top for the 1990s, which is the most relevent decade for this discussion. If we're talking 50s-80s, then you have a point, but not for the 90s.
"With the quality of american programmers going down, is it really surprizing that companies turn to other countries to find qualified computer professionals?"
This is a statement that cannot be left unchallenged. Please provide proof. Moreover, I'd question the scope of this statement, as its presumption is that the quality of all american programmers is going down, meaning that no one in the ever-expanding tech unemployment pool is fit to hold an Indian programmer's jock strap, to borrow an expression.
Slashdot Mirror
Nothing's wrong with subsidizing as a general rule. The problem is when you use profits earned from monopoly power (e.g., is Windows 98 *still* worth its $90 shelf price today?!?) and apply those profits to undermining competitors in another market using anti-competitive means (e.g., exclusionary deals with ISPs in Netscape's case, mandatory Windows/Office bundling in WP/Lotus's case).
Capitalism only works when there is *fair* competition, and monopolies can easily undermine any notion of fairness.
Small problem...what happens if someone does not install the source code or development tools on their box, and runs only the binaries provided for them in the distribution. Is the user still liable?
You seem to have a rosier recollection of the "facts" than I do, and you still haven't countered any of my points.
Re: WordPerfect. The DOS version wasn't bad. It took them too long to get onto Windows, but it should be noted that they were already unprofitable due to marketshare and price erosion.
Re: Lotus. Lotus halted development on their products because it wasn't profitable for them to continue development due to marketshare erosion. IBM bought them in the mid-90s largely because they wanted to bring Notes to all of IBM's server lines (especially AS/400).
Re: Netscape. And did Netscape 4.5 come out before or after Microsoft decided to release Internet Explorer to the market for free, "integrate" it with Windows, and sign ISPs to exclusionary contracts in exchange for "preferred" placement on the Windows desktop?
As I said, Microsoft didn't win on the merits of their products. They won because of the way they did business. It's hard to out-innovate Microsoft when they can use their hefty profit margins on their entrenched products to subsidize developments in other areas. Only three or four of their seven business units are currently profitable, with the others being subsidized so that they can gain marketshare.
But hey, don't let the big picture get in the way of your selective recall.
Yes, the software industry in the early 90s WAS pathetic. Microsoft had already established dominance in the OS market and used that OS dominance to subsidize Office.
Had the rivals been able to sell their software and make a profit margin, they would have continued to out-innovate Microsoft. The sad truth is that Microsoft is very good at using dominance in one market to subsidize entry into another, undercutting those rivals until they can't afford to fund innovation (WordPerfect, Lotus, Netscape, were all tremendous innovators until Microsoft undercut their price, reducing their profit margins and their ability to keep up with R&D). "Better than the competition" is easy when the competition can't afford R&D.
AFAIK, SCO ships GCC 2.95
"better than the competition" didn't happen until the competition had been driven into the ground, or do you forget predatory pricing, mandatory bundling agreements, exclusionary licensing arrangements, and so forth?
When Microsoft has competition, it is never "better than the competition" that wins marketshare for them. It's "good enough and bundled with Windows".
It's things like this that make me believe that the only proper way for this to shake out is for the FSF to sue SCO for GPL infringement and seek a permanent injunction barring them from distributing any code distributed under the GNU General Public License. That should really kick them in the nuts, because they'd have to build their own stuff (compiler, SMB server, among others) for a change.
"good enough" is what put $40b in the bank
However, "good enough" is not equal to "right".
Their case will be shown to be demonstrably weak in a court of law, so their trying of the case in the media is akin to running up the share price on lies and deceit. Of course, unless someone can prove they knew how weak their case was, they're not going to get punished for it.
If your basis for thinking that SCO will win is "evidence" spoon fed to you by SCO, then I've got some swampland in Florida to sell you.
Here's a consideration: what happens if the code in question didn't originate in SCO's codebase? What if it originated in Linux?
Here's another consideration: what happens if the code in question originated BEFORE its integration into UNIX or the "derivatives"? (RCU is an example of this. It was published as a research paper before it was ever productized. Read Cringeley for more information.)
Here's another consideration: what happens if part of SCO's purchase contract from Novell states that Novell has the right to order SCO to take certain actions vis-a-vis the UNIX licensees (such as their June 12 order to stop SCO from terminating IBM's license).
Never mind the fact that IBM CANNOT actually move away from the GPL license. My understanding is that works covered under the GPL cannot be switched to any other license...which makes sense, since part of the GPL's premise is to ensure that GPL'd code can never be privatized.
Hello pot, this is kettle. You're black.
On the other hand, RedHat and everyone else may well be saved because SCO refuses to identify the infringement without an NDA. A legally cumbersome NDA makes it unreasonable for the "offender" to view the evidence...and this was a calculated move on the part of SCO to make sure that no one who could fix the infringement could view the evidence. Refusing to disclose the infringement to the infringer should be viewed negatively by the court.
I don't think it'd be too cool to see the "it's not a bug, it's a feature" patches that Redmond would come up with. :-)
I tried all of the different filesystems, under different kernel versions. They either weren't soup yet, caused stability problems, or performance bottlenecks.
I've been running ReiserFS for 36 months without issue. Can't say the same for JFS, XFS, ext2, or ext3.
Perhaps I'm oversimplifying but to me it looks like grid computing is a almagamation of web services and clustering. Shared resources, directory knows which machines are capable of which jobs, yadda yadda yadda.
At first, I was wondering why a 3d graphics chip company would not support for the platform that is growing fastest among 3d animation studios (note recent news stories about Pixar, Dreamworks, etc., moving from Sun or Windows to Linux/x86). Then I went to the site, and found that it was true, but only to a certain extent. They haven't dropped support - they never officially supported it. They do, however, support developers wanting to write the Linux drivers.
Reading the agreements on SCO's site that were mentioned earlier, SCO may have the right of termination of IBM's contract, but they're blowing smoke when they say they can make customer copies of AIX invalid. The licensing language only provides for copies of System V and AIX in IBM's possession.
The documents have been imaged for permanent storage, so the destruction of the paper records is not a big thing.
The patent application is for a system allowing advertisers to bid for advertising space available on a website and the plumbing to make that all work without human intervention. It is not for advertisements in and of themselves.
However, I really question the non-obviousness of the implementation. Any practitioner of the art would be able to easily create this...there's no value in its disclosure to the IT community, and so the value of the "invention" as an advancement of the arts and sciences is pretty well worthless.
If your judgement is based on your experience at Microsoft, then I'd say that's not a representative sample. Really, considering the length of time its taken Microsoft to get Windows to five 9's on intel hardware, I wouldn't put stock in any claims made about any group.
Although, to be fair, the freedom that America has always had means it's easier for people just in it for the money to go to school and call themselves programmers...they don't really have to prove anything. Certifications aren't much better, because people cram for those and then forget everything 15 minutes after they've passed the text.
I tend to avoid making generalizations about any group of programmers, except when the groups are competent/incompetent, talented/untalented, motivated by technology/motivated by money.
It is apparent that you didn't read it. It clearly puts U.S. productivity near or at the top for the 1990s, which is the most relevent decade for this discussion. If we're talking 50s-80s, then you have a point, but not for the 90s.
"With the quality of american programmers going down, is it really surprizing that companies turn to other countries to find qualified computer professionals?"
This is a statement that cannot be left unchallenged. Please provide proof. Moreover, I'd question the scope of this statement, as its presumption is that the quality of all american programmers is going down, meaning that no one in the ever-expanding tech unemployment pool is fit to hold an Indian programmer's jock strap, to borrow an expression.
Go read some studies yourself. What you say is true, but it's not true anymore:
f
http://www.bls.gov/opub/mlr/2002/06/art4full.pd