The population growth rate is slowing down, so your assumption that growth continues at its current rate is suspect. I don't think that Sterling's number is that unreasonable. If we don't find an AIDS treatment that can be massively deployed in the 3rd world soon, Sterling's number may be on the high side; AIDS is taking out a whole generation in Africa and parts of Asia.
Do you have any clue who James Fallows is? He's the editor of the Atlantic Monthly and a well-known author, with considerable influence in policy circles (meaning that policy wonks in government read his stuff and it has serious influence on foreign policy in the US). He's not some free-lancer who needs Microsoft contract money to avoid starvation.
There's no reason to believe that his motivations aren't what he says they are. Working writers don't have much choice about using Microsoft Word; in many cases their editors and clients require it (yes, they could use other programs if they have a reliable converter to and from Word format, but that's risky -- we're talking career damage if the conversion has flaws).
Be was doomed from the start. What were they thinking? If Linux didn't exist, they'd be just as dead (or you'd be saying that *BSD is wiping them out, except that you seem to like those guys). There is zero chance for a new company to overcome Microsoft on the desktop with the traditional proprietary model, not without a technological change similar to the switch from mainframes to minis or minis to PCs.
Quite a few people (thousands) are making a reasonable living building on GPLed software. Some are now quite wealthy. But even if that weren't so, Sun, HP, and Microsoft are denying us all access to their software for the purpose of building on it and making a reasonable living (yes, one can write applications for those platforms and sell them, but one can do the same with Linux).
GPLed software can even be built on by proprietary companies; you can't link to the application, but you can use it. GCC is the back end for the leading commerical hardware description language simulators, for example, and the companies in question manage to charge $10K per seat and sometimes more.
Your problem is that you think that the BSD subsidy is a natural state of affairs. Sometimes it's a good solution. But it is a subsidy, one someone has to pay for.
It is scary that the person who wrote this nonsense makes part of his living as a journalist, as it is wrong in so many ways.
Stallman acted according to deeply held principles, not a "tantrum". You and I might not agree with those principles, but it discredits you to use such terms. He fought against Symbolics in much the same manner that FreeBSD is currently fighting against Sun: putting out non-copylefted free software. Symbolics could, and did, use RMS's work. If RMS's effort to compete against Symbolics was wrong, then you should, for consistency, attack the FreeBSD people for competing against Sun.
RMS did not "lose the use of his hands" in this effort; he's had some RSI problems, but your claim that he lost the use of his hands in the early 80s is ridiculous.
Symbolics was never in a position to provide a significant source of income to MIT, and in no alternative universe could they be.
RMS developed the GPL because he found that non-copylefted freeware was not effective in competing against Symbolics and others who could take the free base software and add a few proprietary enhancements.
It is true that academic research labs require support from the outside world. But that's no reason why research institutions shouldn't use licenses like the GPL; after all, the GPL is much more generous than the other popular model, which permits non-commercial or research use only and demands licensing for other use, even in cases where the taxpayers footed the bill for the original research (e.g. NCSA's handing of the original Mosaic).
Finally, you misuse the term "commercial developer". All companies and individuals who develop software for profit are commercial developers, no matter what licensing terms they choose to use for their software. Many commercial developers find the GPL superior to a BSD license for use in the code that they decide to release as freeware (whether it represents some or all of their output), as it avoids providing a subsidy to the competition.
People who put out GPLed software are competing in the free market. There are some whiners who don't like this competition; they hope that either their competitors will only sell their software for a high price, or they will license it under terms that allow relatively unskilled programmers to make small enhancements and sell those enhancements.
I get paid from writing proprietary software. I don't cry and wail because I can't grab other people's software and then sell it.
Since no OS (even OpenBSD, as good as they are) is completely impervious to attack, your liability-based solution means everyone on the net has to buy hefty insurance, and the trial lawyers take 1/3 of the cash for every damage award. Sorry, it's the wrong approach.
And what about the Linux newbie with a DSL line and a static IP address? He downloads a distro and pushes the buttons, but the default is an insecure system. Who's liable? The distributor? (You can try to exempt the distributor and say that the newbie is responsible, but no jury's going to buy that -- and the law has to treat Microsoft and Linux vendors equally).
OK, Red Hat can afford it. But Debian has to disband. You've just killed them. The developers can work very hard to be sure they're secure, but can they bet their life savings on it?
There is one thing that should be mandated, possibly by agreement but if that fails, by law. If you operate an ISP and you and your customers are assigned a given segment of IP space, it's trivial to configure your routers so that packets that lie about where they came from (giving a source IP address not in your IP space) can't escape to the rest of the net. It's negligence not to do this. You can make the filtering even tighter, by filtering packets coming from customers (except where there are peering agreements or other arrangements) so they can't spoof the other customers. This kind of filtering is probably going to have to be a legal requirement (or a contractual requirement imposed by the backbone folks on their customers).
While the patent is probably bogus, your message is meaningless, because you include the description, and what matters for patent purposes are the claims.
Recent Linux versions also have a number of kernel options to help with some DoS attacks, and Linux and *BSD kernel developers have been learning from each other on this issue. Just the same, if a recent Linux kernel didn't hold up well in your tests, we should know. Which version did you test?
Good point. Unfortunately, the response of some organizations to the white hat who tries to focus attention on a security flaw is to try to get the white hat prosecuted as a cracker.
We in the Linux community have to pay more attention to our own security. We're going to start to see more and more folks with always-up DSL connections and static IP addresses. If the default configuration as shipped by Red Hat, or Corel, or whoever isn't damn near bulletproof, you know that the DoS freaks are going to own a lot of these boxes, simply because you can assume that there are a lot of people who won't apply security upgrades, who think "I don't need to care about security, nothing on this box matters".
On the contrary, any DSL-connected Unix clone is an attack vehicle, if captured.
It's not good enough to have some specialized Linux distributions that focus on security. The market leaders are the ones that really matter, because if you find a flaw in Red Hat you've found an exploit you can immediately use on thousands of machines.
Let's say that I'm a company that sells laptops, and I'm considering supporting Linux. I hear that Dell's going down this road, so I check in to Slashdot to assess the community reaction. What would I conclude?
For one, it appears that the community expects the work to be done for no money. They scream because the machine isn't cheaper with Linux on it, even though the amount of work to be done to get Linux to work decently with laptops is considerable, far more than with desktops, and support costs can be expected to be higher (simply because there are far fewer experts on Linux-laptop issues than for Windows-laptop issues).
Second, the community screams because support for all platforms isn't instantly available, even though many laptop components don't have Linux drivers present.
So, it would appear that neither profit nor good will is available by doing Linux support for laptops. The users don't want to pay for it, and they'll hate you anyway.
The Linux kernel, standing alone, does not constitute an OS. It's near-useless on its own. And historically, the term "OS" was never used in this way (except perhaps those that claimed that Windows 3.1 was not an OS because it runs on top of DOS).
At minimum, the basic set of services, like the C library and those daemons that the system can't really live without (pretty much all required features that would appear in the LSB document) have to be considered part of the OS. Thus most of, say, the Debian base packages or the Red Hat required packages has to be included in what you call "OS". And generally speaking, these are not exactly the same for different distributions.
Because the author of the GPL says it's wrong
on
GPL for Books?
·
· Score: 2
The FSF does not use the GPL for books and documents, as it makes no sense. If you use the GPL for books, you put an unnecessary burden on publishers (they have to either include the "source code" -- the LaTeX or XML or whatever, as you say, with every copy, or they have to include a written offer to provide the source to every book buyer who asks for three years).
And the GPL is just wrong for some things. The GPL itself is not under the GPL. If it were, anyone could change it!
There are portions of books that the author will want to have included exactly. There are others where the book needs to be modifiable.
But the problem has already been solved. See, for example, the copyright statement on the GCC manual or the Emacs manual:
Copyright (C) 1988, 1989, 1992-1999, 2000 Free Software Foundation, Inc.
Permission is granted to make and distribute verbatim copies of this manual provided the copyright notice and this permission notice are preserved on all copies.
Permission is granted to copy and distribute modified versions of this manual under the conditions for verbatim copying, provided also that the sections entitled ``GNU General Public License'' and ``Funding for Free Software'' are included exactly as in the original, and provided that the entire resulting derived work is distributed under the terms of a permission notice identical to this one.
Permission is granted to copy and distribute translations of this manual into another language, under the above conditions for modified versions, except that the sections entitled ``GNU General Public License'' and ``Funding for Free Software'', and this permission notice, may be included in translations approved by the Free Software Foundation instead of in the original English.
Notice that this license is much simpler than the GPL and accomplishes the desired purpose.
Nobody builds processors that directly run the ix86 instruction set any more. Both Intel and AMD put out processors that execute RISC-like instructions at the core, and then surround it with a lot of machinery to translate the user level instructions into these operations on the fly. There's no reason why Transmeta's approach can't be competitive to that, especially given caching of translated instructions.
What we're seeing is the rebirth of microcode. This approach makes the most sense when executing a tightly coded instruction stream -- CISC instructions, or maybe bytecodes (there might be a performance advantage in doing JIT translation from Java bytecodes to VLIW instructions, but then again the combination of the two JIT translations, from bytecodes to ix86 and from ix86 to VLIW might be nearly as good and cheaper in the long run.
As the trademark holder, Linus is perfectly well allowed to leave legitimate and useful Linux sites alone, and only harrass those sites who, say, use a Linux name to attract viewers for something that has nothing to do with Linux (e.g. to get click-through fees from porn sites).
The only reason to shiver is if you trust sleazoid domain name squatters more than you trust Linus Torvalds, or if you have fantasies of making big bucks by registering a domain name and reselling it (sorry, the good names are all gone now).
Inktomi sells their technology to other companies; they don't operate a search engine under their own name. HotBot is Inktomi-based; there are others as well but I don't know who.
Google is one of the best search engines available for most purposes, because it ignores meta tags, and scores pages higher based on links to the site from other high-scoring pages (this is a recursive definition but the recursion bottoms out).
The result of this is that it gives useful results even when very common words are used. Try searching for Linux on Google. The first ten results are
linux.org
linux.com
www.debian.org
www.linuxworld.com
linux.davecentral.com
www.varesearch.com (VA Linux)
linux.corel.com
www.li.org (Linux International)
lwn.net (Linux Weekly News)
www.linuxhq.com
While a human being might be able to come up with a better list, a machine came up with that list, based solely on the structure of the web. (I wonder why linux.davecentral.com rates so high -- possibly because it's attached to a high-ranking site, davecentral.com).
The idea is that someone creates an organization, and then requires everyone to be in the organization as a condition for software distribution. Then the modified GPLed program is only distributed to club members, and all the club members agree to only distribute the program within the club. In a sense, the Trillian project (which is porting the GNU tools and Linux to the IA64 architecture, which is still under nondisclosure agreements) is such a club.
So, does the fact that this can be done break the GPL protections? No, because it doesn't get around the requirement to provide sources to everyone who gets binaries. Attempts to do this kind of thing for a different reason (e.g. charge everyone big bucks for being in the club and forbid them from sharing information with outsiders) may run afoul of antitrust provisions in the US and the EU (forcing people to be in a club before you do business with them may not be legal, depending on the circumstances).
RMS often points out that the GPL (and other licenses) shouldn't be written, or read, as if they represent the whole of the law. Just because the GPL doesn't exclude some possibility doesn't mean that it is legal. It may be illegal for another reason.
The press release brags about "Secure Computing's patented Type Enforcement technology". Clearly, to make this work they need to put their type enforcement stuff in the kernel. However, the GPL in Clause 7 specifically states
7. If, as a consequence of a court judgment or allegation of patent infringement or for any other reason (not limited to patent issues), conditions are imposed on you (whether by court order, agreement or otherwise) that contradict the conditions of this License, they do not excuse you from the conditions of this License. If you cannot distribute so as to satisfy simultaneously your obligations under this License and any other pertinent obligations, then as a consequence you may not distribute the Program at all. For example, if a patent license would not permit royalty-free redistribution of the Program by all those who receive copies directly or indirectly through you, then the only way you could satisfy both it and this License would be to refrain entirely from distribution of the Program.
This means that Secure Computing must grant a royalty-free license to all direct or indirect recipients to use their patented technologies in Linux kernels. Other clauses of the GPL forbid them from restricting redistribution. So are they giving up hope of making money on their patent? Do they know this?
The C library is at least as critical to the success of Linux as the kernel, but few people know the name of the person who's mainly responsible for its success, Ulrich Drepper.
Actually it is the GNU C++ compiler that is most improved; it is now dramatically better than it was in 1998.
Linus's warnings against egcs were not because of egcs bugs, but rather because of Linux bugs (the 2.0.x kernel doesn't do constraints on assembly language instructions properly, so when the optimizer was improved it broke code).
Slashdot Mirror
The population growth rate is slowing down, so your assumption that growth continues at its current rate is suspect. I don't think that Sterling's number is that unreasonable. If we don't find an AIDS treatment that can be massively deployed in the 3rd world soon, Sterling's number may be on the high side; AIDS is taking out a whole generation in Africa and parts of Asia.
What do you think the remark about intellectual property lawyers acting like Mafia dons was referring to?
Do you have any clue who James Fallows is? He's the editor of the Atlantic Monthly and a well-known author, with considerable influence in policy circles (meaning that policy wonks in government read his stuff and it has serious influence on foreign policy in the US). He's not some free-lancer who needs Microsoft contract money to avoid starvation.
There's no reason to believe that his motivations aren't what he says they are. Working writers don't have much choice about using Microsoft Word; in many cases their editors and clients require it (yes, they could use other programs if they have a reliable converter to and from Word format, but that's risky -- we're talking career damage if the conversion has flaws).
Be was doomed from the start. What were they thinking? If Linux didn't exist, they'd be just as dead (or you'd be saying that *BSD is wiping them out, except that you seem to like those guys). There is zero chance for a new company to overcome Microsoft on the desktop with the traditional proprietary model, not without a technological change similar to the switch from mainframes to minis or minis to PCs.
Quite a few people (thousands) are making a reasonable living building on GPLed software. Some are now quite wealthy. But even if that weren't so, Sun, HP, and Microsoft are denying us all access to their software for the purpose of building on it and making a reasonable living (yes, one can write applications for those platforms and sell them, but one can do the same with Linux).
GPLed software can even be built on by proprietary companies; you can't link to the application, but you can use it. GCC is the back end for the leading commerical hardware description language simulators, for example, and the companies in question manage to charge $10K per seat and sometimes more.
Your problem is that you think that the BSD subsidy is a natural state of affairs. Sometimes it's a good solution. But it is a subsidy, one someone has to pay for.
It is scary that the person who wrote this nonsense makes part of his living as a journalist, as it is wrong in so many ways.
Stallman acted according to deeply held principles, not a "tantrum". You and I might not agree with those principles, but it discredits you to use such terms. He fought against Symbolics in much the same manner that FreeBSD is currently fighting against Sun: putting out non-copylefted free software. Symbolics could, and did, use RMS's work. If RMS's effort to compete against Symbolics was wrong, then you should, for consistency, attack the FreeBSD people for competing against Sun.
RMS did not "lose the use of his hands" in this effort; he's had some RSI problems, but your claim that he lost the use of his hands in the early 80s is ridiculous.
Symbolics was never in a position to provide a significant source of income to MIT, and in no alternative universe could they be.
RMS developed the GPL because he found that non-copylefted freeware was not effective in competing against Symbolics and others who could take the free base software and add a few proprietary enhancements.
It is true that academic research labs require support from the outside world. But that's no reason why research institutions shouldn't use licenses like the GPL; after all, the GPL is much more generous than the other popular model, which permits non-commercial or research use only and demands licensing for other use, even in cases where the taxpayers footed the bill for the original research (e.g. NCSA's handing of the original Mosaic).
Finally, you misuse the term "commercial developer". All companies and individuals who develop software for profit are commercial developers, no matter what licensing terms they choose to use for their software. Many commercial developers find the GPL superior to a BSD license for use in the code that they decide to release as freeware (whether it represents some or all of their output), as it avoids providing a subsidy to the competition.
People who put out GPLed software are competing in the free market. There are some whiners who don't like this competition; they hope that either their competitors will only sell their software for a high price, or they will license it under terms that allow relatively unskilled programmers to make small enhancements and sell those enhancements.
I get paid from writing proprietary software. I don't cry and wail because I can't grab other people's software and then sell it.
Since no OS (even OpenBSD, as good as they are) is completely impervious to attack, your liability-based solution means everyone on the net has to buy hefty insurance, and the trial lawyers take 1/3 of the cash for every damage award. Sorry, it's the wrong approach.
And what about the Linux newbie with a DSL line and a static IP address? He downloads a distro and pushes the buttons, but the default is an insecure system. Who's liable? The distributor? (You can try to exempt the distributor and say that the newbie is responsible, but no jury's going to buy that -- and the law has to treat Microsoft and Linux vendors equally).
OK, Red Hat can afford it. But Debian has to disband. You've just killed them. The developers can work very hard to be sure they're secure, but can they bet their life savings on it?
There is one thing that should be mandated, possibly by agreement but if that fails, by law. If you operate an ISP and you and your customers are assigned a given segment of IP space, it's trivial to configure your routers so that packets that lie about where they came from (giving a source IP address not in your IP space) can't escape to the rest of the net. It's negligence not to do this. You can make the filtering even tighter, by filtering packets coming from customers (except where there are peering agreements or other arrangements) so they can't spoof the other customers. This kind of filtering is probably going to have to be a legal requirement (or a contractual requirement imposed by the backbone folks on their customers).
Doesn't have be a "Yahoo! competitor" -- it can be some lamer day trader with a short position on his ETrade account.
You write:
But is that really true? If every router refused to pass packets that clearly lie about their origin, IP spoofing would be a lot harder to do.
While the patent is probably bogus, your message is meaningless, because you include the description, and what matters for patent purposes are the claims.
Recent Linux versions also have a number of kernel options to help with some DoS attacks, and Linux and *BSD kernel developers have been learning from each other on this issue. Just the same, if a recent Linux kernel didn't hold up well in your tests, we should know. Which version did you test?
Good point. Unfortunately, the response of some organizations to the white hat who tries to focus attention on a security flaw is to try to get the white hat prosecuted as a cracker.
We in the Linux community have to pay more attention to our own security. We're going to start to see more and more folks with always-up DSL connections and static IP addresses. If the default configuration as shipped by Red Hat, or Corel, or whoever isn't damn near bulletproof, you know that the DoS freaks are going to own a lot of these boxes, simply because you can assume that there are a lot of people who won't apply security upgrades, who think "I don't need to care about security, nothing on this box matters".
On the contrary, any DSL-connected Unix clone is an attack vehicle, if captured.
It's not good enough to have some specialized Linux distributions that focus on security. The market leaders are the ones that really matter, because if you find a flaw in Red Hat you've found an exploit you can immediately use on thousands of machines.
Let's say that I'm a company that sells laptops, and I'm considering supporting Linux. I hear that Dell's going down this road, so I check in to Slashdot to assess the community reaction. What would I conclude?
For one, it appears that the community expects the work to be done for no money. They scream because the machine isn't cheaper with Linux on it, even though the amount of work to be done to get Linux to work decently with laptops is considerable, far more than with desktops, and support costs can be expected to be higher (simply because there are far fewer experts on Linux-laptop issues than for Windows-laptop issues).
Second, the community screams because support for all platforms isn't instantly available, even though many laptop components don't have Linux drivers present.
So, it would appear that neither profit nor good will is available by doing Linux support for laptops. The users don't want to pay for it, and they'll hate you anyway.
The Linux kernel, standing alone, does not constitute an OS. It's near-useless on its own. And historically, the term "OS" was never used in this way (except perhaps those that claimed that Windows 3.1 was not an OS because it runs on top of DOS).
At minimum, the basic set of services, like the C library and those daemons that the system can't really live without (pretty much all required features that would appear in the LSB document) have to be considered part of the OS. Thus most of, say, the Debian base packages or the Red Hat required packages has to be included in what you call "OS". And generally speaking, these are not exactly the same for different distributions.
The FSF does not use the GPL for books and documents, as it makes no sense. If you use the GPL for books, you put an unnecessary burden on publishers (they have to either include the "source code" -- the LaTeX or XML or whatever, as you say, with every copy, or they have to include a written offer to provide the source to every book buyer who asks for three years).
And the GPL is just wrong for some things. The GPL itself is not under the GPL. If it were, anyone could change it!
There are portions of books that the author will want to have included exactly. There are others where the book needs to be modifiable.
But the problem has already been solved. See, for example, the copyright statement on the GCC manual or the Emacs manual:
Notice that this license is much simpler than the GPL and accomplishes the desired purpose.
Nobody builds processors that directly run the ix86 instruction set any more. Both Intel and AMD put out processors that execute RISC-like instructions at the core, and then surround it with a lot of machinery to translate the user level instructions into these operations on the fly. There's no reason why Transmeta's approach can't be competitive to that, especially given caching of translated instructions.
What we're seeing is the rebirth of microcode. This approach makes the most sense when executing a tightly coded instruction stream -- CISC instructions, or maybe bytecodes (there might be a performance advantage in doing JIT translation from Java bytecodes to VLIW instructions, but then again the combination of the two JIT translations, from bytecodes to ix86 and from ix86 to VLIW might be nearly as good and cheaper in the long run.
As the trademark holder, Linus is perfectly well allowed to leave legitimate and useful Linux sites alone, and only harrass those sites who, say, use a Linux name to attract viewers for something that has nothing to do with Linux (e.g. to get click-through fees from porn sites).
The only reason to shiver is if you trust sleazoid domain name squatters more than you trust Linus Torvalds, or if you have fantasies of making big bucks by registering a domain name and reselling it (sorry, the good names are all gone now).
Hotbot uses Inktomi technology. They don't use Inktomi's database (I don't know who does).
Inktomi sells their technology to other companies; they don't operate a search engine under their own name. HotBot is Inktomi-based; there are others as well but I don't know who.
Google is one of the best search engines available for most purposes, because it ignores meta tags, and scores pages higher based on links to the site from other high-scoring pages (this is a recursive definition but the recursion bottoms out).
The result of this is that it gives useful results even when very common words are used. Try searching for Linux on Google. The first ten results are
While a human being might be able to come up with a better list, a machine came up with that list, based solely on the structure of the web. (I wonder why linux.davecentral.com rates so high -- possibly because it's attached to a high-ranking site, davecentral.com).
ObAdvocacy: and Google runs on Linux.
The idea is that someone creates an organization, and then requires everyone to be in the organization as a condition for software distribution. Then the modified GPLed program is only distributed to club members, and all the club members agree to only distribute the program within the club. In a sense, the Trillian project (which is porting the GNU tools and Linux to the IA64 architecture, which is still under nondisclosure agreements) is such a club.
So, does the fact that this can be done break the GPL protections? No, because it doesn't get around the requirement to provide sources to everyone who gets binaries. Attempts to do this kind of thing for a different reason (e.g. charge everyone big bucks for being in the club and forbid them from sharing information with outsiders) may run afoul of antitrust provisions in the US and the EU (forcing people to be in a club before you do business with them may not be legal, depending on the circumstances).
RMS often points out that the GPL (and other licenses) shouldn't be written, or read, as if they represent the whole of the law. Just because the GPL doesn't exclude some possibility doesn't mean that it is legal. It may be illegal for another reason.
The press release brags about "Secure Computing's patented Type Enforcement technology". Clearly, to make this work they need to put their type enforcement stuff in the kernel. However, the GPL in Clause 7 specifically states
This means that Secure Computing must grant a royalty-free license to all direct or indirect recipients to use their patented technologies in Linux kernels. Other clauses of the GPL forbid them from restricting redistribution. So are they giving up hope of making money on their patent? Do they know this?
The C library is at least as critical to the success of Linux as the kernel, but few people know the name of the person who's mainly responsible for its success, Ulrich Drepper.
Actually it is the GNU C++ compiler that is most improved; it is now dramatically better than it was in 1998.
Linus's warnings against egcs were not because of egcs bugs, but rather because of Linux bugs (the 2.0.x kernel doesn't do constraints on assembly language instructions properly, so when the optimizer was improved it broke code).