an implementation of the open ECMA-standard components of.NET, done without any knowledge of MS's implementation
The ECMA-standard components of.NET alone (for which MS has promised royalty-free licensing of any patents they may get) do not give a useful platform. Most C# programs use non-ECMA class library components in essential ways.
How much the developers knew about Microsoft's MS's implementation is possibly relevant if MS claims contract violation (e.g. violation of some EULA clause) or copyright violation, but it is irrelevant in the IMO much more likely case of an attack based on patent claims.
Even if the MS.Net API patent is granted, there are several reasons why it may be difficult or impossible for them to enforce it gainst Mono and DotGNU. For example, there is the matter of anti-trust law. Patent law says that when a patent is unenforcable because of anti-trust law, that makes the patent invalid. Of course Novell (for the Mono project) and the Free Software Foundation (for the DotGNU project) will have to prove this in court. Until this is done, Microsoft might be able to make hell hot for Novell, or Novell might feel forced to agree to some patent license from MS that makes Mono non-free. (Since DotGNU Portable.Net development is based outside the US where US patents have no legal force, DotGNU is much less vulnerable to bullying from MS.)
Actually, "show all of your code" is a good description of part of what RedHat is aking for:
20. A copy of the source code of each and every version of UNIX software in which SCO claims to own any intellectual property rights or have any other legal interest, and a copy of the source code of each and every version of Linux software in SCO's possession custody, or control that was created on or after January 1, 2000, whether or not released for public use, licensing or sale, and a copy of the source code of each and every version of UNIX software that SCO has sold, offered for sale, licensed, and offered to license.
21. The source code of any SCO Linux product and any third party Linux product identified in response to Interrogatory No. 6.
22. A copy of the source code for each version of the following: UnixWare, UnixWare 7, UnixWare 7.1.2, UnixWare 7.1.3, Open UNIX, Open UNIX 8, Reliant HA, NeTraverse Merge, Merge, Merge 5, SCO OpenServer, SCO OpenServer Release 5.0.5, SCO OpenServer Release 5.0.6, SCO OpenServer Release 5.0.6, SCO OpenServer Release 5.0.7, OpenServer Kernel Personality, OpenServer Kernel Personality for UnixWare 7.1.3, SCO Linux, SCO Linux 4.0 Server for the Itanium Processor Family, SCO Linux 4, OpenLinux, OpenLinux 3.1, OpenLinux 3.1.1, OpenLinux 64, UnitedLinux, UnitedLinux1.0, UnitedLinux 1.0 Service Pack 2, UnitedLinux 1.0 ServicePack 1.
No, it's not like the whole world has gone mad. Not every country has that "software inventions are patentable" bug in their legal systems (at least not yet).
I'd like to have a list of countries which do not have this problem. How would I go about finding / creating such a list? (tried google already, but possibly not with the best combination of serach terms)
Traditional brick-and-mortar retailers also have their eyes on lost money. They said they stand to lose money as shoppers turn to tax-free Internet purchases.
In a globalized economy, taxing e-commerce isn't going to work well. For the sake of fairness, traditional brick-and-mortar retailers shouldn't have to suffer from sales tax either. Time to move to a totally different tax system; I'd propose to tax energy consumption and nothing else (not even income tax); calibrate it so that the total tax burden remains unchanged.
Decreasing regulation is good whenever it can be done without undesireable side-effects such as e.g. increased risk of disasters (such as power failures, plane crashes, etc) or allowing profit-oriented companies to establish monopolies.
I agree that there are too many rules and regulations today; their total number should be reduced even as new technologies come up and may require some degree of regulation.
Why shouldn't VOIP providers be required by law to follow the same rules as traditional phone serve companies?
The problem is that these legislators want to apply a telephone operator licences thing that was designed for a now-past era when telephone companies had natural monopolies.
They should make some new rules which make sense for companies which make a buck from offering some kind of communication service via the internet, and then apply the same tule also to traditional telephone companies - there's no reason why companies which use some particular style of technology should be discriminated against.
Yes. There should be laws with some teeth to properly punish those who would intercept or re-route internet traffic which was not meant for them. Verisign comes to mind...
As long as only very few people used voice-over-IP, there was clearly no need to regulate. Any problems with voice-over-IP telephone would affect only the small percentage of people who use it, but not the economy as a whole.
When a technology becomes mainstream so that a big portion of the economy depends on it, or the privacy of a large part of the polulation depends on it, then it needs to be regulated.
You *need* larger scale, controlled tests to determine if a problem is *significant* before making any changes.
For many issues you don't need more than a single observation of user frustration - if you can see an obvious improvement, implement it.
For other issues, you will indeed want to observe multiple users. However, in practice there is rarely any benefit in wanting to observe more than five users concering any particular issue.
Are we supposed to assume that creative and logical are now mutually exclusive?
The human thought process is normally to some degree creative and to some degree logial.
For some tasks, too much logical thinking with too little creativity will give bad results.
The point here is that with UI design, it's different: Any significant amount of creativity in your UI design means that the UI is different from what prople are used to, and that will generally reduce usability. To create a decent UI, you need logical thinking, and lots of it.
That isn't true - at least for most programs it isn't true - most programs are written for people who are tring to have fun, or who are trying to accomplish some task. The assumption "users are idiots" is just as wrong as assuming the user's thought process to be in any way similar to how the programmer thinks.
The only way to create a program which is suitable for use by others is to watch others try to use the program. That, and nothing else, gives an understanding of how users think.
Thus, at any time, the value of the "intellectual property" contained within the open-source company is effectively zero.
Apart from the value of brands/trademarks, this is true. That however is not a reson for investors to avoid investing in open source / Free Software companies. Every investor who knows what he's doing will estimate the risks of the various investments (at least how big the worst-case loss is), and the investor will diversify in order to avoid as far as possible the possibility of unacceptably big losses.
If anything, this should make investing in open source / Free Software companies more attractive because for this type of businesses, estimating the worst-case loss is trivial - it's obviously the total investment.
The true problem with investing in open source / Free Software companies is that currently there is so little knowledge and experience on how to successfully run such a business that all such investments are high-risk but at the same time open source / Free Software business does not create any effective monopolies. Without the possibility of creating an effective monopoly, there is not the possibility of a very high ROI, and without this possibility high-risk investments don't make sense.
Let's get it right. This is not a 'free speech' issue. It is an corporate and scientific honesty issue. In fact, it was the employer excercising their rights to fire an employee for making statements they didn't like
I agree that @Stake's decision to fire Geer was not a 'free speech' issue. In fact I think the firing is understandable because
The particular truth which Geer had been pointing out is extremely dangerous for Microsoft's monopolistic strategies. Think about what will happen when this point is widely publicised and taken seriously by those who make purchase decisions for mission critical IT infrastructure of US government institutions. Not only would MS lose some significant revenue (as those institutions would have to make room in their IT budgets for buying stuff from a competitor of MS), but perhaps even more significantly they'd lose their current effective monopoly in one fell swoop.
The risk of losing MS as a customer is probably unacceptable to the VC shareholders of @Stake.
Therefore, this is just another example of venture capital funding corrupting a company.
However, there is a Free Speech issue here:
The real problem is that these events, together with earlier events in which MS abused their position of power to strike back at those who had the courage to speak up in the antitrust trials, will cause many people to think twice before they speak up about this truth. The article mentions the reluctance of many academics to get involved in the discussion for fear of losing funding.
The free speach issue is that MS has more power than any business should have, and they're abusing this power. It cannot be avoided that government has this power, and that's why there's a need for the
First Amendment which intends to prevent governments from abusing the power they have.
The First Amendment does not help if the dirty work of suppressing Free Speech is done by a company and not by government.
In this context I'd like to remind everyone that Microsoft apparantly
would have no hesitations to kill a Free Speech website like Slashdot if they believe that to be in their business interests, and they think they can get away with it.
How can mega-corporations like MS be stopped from suppressing Free Speech?
As they did with everyone else with whom they've "allied", the alliance will last exactly as long as it takes for Microsoft to get any proprietary information they want, then they'll kill the deal.
This particular business alliance isn't about MS getting information from the alliance partners, it's about both sides increasing their ability to sell their commercial products. I predict this alliance to be stable for the foreseeable future (until MS loses their dominating position in the PC OS market) for unrelated reasons.
That is assuming the toaster would still be http-reachable when it "breaks".
No, I wasn't assuming that. When the toaster doesn't respond at all, the program that does the diagnostics check on all toasters knows that it (or its network connection) is broken.
Well now we finally get to see how Microsoft treat Free software that they (think they) 'own'.
Is this Tron kernel thing really Free Software? Their FAQ doesn't sound like it...
Q14: What are the license types available for T-Kernel?
The T-Engine development kit includes T-Kernel and a license to execute it
on the kit. You will hence be able to execute T-Kernel by buying the kit.
If you would like to use T-Kernel on hardware other than the kit, such as
your final product, you will need to conclude a separate license agreement
with Personal Media Corporation, who is the developer of T-Kernel.
If this is the case, you can choose from the following licenses:
(1) Project license by product type (with source code)
This license is granted by the type of the product. You may apply for this
license, for example, when you produce a car navigation equipment in which
T-Kernel is embedded. You have only to pay the license fee once regardless
of the number of copies you produce as long as the copies are of the same
product type. With this license, the more copies you make, the less your unit
cost will be. However, you will need another license agreement for a
different product or another type of the licensed product. This license is
suitable for T-Kernel-embedded products to be manufactured on a large
scale.
(2) License by copy (without source code)
With this license, the license fee needs to be paid for each copy. In this case,
the source code is not given to you, and thereby you are required to run the
copy on hardware that is compatible with the T-Engine development kit.
This license is suitable for T-Kernel-embedded products of which many
types are produced but only a small number of copies are made for each type,
or for relatively large-sized T-Kernel-embedded products such as those used
for plant control.
Please contact Personal Media Corporation for more details about the
licensing details or licenses other than the above.
For any device with complicated enough functionality that it makes sense to put a microprocessor into it, it may actually be a useful feature for the device to be able to answer http requests with an XML document that provides a system status report. If you have 100 toasters or whatever, this will allow to check conveniently and efficiently which ones are broken.
Read the license that comes with the kernel. It is/not/ a vanilla GPL. If it were, proprietary kernel modules would be illegal.
The COPYING file in my kernel sources directory consists of a copy of the GNU GPL v2 plus some clarifications which are totally irrelevant to the issue of proprietary kernel modules. In particular, kernel modules are not "user programs that use kernel services by normal system calls".
Note that this does not make it illegal to use the kernel together with proprietary kernel modules - the GPL makes no restrictions whatsoever on how a program may be used. The GPL just grants rights to change and redistribute the kernel under certain conditions. According to how the FSF interprets the GPL, these conditions are not met if you distribute a copy of the kernel together with a proprietary kernel module. You can distribute the kernel if you don't distribute proprietary kernel modules along with it, or you can distribute proprietary kernel modules if you don't distribute a copy of the kernel along with them. End-users (but not shops which will re-sell PCs) are free to get the two components from two different sources and put them together themselves.
Did someone give you a copy of the kernel under a different license? If yes, please tell us the specifics.
yes, no shit, binary kernel modules are fine. That's not what this is.
According to how the FSF interprets the GPL, if you distribute a kernel module together with the kernel (for example by them both being installed on some hardware thing which you distribute) the terms of the GPL require you to make both the kernel and the kernel module available (with full source code) to the recipient of that thing which you distribute.
The only reason why the distinction between static and dynamic linking is relevant is that this interpretation of the GPL has not yet been tested in court. No-one who has seriously looked into the matter doubts that the copyleft of the GPL can be enforced when code is linked statically with GPL'd code. Whether the copyleft property of the GPL can also be enforced for dynamic linking is less clear.
My view is that dynamic linking does not reduce the legal power of the GPL's copyleft properties, but it makes it possible for software vendors to work around the copyleft, as follows: If you have an nVidia video card and install SuSE Linux, the installation program will offer you the option of installing nVidia's proprietary divers. If you accept this option, the non-free kernel module for that will be downloaded via the itnernet and installed. Now nVidia is not violating the GPL because they're distributing only their own code, and SuSE is not violating the GPL because they're not distributing the non-free code.
However, if you install Linux (the GPL'd kernel of the GNU/Linux system) and this driver on a PC, then you may not sell that PC, because in distributing the PC you would violate the GPL.
but the kernel has an exception to the license, allowing people to link dynamically to the kernel without being bound by the GPL
This is not true.
The kernel readme has a clarification that making system calls is considered "using" the kernel, not "linking" to it. There is no linking exception (for dynamic linking) like you describe.
nVidia's drivers are no ok to be distributed together with the kernel. They can still be distributed separately. The only reason why nVidia is not in violation of the GPL is that they're not distributing a copy of the kernel with the driver.
This is the core of the matter, and why google is so successful. We believe that they are unbiased, and therefore trust their results.
Google is great for finding webpages with worthwhile information related to some keywords.
It seems that the goal of A9 is to solve a closely related, but different problem: Finding worthwhile e-commerce sites related to some keyword. This is somnewhat related to Amazon's core business of empowering people to find and order the books they want, via the internet. So I think Amazon may have some insights on this problems and into which solution strategies are likely to work.
It will be interesting to see what strategy A9 will use for doing the "how worthwhile is this site likely to be" rating of e-commerce sites. If they come up with something that works as well as Google's link popularity concept works for informative websites, then A9 will certainly be successful in their chosen niche.
Slashdot Mirror
The ECMA-standard components of .NET alone (for which MS has promised royalty-free licensing of any patents they may get) do not give a useful platform. Most C# programs use non-ECMA class library components in essential ways.
How much the developers knew about Microsoft's MS's implementation is possibly relevant if MS claims contract violation (e.g. violation of some EULA clause) or copyright violation, but it is irrelevant in the IMO much more likely case of an attack based on patent claims.
Even if the MS .Net API patent is granted, there are several reasons why it may be difficult or impossible for them to enforce it gainst Mono and DotGNU. For example, there is the matter of anti-trust law. Patent law says that when a patent is unenforcable because of anti-trust law, that makes the patent invalid. Of course Novell (for the Mono project) and the Free Software Foundation (for the DotGNU project) will have to prove this in court. Until this is done, Microsoft might be able to make hell hot for Novell, or Novell might feel forced to agree to some patent license from MS that makes Mono non-free. (Since DotGNU Portable.Net development is based outside the US where US patents have no legal force, DotGNU is much less vulnerable to bullying from MS.)
Actually, "show all of your code" is a good description of part of what RedHat is aking for:
No, it's not like the whole world has gone mad. Not every country has that "software inventions are patentable" bug in their legal systems (at least not yet).
I'd like to have a list of countries which do not have this problem. How would I go about finding / creating such a list? (tried google already, but possibly not with the best combination of serach terms)
In a globalized economy, taxing e-commerce isn't going to work well. For the sake of fairness, traditional brick-and-mortar retailers shouldn't have to suffer from sales tax either. Time to move to a totally different tax system; I'd propose to tax energy consumption and nothing else (not even income tax); calibrate it so that the total tax burden remains unchanged.
Decreasing regulation is good whenever it can be done without undesireable side-effects such as e.g. increased risk of disasters (such as power failures, plane crashes, etc) or allowing profit-oriented companies to establish monopolies.
I agree that there are too many rules and regulations today; their total number should be reduced even as new technologies come up and may require some degree of regulation.
The problem is that these legislators want to apply a telephone operator licences thing that was designed for a now-past era when telephone companies had natural monopolies.
They should make some new rules which make sense for companies which make a buck from offering some kind of communication service via the internet, and then apply the same tule also to traditional telephone companies - there's no reason why companies which use some particular style of technology should be discriminated against.
Yes. There should be laws with some teeth to properly punish those who would intercept or re-route internet traffic which was not meant for them. Verisign comes to mind...
When a technology becomes mainstream so that a big portion of the economy depends on it, or the privacy of a large part of the polulation depends on it, then it needs to be regulated.
For many issues you don't need more than a single observation of user frustration - if you can see an obvious improvement, implement it.
For other issues, you will indeed want to observe multiple users. However, in practice there is rarely any benefit in wanting to observe more than five users concering any particular issue.
The human thought process is normally to some degree creative and to some degree logial.
For some tasks, too much logical thinking with too little creativity will give bad results.
The point here is that with UI design, it's different: Any significant amount of creativity in your UI design means that the UI is different from what prople are used to, and that will generally reduce usability. To create a decent UI, you need logical thinking, and lots of it.
That isn't true - at least for most programs it isn't true - most programs are written for people who are tring to have fun, or who are trying to accomplish some task. The assumption "users are idiots" is just as wrong as assuming the user's thought process to be in any way similar to how the programmer thinks.
The only way to create a program which is suitable for use by others is to watch others try to use the program. That, and nothing else, gives an understanding of how users think.
Apart from the value of brands/trademarks, this is true. That however is not a reson for investors to avoid investing in open source / Free Software companies. Every investor who knows what he's doing will estimate the risks of the various investments (at least how big the worst-case loss is), and the investor will diversify in order to avoid as far as possible the possibility of unacceptably big losses.
If anything, this should make investing in open source / Free Software companies more attractive because for this type of businesses, estimating the worst-case loss is trivial - it's obviously the total investment.
The true problem with investing in open source / Free Software companies is that currently there is so little knowledge and experience on how to successfully run such a business that all such investments are high-risk but at the same time open source / Free Software business does not create any effective monopolies. Without the possibility of creating an effective monopoly, there is not the possibility of a very high ROI, and without this possibility high-risk investments don't make sense.
I agree that @Stake's decision to fire Geer was not a 'free speech' issue. In fact I think the firing is understandable because
- The particular truth which Geer had been pointing out is extremely dangerous for Microsoft's monopolistic strategies. Think about what will happen when this point is widely publicised and taken seriously by those who make purchase decisions for mission critical IT infrastructure of US government institutions. Not only would MS lose some significant revenue (as those institutions would have to make room in their IT budgets for buying stuff from a competitor of MS), but perhaps even more significantly they'd lose their current effective monopoly in one fell swoop.
- The risk of losing MS as a customer is probably unacceptable to the VC shareholders of @Stake.
Therefore, this is just another example of venture capital funding corrupting a company.However, there is a Free Speech issue here:
The real problem is that these events, together with earlier events in which MS abused their position of power to strike back at those who had the courage to speak up in the antitrust trials, will cause many people to think twice before they speak up about this truth. The article mentions the reluctance of many academics to get involved in the discussion for fear of losing funding.
The free speach issue is that MS has more power than any business should have, and they're abusing this power. It cannot be avoided that government has this power, and that's why there's a need for the First Amendment which intends to prevent governments from abusing the power they have.
The First Amendment does not help if the dirty work of suppressing Free Speech is done by a company and not by government. In this context I'd like to remind everyone that Microsoft apparantly would have no hesitations to kill a Free Speech website like Slashdot if they believe that to be in their business interests, and they think they can get away with it.
How can mega-corporations like MS be stopped from suppressing Free Speech?
This particular business alliance isn't about MS getting information from the alliance partners, it's about both sides increasing their ability to sell their commercial products. I predict this alliance to be stable for the foreseeable future (until MS loses their dominating position in the PC OS market) for unrelated reasons.
No, I wasn't assuming that. When the toaster doesn't respond at all, the program that does the diagnostics check on all toasters knows that it (or its network connection) is broken.
Is this Tron kernel thing really Free Software? Their FAQ doesn't sound like it...
For any device with complicated enough functionality that it makes sense to put a microprocessor into it, it may actually be a useful feature for the device to be able to answer http requests with an XML document that provides a system status report. If you have 100 toasters or whatever, this will allow to check conveniently and efficiently which ones are broken.
The agreement will allow the Windows CE .NET platform for digital devices to work on top of T-Kernel.
The COPYING file in my kernel sources directory consists of a copy of the GNU GPL v2 plus some clarifications which are totally irrelevant to the issue of proprietary kernel modules. In particular, kernel modules are not "user programs that use kernel services by normal system calls".
Note that this does not make it illegal to use the kernel together with proprietary kernel modules - the GPL makes no restrictions whatsoever on how a program may be used. The GPL just grants rights to change and redistribute the kernel under certain conditions. According to how the FSF interprets the GPL, these conditions are not met if you distribute a copy of the kernel together with a proprietary kernel module. You can distribute the kernel if you don't distribute proprietary kernel modules along with it, or you can distribute proprietary kernel modules if you don't distribute a copy of the kernel along with them. End-users (but not shops which will re-sell PCs) are free to get the two components from two different sources and put them together themselves.
Did someone give you a copy of the kernel under a different license? If yes, please tell us the specifics.
According to how the FSF interprets the GPL, if you distribute a kernel module together with the kernel (for example by them both being installed on some hardware thing which you distribute) the terms of the GPL require you to make both the kernel and the kernel module available (with full source code) to the recipient of that thing which you distribute.
The only reason why the distinction between static and dynamic linking is relevant is that this interpretation of the GPL has not yet been tested in court. No-one who has seriously looked into the matter doubts that the copyleft of the GPL can be enforced when code is linked statically with GPL'd code. Whether the copyleft property of the GPL can also be enforced for dynamic linking is less clear.
My view is that dynamic linking does not reduce the legal power of the GPL's copyleft properties, but it makes it possible for software vendors to work around the copyleft, as follows: If you have an nVidia video card and install SuSE Linux, the installation program will offer you the option of installing nVidia's proprietary divers. If you accept this option, the non-free kernel module for that will be downloaded via the itnernet and installed. Now nVidia is not violating the GPL because they're distributing only their own code, and SuSE is not violating the GPL because they're not distributing the non-free code.
However, if you install Linux (the GPL'd kernel of the GNU/Linux system) and this driver on a PC, then you may not sell that PC, because in distributing the PC you would violate the GPL.
Could it be that MS might be using their market position to put pressure on hardware vendors to keep their specs and APIs locked up under NDAs?
This is not true.
The kernel readme has a clarification that making system calls is considered "using" the kernel, not "linking" to it. There is no linking exception (for dynamic linking) like you describe.
nVidia's drivers are no ok to be distributed together with the kernel. They can still be distributed separately. The only reason why nVidia is not in violation of the GPL is that they're not distributing a copy of the kernel with the driver.
why do you feel that Microsoft's patent filings are any better than Amazon's?
Google is great for finding webpages with worthwhile information related to some keywords.
It seems that the goal of A9 is to solve a closely related, but different problem: Finding worthwhile e-commerce sites related to some keyword. This is somnewhat related to Amazon's core business of empowering people to find and order the books they want, via the internet. So I think Amazon may have some insights on this problems and into which solution strategies are likely to work.
It will be interesting to see what strategy A9 will use for doing the "how worthwhile is this site likely to be" rating of e-commerce sites. If they come up with something that works as well as Google's link popularity concept works for informative websites, then A9 will certainly be successful in their chosen niche.