There's something to be said for being responsible for your own problems. When you outsource, you expect 100% reliability. When it fails, you are helpless. There is nothing you can do. When phone calls come streaming in, all you can say is "I don't know" when people ask when it will be up again. That makes you look bad.
If you are handling it yourself, you know what's going on. You are responsible if it fails. You're not twiddling your thumbs waiting for someone else to fix the problem. You can give estimates and tell your bosses about progress.
No, there is no requirement to download the code, but i was referring to what is feasible distribution of source code. I doubt many people are going to want to deal with the snail mail option of providing source when they are distributing the binary to millions of people by internet.
You know as well as anyone that software patents can (and often are) violated unintentionally. Plus, patents can show up years later when a patent is acquired by a merger or product purchase. Only a fool believes they can write non-trivial software without running afoul of some software patent. Thankfully, few software patent holders seem interested in litigating such things.
The point is, something like the OIN only protects other open source projects from their patents. If Microsoft agrees to never enforce their patents against an open source project, there is no agreement that those projects will never enforce patents against them. That's a rather lopsided bargain, don't you think?
My point is this. Microsoft is unlikely to ever use GPL code in their products (other than products they inherit such as SFU nee Interix nee OpenNT). So any patent claim that protects GPL will never apply to them. However, if Microsoft agrees to not enforce any patents against GPL, then any patent bomb added to the GPL could could then be used against Microsoft for their non-GPL software, and Microsoft would not be able to counter-sue because they've already agreed never to enfore their patents, thus they've lost the #1 advantage of a patent portfolio (using it to force a cross licensing deal).
Now, before you or anyone else jumps to conclusions, i'm not suggesting that Microsoft use GPL code without adhering to the license, I mean suppose they have code that is found to independantly violate a patent that has been pledged to open source code. Because of Microsoft's pledge to never enforce patents against open source, the open source entities who now own the code can sue Microsoft at will and Microsoft will be unable to defend themselves.
What part of "Now, certainly, you could get the code from one of my customers" didn't you understand?
Yes, they could, but there is no requirement for them to do so, just like the BSD license makes no requriement for them to do so unless they themselves give the binaries to someone.
Actually, it's misnamed. It should be called "Active Sites", not Active Servers. You can see this by looking at the table above it that is in fact labled "Active Sites" and notice the numbers are the same. The methodology does not count market share, it counts hostnames that have "unique content", ie are not parking or squatting sites. This is no more accurate than the full survey, it just makes an attempt to filter out parking.
You don't have to follow 3(b) if you follow 3(a). But you are not allowed to prevent any of the people you give source code under 3(a) from giving it to anyone they like.
Which is what I said further down when I said you couldn't prevent your customers from providing the source code.
While yes, IE was part of the plus pack, it was also available for free download from MS's FTP and Web site. Also, as of Windows 95A (the OSR1) it was included with the OS, that was about 6 months later. Didn't matter though, cause IE1 sucked rocks, as did IE2. IE3 was the first usable version, on a par with Netscape and IE4 started to leave NS in the dust.
Also, most people accuse Microsoft of deliberately trying to break standards, when that's not the case at all. The fact is, there *WERE* no decent standard, or standards bodies making the standards back then. The IETF punted, and the W3C was basically useless. Netscape and Microsoft were the only ones pushing HTML forward. IE4 was actually more standards compliant than Netscape was. IE5 and IE6 were more compliant than Mozilla was when they were released. But, Microsoft just let IE rot, and as Mozilla (then FF) got better, IE didn't, which many confused with "deliberately breaking the standards". IE just didn't mature to the newer understanding of the standards while Moz and FF did.
While that may be true, it indicates an inherant flaw in your reasoning. If domain parking sites are meaningless, then you're also admitting that Apache was propped up by domain parking sites as well, so that was also meaningless.
Looking at the Netcraft "active sites" stats excludes domain parking, and guess what? The numbers are about the same. Kind of blows your theory.
Apart from the fact that Netcraft's survey doesn't count market share, it only counts hostnames (which you know, because you've championed efforts to play that game, right down to lying about the fact that your parking servers ran lighttpd but claimed to be apache in order to game the stats), that's a 14% difference, which means that if IIS gains 7% of Apache's hostnames, they're even. 7% is not a huge margin and can change in a matter of months.
We all know you harbor a major grudge against Microsoft because you think they got you fired (and maybe they did), but you're being pretty unreasonable.
They can disavow it where we are concerned, because we're not aiming one at them.
That's not true. Many open source companies, like Red Hat, Novel, IBM, Sun, Oracle, etc.. have vast patent portfolios, and while some of these companies have agreed not to enforce those patents against open source, nothing prevents them from enforcing those patents against Microsoft. So let's say Microsoft agrees to never enforce any patent against Open Source software, what stops IBM from planting patent bombs in Open Source software to use against Microsoft, since there is no "open source entity" to promise that no open source patents will ever be used against corporations, it's a catch-22.
I don't think you've really thought this whole patent thing through, nor have you analyzed all the ways that corporations can get their ass handed to them by participating in such actions.
If you develop and don't get given money then the GPL "pays" you in the form of reciprocal freedom.
No, it doesn't. You really don't understand the GPL at all. There is no requirement that anyone give you their changes in return for you giving them your code. I am completely within my legal right to take your GPL'd code, make changes and not give them to you. Even if I distribute the code to my customers, I don't need to give you the changes. I need only give my customers the code (if they ask for it). I can deny you the code if you ask for it. There is nothing in the GPL which requires me to give my changes to random people who i have not given the binaries to.
Now, certainly, you could get the code from one of my customers, and that's a possibility, and I can't prevent them or you from that, but if you do that, you have to get the code from them, not me. I only have to provide it to the people I distributed the code to.
The point is, the GPL does not do what most people think it does. Most people completely misunderstand the rights granted by the GPL, and the restrictions required by it. For instsance, most people don't realize that if they take some GPL code, like a Fedora distro, make a few tiny changes, and then post the ISO then they are required to supply all upstream code. You can't just point them to Fedora, you are legally required to provide it yourself, including whatever bandwith costs that would encumber you with. The MEPIS developer lost a court battle on that one.
Nit pickers should not have their own nits to be picked.
The "Active Servers" graph is not what you think it is. Their methodology is that they count a unique site as an "active site", and somehow they put the word "servers" in there. Active sites is meant to filter out domain squating and domain parking sites where the pages are substantially similar to each other. It does not indicate a stand-alone server.
Although Netcraft does count unique servers (to the best of their ability) they don't publish this information for free, except for once a number of years ago. The public survey counts only hostnames, not servers.
No, this is not what the TPM chip has been marketed as. And no, it's not what Microsoft has marketed it as. The TPM chip is not used by the DRM system at all, and I have read no plans of using it as such. Not that it couldn't be used as such, but it's only the anti-DRM advocates who complain about this supposed connection that doesn't seem to exist.
I really blame the slashdot editors. This has come up often enough over the years that they should know this by now. But, they also know that patent arguments (valid or not) will generate tons of page views, so they seem willing to put anything like this on the front page without even considering it's validity or not.
This is so typical of slashdot patent troll arguments. You focus on one claim, ignoring the fact that individual claims are irrelevant. A patent must be taken *AS A WHOLE*. The claims are logically and'd, not or'd. All claims must be true for the patent to be enforceable. This patent does *NOT* patent general filesystem trees, that's merely one of the claims that must be true for the patent to be enforceable.
Patents are written from general to specific, with each claim becoming more and more specific. Thus you could see a patent that goes like this:
1. A computer system 2. Claim 1 where the computer system is green. 3. Claim 2 where a clown tap dances 4. Claim 3 where Elmer Fudd shoots himself in the face...
Now, the typical slashdot patent troll would scream "ZOMG, they've patented the color green!!!!!!1111"
"allows remote attackers to cause a denial of service (application crash) and possibly execute arbitrary code via crafted URLs that are not properly handled using certain rewrite rules."
That seems critical to me. Also, while a patch may appear in CVS/SVN within a week, it typically doesn't make it out to the distro users for several weeks afterwards. For instance, this flaw was published on July 6th, but it didn't make it to (for example) Debian until August 1st. http://www.debian.org/security/2006/dsa-1131
I don't really trust the way that apache categorizes their vulnerabilities as they list a DoS attack as critical, but a remote arbitrary code execution flaw as "important". So who knows.
I wasn't aware of the February vulnerability, but the 2006 vulnerability is not a flaw in IIS, it's a flaw in ASP which is exposed via IIS (because that's really the only way to expose it). It would be like a flaw in mod_perl or something being attributed to IIS. In addition, ASP is not enabled by default, so it's a less critical flaw.
The February vulnerability appears more serious, however, it's still mitigated by the fact that the attacker can only execute code as the worker process, which severely limits things in the default configuration. But still, only one critical vulnerability in 5 years, with that vulnerabilty being only a few months ago?
Uhh.. you really have no idea what you're talking about. First, MFC is a library, C# is a language. Second, C# was developed by Anders Hjelberg, who Microsoft hired away from Borland. He's the guy that basically wrote Delphi. And no, he did not create MFC.
Bigger question is, why isn't it possible to optimize Vista for SSDs? You speak as if Microsoft would have to redesign everything top to bottom in order to support them
The problem is related to the entire core Virtual Memory system in Windows. It assumes a different kind of subsystem than is provided by SSD's. The good news is that Microsoft has already done much of the work for their Embedded Windows project, but that work hasn't been done to Vista.
Windows (NT based) was designed to run on hardware that was a lower spec than it really needed, so it was designed to agressivly prune working sets, use page backed executables, and many other things which improve the performance if your memory and CPU are a bit low. Unfortunately, Microsoft has not changed this design in recent versions of Windows, even though other parts of the OS are far less stingy with resources. Microsoft really needs to retune the kernel to todays hardware.
And yes, IIS and ASP.NET are very easy to develop for. You can download the free Visual Studio Express, Free Sql Server Express, and have web pages up and running in seconds after installation. This gives you the ability to use a debugger to debug web pages and many other features that are a real pain in the ass to get working in the patchwork tools required for PHP and JSP or even RoR. No configuration needed.
Certainly, developing non-trivial web applications is just as difficult over the long haul, but the Microsoft tools make it VERY easy to get started.
Slashdot Mirror
There's something to be said for being responsible for your own problems. When you outsource, you expect 100% reliability. When it fails, you are helpless. There is nothing you can do. When phone calls come streaming in, all you can say is "I don't know" when people ask when it will be up again. That makes you look bad.
If you are handling it yourself, you know what's going on. You are responsible if it fails. You're not twiddling your thumbs waiting for someone else to fix the problem. You can give estimates and tell your bosses about progress.
No, there is no requirement to download the code, but i was referring to what is feasible distribution of source code. I doubt many people are going to want to deal with the snail mail option of providing source when they are distributing the binary to millions of people by internet.
Except that such a promise is not GPL compatible. The GPL requires that patents be licensed freely, under no conditions.
You know as well as anyone that software patents can (and often are) violated unintentionally. Plus, patents can show up years later when a patent is acquired by a merger or product purchase. Only a fool believes they can write non-trivial software without running afoul of some software patent. Thankfully, few software patent holders seem interested in litigating such things.
The point is, something like the OIN only protects other open source projects from their patents. If Microsoft agrees to never enforce their patents against an open source project, there is no agreement that those projects will never enforce patents against them. That's a rather lopsided bargain, don't you think?
You misunderstand.
My point is this. Microsoft is unlikely to ever use GPL code in their products (other than products they inherit such as SFU nee Interix nee OpenNT). So any patent claim that protects GPL will never apply to them. However, if Microsoft agrees to not enforce any patents against GPL, then any patent bomb added to the GPL could could then be used against Microsoft for their non-GPL software, and Microsoft would not be able to counter-sue because they've already agreed never to enfore their patents, thus they've lost the #1 advantage of a patent portfolio (using it to force a cross licensing deal).
Now, before you or anyone else jumps to conclusions, i'm not suggesting that Microsoft use GPL code without adhering to the license, I mean suppose they have code that is found to independantly violate a patent that has been pledged to open source code. Because of Microsoft's pledge to never enforce patents against open source, the open source entities who now own the code can sue Microsoft at will and Microsoft will be unable to defend themselves.
See the complexities?
What part of "Now, certainly, you could get the code from one of my customers" didn't you understand?
Yes, they could, but there is no requirement for them to do so, just like the BSD license makes no requriement for them to do so unless they themselves give the binaries to someone.
Actually, it's misnamed. It should be called "Active Sites", not Active Servers. You can see this by looking at the table above it that is in fact labled "Active Sites" and notice the numbers are the same. The methodology does not count market share, it counts hostnames that have "unique content", ie are not parking or squatting sites. This is no more accurate than the full survey, it just makes an attempt to filter out parking.
Their methodology is here: http://www.netcraft.com/survey/index-200007.html#active
You don't have to follow 3(b) if you follow 3(a). But you are not allowed to prevent any of the people you give source code under 3(a) from giving it to anyone they like.
Which is what I said further down when I said you couldn't prevent your customers from providing the source code.
Not exactly true.
While yes, IE was part of the plus pack, it was also available for free download from MS's FTP and Web site. Also, as of Windows 95A (the OSR1) it was included with the OS, that was about 6 months later. Didn't matter though, cause IE1 sucked rocks, as did IE2. IE3 was the first usable version, on a par with Netscape and IE4 started to leave NS in the dust.
Also, most people accuse Microsoft of deliberately trying to break standards, when that's not the case at all. The fact is, there *WERE* no decent standard, or standards bodies making the standards back then. The IETF punted, and the W3C was basically useless. Netscape and Microsoft were the only ones pushing HTML forward. IE4 was actually more standards compliant than Netscape was. IE5 and IE6 were more compliant than Mozilla was when they were released. But, Microsoft just let IE rot, and as Mozilla (then FF) got better, IE didn't, which many confused with "deliberately breaking the standards". IE just didn't mature to the newer understanding of the standards while Moz and FF did.
While that may be true, it indicates an inherant flaw in your reasoning. If domain parking sites are meaningless, then you're also admitting that Apache was propped up by domain parking sites as well, so that was also meaningless.
Looking at the Netcraft "active sites" stats excludes domain parking, and guess what? The numbers are about the same. Kind of blows your theory.
Apart from the fact that Netcraft's survey doesn't count market share, it only counts hostnames (which you know, because you've championed efforts to play that game, right down to lying about the fact that your parking servers ran lighttpd but claimed to be apache in order to game the stats), that's a 14% difference, which means that if IIS gains 7% of Apache's hostnames, they're even. 7% is not a huge margin and can change in a matter of months.
We all know you harbor a major grudge against Microsoft because you think they got you fired (and maybe they did), but you're being pretty unreasonable.
They can disavow it where we are concerned, because we're not aiming one at them.
That's not true. Many open source companies, like Red Hat, Novel, IBM, Sun, Oracle, etc.. have vast patent portfolios, and while some of these companies have agreed not to enforce those patents against open source, nothing prevents them from enforcing those patents against Microsoft. So let's say Microsoft agrees to never enforce any patent against Open Source software, what stops IBM from planting patent bombs in Open Source software to use against Microsoft, since there is no "open source entity" to promise that no open source patents will ever be used against corporations, it's a catch-22.
I don't think you've really thought this whole patent thing through, nor have you analyzed all the ways that corporations can get their ass handed to them by participating in such actions.
If you develop and don't get given money then the GPL "pays" you in the form of reciprocal freedom.
No, it doesn't. You really don't understand the GPL at all. There is no requirement that anyone give you their changes in return for you giving them your code. I am completely within my legal right to take your GPL'd code, make changes and not give them to you. Even if I distribute the code to my customers, I don't need to give you the changes. I need only give my customers the code (if they ask for it). I can deny you the code if you ask for it. There is nothing in the GPL which requires me to give my changes to random people who i have not given the binaries to.
Now, certainly, you could get the code from one of my customers, and that's a possibility, and I can't prevent them or you from that, but if you do that, you have to get the code from them, not me. I only have to provide it to the people I distributed the code to.
The point is, the GPL does not do what most people think it does. Most people completely misunderstand the rights granted by the GPL, and the restrictions required by it. For instsance, most people don't realize that if they take some GPL code, like a Fedora distro, make a few tiny changes, and then post the ISO then they are required to supply all upstream code. You can't just point them to Fedora, you are legally required to provide it yourself, including whatever bandwith costs that would encumber you with. The MEPIS developer lost a court battle on that one.
Nit pickers should not have their own nits to be picked.
The "Active Servers" graph is not what you think it is. Their methodology is that they count a unique site as an "active site", and somehow they put the word "servers" in there. Active sites is meant to filter out domain squating and domain parking sites where the pages are substantially similar to each other. It does not indicate a stand-alone server.
Although Netcraft does count unique servers (to the best of their ability) they don't publish this information for free, except for once a number of years ago. The public survey counts only hostnames, not servers.
No, this is not what the TPM chip has been marketed as. And no, it's not what Microsoft has marketed it as. The TPM chip is not used by the DRM system at all, and I have read no plans of using it as such. Not that it couldn't be used as such, but it's only the anti-DRM advocates who complain about this supposed connection that doesn't seem to exist.
I really blame the slashdot editors. This has come up often enough over the years that they should know this by now. But, they also know that patent arguments (valid or not) will generate tons of page views, so they seem willing to put anything like this on the front page without even considering it's validity or not.
This is so typical of slashdot patent troll arguments. You focus on one claim, ignoring the fact that individual claims are irrelevant. A patent must be taken *AS A WHOLE*. The claims are logically and'd, not or'd. All claims must be true for the patent to be enforceable. This patent does *NOT* patent general filesystem trees, that's merely one of the claims that must be true for the patent to be enforceable.
Patents are written from general to specific, with each claim becoming more and more specific. Thus you could see a patent that goes like this:
1. A computer system ...
2. Claim 1 where the computer system is green.
3. Claim 2 where a clown tap dances
4. Claim 3 where Elmer Fudd shoots himself in the face
Now, the typical slashdot patent troll would scream "ZOMG, they've patented the color green!!!!!!1111"
Uhh.. what?
http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2006-3747
"allows remote attackers to cause a denial of service (application crash) and possibly execute arbitrary code via crafted URLs that are not properly handled using certain rewrite rules."
That seems critical to me. Also, while a patch may appear in CVS/SVN within a week, it typically doesn't make it out to the distro users for several weeks afterwards. For instance, this flaw was published on July 6th, but it didn't make it to (for example) Debian until August 1st. http://www.debian.org/security/2006/dsa-1131
I don't really trust the way that apache categorizes their vulnerabilities as they list a DoS attack as critical, but a remote arbitrary code execution flaw as "important". So who knows.
ASP is not use that much anymore, ASP != ASP.NET. The flaw was in classic ASP.
I wasn't aware of the February vulnerability, but the 2006 vulnerability is not a flaw in IIS, it's a flaw in ASP which is exposed via IIS (because that's really the only way to expose it). It would be like a flaw in mod_perl or something being attributed to IIS. In addition, ASP is not enabled by default, so it's a less critical flaw.
The February vulnerability appears more serious, however, it's still mitigated by the fact that the attacker can only execute code as the worker process, which severely limits things in the default configuration. But still, only one critical vulnerability in 5 years, with that vulnerabilty being only a few months ago?
Uhh.. you really have no idea what you're talking about. First, MFC is a library, C# is a language. Second, C# was developed by Anders Hjelberg, who Microsoft hired away from Borland. He's the guy that basically wrote Delphi. And no, he did not create MFC.
Uhh.. no, that's not true at all. Since IIS6 was released in 2003, there hasn't been a single critical security vulnerability in IIS. Not one.
The same can't be said for Apache.
You really need to come into this century.
http://en.wikipedia.org/wiki/Powershell
Bigger question is, why isn't it possible to optimize Vista for SSDs? You speak as if Microsoft would have to redesign everything top to bottom in order to support them
The problem is related to the entire core Virtual Memory system in Windows. It assumes a different kind of subsystem than is provided by SSD's. The good news is that Microsoft has already done much of the work for their Embedded Windows project, but that work hasn't been done to Vista.
Windows (NT based) was designed to run on hardware that was a lower spec than it really needed, so it was designed to agressivly prune working sets, use page backed executables, and many other things which improve the performance if your memory and CPU are a bit low. Unfortunately, Microsoft has not changed this design in recent versions of Windows, even though other parts of the OS are far less stingy with resources. Microsoft really needs to retune the kernel to todays hardware.
Nothing without a decent command line is "easy to develop for in my book
You might want to come into this century. http://en.wikipedia.org/wiki/Windows_PowerShell
And yes, IIS and ASP.NET are very easy to develop for. You can download the free Visual Studio Express, Free Sql Server Express, and have web pages up and running in seconds after installation. This gives you the ability to use a debugger to debug web pages and many other features that are a real pain in the ass to get working in the patchwork tools required for PHP and JSP or even RoR. No configuration needed.
Certainly, developing non-trivial web applications is just as difficult over the long haul, but the Microsoft tools make it VERY easy to get started.