(Although, I wouldn't call BetaCam, etc an evolution - it's really a different product aimed at a different market, not where you'd see Star Wars movies. And you must have a better 20 year old BetaMax than me and my memories...)
I guess what I meant to say is that the rest of the world has decided that a multi-user OS (like Unix) is really what should be running on people's desktops. Obviously, Microsoft isn't worried about "doing Unix", but is chasing some of the same design principles.
And you are right, the rest of the world is doing Unix. Because it's cheap. Because you can download some or all of what you need from the net. Nobody really wants to go through the crappy job of rewriting ls and mkdir and so on. So they use GNU or BSD code (or in the case of Microsoft, OS/2 code).
Microsoft is in the process of slowly killing their 'fast single-user' OS. By this time next year, Apple will probably have already killed their's. Amiga has already been killed. (couldn't resist!)
So, when the world is trying to become Unix, why would people want to build something else? (And if one did want a stripped down single user OS, they could build one on top of the Linux kernel. Much of 'Unix' is actually the toolset and the standard userspace stuff.)
"Is there anything special about Unix that has made it particularly suitable for adoption by free software advocates?"
Unix was designed from the beginning to be portable and source compatible. That means that lots of open source code predates Linux, so Linux nicely avoided the "no applications" problem that has plauged things like BeOS and OS/2.
Also, since Unix has been so widely licenced and adopted by different companies and universities, there's a larger base of users familiar with it's internals. Maybe people just like the design, too. But people like the design of Multics and OS/400 too, just that not enough people have seen those OSes to get a serious reverse engineering project going.
I'm curious what you guys are using for XML/XSL processors. Perl? Java?
The reason I ask is that the one experiment I did with XML/XSL was with Microsoft's IE5 thing. It did a simple parse of a 815KB file almost instantly. Of course, MS's stuff is supposedly all non-compliant, so I'm open to other solutions. Just wondering if there is any benchmarks or informal speed comparisons out there...
1) Professional equipment does not use the consumer "BetaMax" format. They only thing the pro stuff has in common with BetaMax is the size of the tape shell.
2) BetaMax never looked better than SuperVHS. Maybe the late and expensive SuperBeta (or whatever) format did, but too little too late.
3) Beta wasn't really a "far superior" format. It was more expensive with only slightly better image quality than VHS. Any 1988 $700 BetaMax you find will look pretty crappy compared to even a $100 2000 VHS machine. Plus the tape length was too short, so many Hollywood movies had to be edited down, or shipped in crappy "EP" speed.
This is all horridly offtopic, but if you are going to open your mouth about Beta, at least try to be somewhat factually correct.
Most of the "GNU System" (gcc, emacs, the user tools) were developed originally on closed source OSes. The GPL specifically allows the use of operating system libraries. Commercial closed source companies also can legally work on GPL code, Samba being a big example of something that supported by the commercial Unix guys.
On the other hand, you do see people arguing that now that Linux exists, there shouldn't be any reason to maintain the GNU system on other platforms. Maybe Palm is an exception to these folks because Linux is not a suitible replacement. However, one could make the same argument for many of the applications on a Windows system versus a Linux one. All I know is that I use GNU stuff on Windows all of the time, and I'm glad it's there and it's legal.
There's also the "non-discrimation" issue, which means that you can't tell the end user what platform he can compile his GPL source on.
So, this is a cultural issues much more than a legal one. That doesn't change your problem that someone needs to step up and maintain the 0.60 toolkit on CygWin. A GPL author isn't required a to support any particular platform, even including Linux, but he can't stop a port or a fork either -- that's the good thing about the GPL.
Hey Bruce -- this is a similar question to what I asked you yesterday. Thanks for restating the problem more articulately and bringing it up with Stallman.
I already know of a system where this sort of thing could potentially be an issue. It's a distributed component directory based on Microsoft COM and ActiveDirectory. It allows developers to call network components where they might not necessarily know where the component is, much less who wrote it and what the license is. (Once can imagine a similar CORBA-based system, especially now that CORBA implementation are being included in Linux distributions.)
I can imagine a situation where a developer might take GPL code, link it with the COM libraries (this would be legal under the GPL, I think), and then register it with this directory. Other closed source, non-GPL programs could then interact with this component. This would be legal under the current GPL, I think, because the software itself is only talking directly to the OS's COM libraries.
This issue might have already been worked out with things like pipes and RPC, however it seems like people have different opinions on the real answer. It will be interesting to see what RMS says about it.
"I don't see why they didn't release a fixed dll."
They have! Newer versions of FrontPage and Interdev do not contain the "weenies" problem.
You seem to be confused -- this is not a flaw in the base IIS product, and the solution is not to uninstall IIS. It's a packaging mistake - where MS included a InterDev 1.0-specific DLL in with the IIS webserver install. They probably did it for political reasons (to make using InterDev 1.0 easier.
Think if RedHat accidentially included a unnecessary and unsecure library in an RPM package. What would their security advisory say? Probably the same thing - delete the library.
Bruce -- The whole point of systems like CORBA or MS DCOM is to allow you to call components across the network that you might know nothing about (except their interfaces). Binding can all be done at runtime.
How does do you think the GPL handles these situations? Can a closed component talk to a GPL one, or is that a problem... The CORBA and COM libraries themselves may-or-may not be considered a 'base operating system' library, depending.
Are you kidding? Try reading the GPL and then compare it to the lore spread on usenet or slashdot. The more I hear about the GPL the more confused I get.
First of all, the GPL does not contain the word "link" and does not directly address "linking" at all. So what is it? Calling a library function? Calling a binary object over CORBA or COM? Is this compile time or run time or both?
Second the "major components of the operating system" clause. What does this mean? According to RMS, this does not include Java, even though Java is included in virtually every OS.
The big problem with the GPL is that there's a big divergance between (1) What it says, which is fairly simple and (2) What RMS, GNU, and everyone else tells you it says and (3) what an author might want it to say.
A better analogy would be Microsoft saying "You can't distribute software linked against MFC unless you've paid us for a MFC developers licence (aka a copy of Visual C++)." Troll Tech could say the same with Qt for Windows.
This would be a seperate issue from pirating Visual C++ to begin with to build your app.
Note that I have no idea if Microsoft/Troll actually has this in their licences, but they could. Similar to the GPL requiring 'runtime' restrictions on the usage of GPL libraries.
Star Wars was hardly at the height of popularity in 1983
Wasn't "Return of the Jedi" released in 1983? Not the height of Star Wars' popularity (that would be around 1977), but "Jedi" was the most popular movie released that year, so Star Wars was still a huge name in popular culture.
It doesn't need to be true cryptographically -- it's an ACL issue, that's all. If you want to grant author rights to your website using clear HTTP authentication, there's nothing stopping you with IIS or Apache or anything else.
Come on -- if the "public as a whole" was ready to worry about security problems, Linux would have about seven users. Every major Linux component has been rootkitted about nine times, here a competant NT admin only needs to delete one obsolete DLL.
The "public" doesn't give a rats ass about this stuff. Nerd politics at it's finest.
The only thing that's weird is that the NT community is so willing to believe that Microsoft backdoored their software. Even to the point of making hay about the whole matter before it was verified.
Small Dick is right -- this bug was reported by none other than Russ Cooper, who runs the "NTBugTraq" list. That makes it "news". (However, notice how Slashdot fails to report root holes in popular Unix software like Bind and WUFTP....)
Cooper shot off his mouth to none other than the Wall Street Journal, before posting to his own list. Slashdot even had the story several hours before I got it in the mail. Once the folks on NTBugTraq looked at it, they determined that "netscapeengineersareweenies" was not a password, but did find another exploit concerning this DLL.
While Cooper's list is an important source for NT admins, right now Russ is definately wearing a boob on his head.
As for the facts of the matter, the DLL in question is only needed to support Visual InterDev 1.0 (obsolete), so by all means it should be deleted. Shame on Microsoft for including it in the base IIS product to begin with.
Yes, everyone knows a senior citizen who is angry that Medicare doesn't cover their expensive drug bills. Everyone has ridden on a crowded airline and eaten their crappy food. Everyone knows someone who has died from cancer.
On the other hand everyone sees Microsoft's name all over their computer. Computer good, Internet good, Microsoft good. Or as good as they understand. Almost nobody has seen their real bastardly tactics, unless you are in the IT business, and even then your career probably depends on Microsoft, so be careful what you think.
Uhhh, Clinton and the Democrats (and most Republicans) supported the capital gains cut on it's political and economic merits, not some arguable moral benefits.
Your argument embodies the hypocracy of these clowns. If a capital gains cut is good for Americans and American families, then argue the merits of that. The government has no mandate to uphold "Christian virtue", and by buying into a policy argument centered on that you have exposed yourself as one of the theocratic droids that have made Reed so successful.
Nobody's said it yet, but does it uphold "Christian virtue" to scotch the anti-trust trial against Microsoft? Maybe now that Reed's pockets have been lined it does. (He sure didn't worry about implying that John McCain was pro-choice - backstabbing one of his own for the money.) I'm sure that you are waiting for your orders from on-high, just as soon as God sends a messenger to articulate these important details of Christian public policy, you'll come out with an argument one way or another.
What you should understand is that the Republican party is made up of two major wings -- the majority of people active in the party are from the Christian Right, working to promote moral values in government (specifically the elimination of Abortion). However, the majority of the party's funders are from the upper class of society and have a quite different pro-business, lower taxes agenda.
Consequentally, these two wings are always making deals with one another, leading to all sorts of interesting contridictions. The Christian folks back the immoral but pro-business of Newt Gingrich. The moderate Doles and Bushes promise to uphold the anti-abortion cause. And so on and so on. If it's beyond your imagination to figure out why Reed would claim that a capital gains cut was a "Christian" policy, you should really start praying, this time for more brains.
(Sure, the democrats have the same sort of divergant factions in their party. But the contradictions aren't so great: Civil Rights folks, labor unions, trial lawyers, and Hollywood types have more in agreement than the faultering "Reagan" coalition of the Republicans.)
"A driver version should not be tied to a kernel version in the first place. "
Linus and everyone else on linux-kernel disagrees with you.
But don't worry -- someone (probably Intel) will donate a UDI implementation for Linux. Just don't expect any of the normal Linux device driver authors to write to it.
This guy is seriously whacked. He once told his legion that changing the capital gains tax laws was the "christian" and "moral" thing to do. If that doesn't make him a dancing monkey for the country club set, I don't know what does.
In the original series, they alluded to a huge war with the Romulans where they were using Nuclear weapons against each other. That might make an interesting series because the 'tech' aspect would be much more primitive and less able to drive the plot.
Oh, the flaps, the velcro straps, the guts pushing out. Ugh, those uniforms were bad. The fact that the actors were fiddling with them all of the time on screen only drew attention to the fact that they were bad.
Slashdot Mirror
Terraserver.com has actually been down for a couple days. There are way more Area 51 nuts than Slashdot can take credit for.
terraserver.microsoft.com is still up - maybe you can find the Area 51 stuff there.
Har, I gracefully submit to your Beta mastery!
(Although, I wouldn't call BetaCam, etc an evolution - it's really a different product aimed at a different market, not where you'd see Star Wars movies. And you must have a better 20 year old BetaMax than me and my memories...)
I guess what I meant to say is that the rest of the world has decided that a multi-user OS (like Unix) is really what should be running on people's desktops. Obviously, Microsoft isn't worried about "doing Unix", but is chasing some of the same design principles.
And you are right, the rest of the world is doing Unix. Because it's cheap. Because you can download some or all of what you need from the net. Nobody really wants to go through the crappy job of rewriting ls and mkdir and so on. So they use GNU or BSD code (or in the case of Microsoft, OS/2 code).
Microsoft is in the process of slowly killing their 'fast single-user' OS.
By this time next year, Apple will probably have already killed their's.
Amiga has already been killed. (couldn't resist!)
So, when the world is trying to become Unix, why would people want to build something else? (And if one did want a stripped down single user OS, they could build one on top of the Linux kernel. Much of 'Unix' is actually the toolset and the standard userspace stuff.)
"Is there anything special about Unix that has made it particularly suitable for adoption by free software advocates?"
Unix was designed from the beginning to be portable and source compatible. That means that lots of open source code predates Linux, so Linux nicely avoided the "no applications" problem that has plauged things like BeOS and OS/2.
Also, since Unix has been so widely licenced and adopted by different companies and universities, there's a larger base of users familiar with it's internals. Maybe people just like the design, too. But people like the design of Multics and OS/400 too, just that not enough people have seen those OSes to get a serious reverse engineering project going.
I'm curious what you guys are using for XML/XSL processors. Perl? Java?
The reason I ask is that the one experiment I did with XML/XSL was with Microsoft's IE5 thing. It did a simple parse of a 815KB file almost instantly. Of course, MS's stuff is supposedly all non-compliant, so I'm open to other solutions. Just wondering if there is any benchmarks or informal speed comparisons out there...
1) Professional equipment does not use the consumer "BetaMax" format. They only thing the pro stuff has in common with BetaMax is the size of the tape shell.
2) BetaMax never looked better than SuperVHS. Maybe the late and expensive SuperBeta (or whatever) format did, but too little too late.
3) Beta wasn't really a "far superior" format. It was more expensive with only slightly better image quality than VHS. Any 1988 $700 BetaMax you find will look pretty crappy compared to even a $100 2000 VHS machine. Plus the tape length was too short, so many Hollywood movies had to be edited down, or shipped in crappy "EP" speed.
This is all horridly offtopic, but if you are going to open your mouth about Beta, at least try to be somewhat factually correct.
Most of the "GNU System" (gcc, emacs, the user tools) were developed originally on closed source OSes. The GPL specifically allows the use of operating system libraries. Commercial closed source companies also can legally work on GPL code, Samba being a big example of something that supported by the commercial Unix guys.
On the other hand, you do see people arguing that now that Linux exists, there shouldn't be any reason to maintain the GNU system on other platforms. Maybe Palm is an exception to these folks because Linux is not a suitible replacement. However, one could make the same argument for many of the applications on a Windows system versus a Linux one. All I know is that I use GNU stuff on Windows all of the time, and I'm glad it's there and it's legal.
There's also the "non-discrimation" issue, which means that you can't tell the end user what platform he can compile his GPL source on.
So, this is a cultural issues much more than a legal one. That doesn't change your problem that someone needs to step up and maintain the 0.60 toolkit on CygWin. A GPL author isn't required a to support any particular platform, even including Linux, but he can't stop a port or a fork either -- that's the good thing about the GPL.
Hey Bruce -- this is a similar question to what I asked you yesterday. Thanks for restating the problem more articulately and bringing it up with Stallman.
I already know of a system where this sort of thing could potentially be an issue. It's a distributed component directory based on Microsoft COM and ActiveDirectory. It allows developers to call network components where they might not necessarily know where the component is, much less who wrote it and what the license is. (Once can imagine a similar CORBA-based system, especially now that CORBA implementation are being included in Linux distributions.)
I can imagine a situation where a developer might take GPL code, link it with the COM libraries (this would be legal under the GPL, I think), and then register it with this directory. Other closed source, non-GPL programs could then interact with this component. This would be legal under the current GPL, I think, because the software itself is only talking directly to the OS's COM libraries.
This issue might have already been worked out with things like pipes and RPC, however it seems like people have different opinions on the real answer. It will be interesting to see what RMS says about it.
"I don't see why they didn't release a fixed dll."
They have! Newer versions of FrontPage and Interdev do not contain the "weenies" problem.
You seem to be confused -- this is not a flaw in the base IIS product, and the solution is not to uninstall IIS. It's a packaging mistake - where MS included a InterDev 1.0-specific DLL in with the IIS webserver install. They probably did it for political reasons (to make using InterDev 1.0 easier.
Think if RedHat accidentially included a unnecessary and unsecure library in an RPM package. What would their security advisory say? Probably the same thing - delete the library.
Bruce -- The whole point of systems like CORBA or MS DCOM is to allow you to call components across the network that you might know nothing about (except their interfaces). Binding can all be done at runtime.
How does do you think the GPL handles these situations? Can a closed component talk to a GPL one, or is that a problem... The CORBA and COM libraries themselves may-or-may not be considered a 'base operating system' library, depending.
Are you kidding? Try reading the GPL and then compare it to the lore spread on usenet or slashdot. The more I hear about the GPL the more confused I get.
First of all, the GPL does not contain the word "link" and does not directly address "linking" at all. So what is it? Calling a library function? Calling a binary object over CORBA or COM? Is this compile time or run time or both?
Second the "major components of the operating system" clause. What does this mean? According to RMS, this does not include Java, even though Java is included in virtually every OS.
The big problem with the GPL is that there's a big divergance between (1) What it says, which is fairly simple and (2) What RMS, GNU, and everyone else tells you it says and (3) what an author might want it to say.
A better analogy would be Microsoft saying "You can't distribute software linked against MFC unless you've paid us for a MFC developers licence (aka a copy of Visual C++)." Troll Tech could say the same with Qt for Windows.
This would be a seperate issue from pirating Visual C++ to begin with to build your app.
Note that I have no idea if Microsoft/Troll actually has this in their licences, but they could. Similar to the GPL requiring 'runtime' restrictions on the usage of GPL libraries.
Star Wars was hardly at the height of popularity in 1983
Wasn't "Return of the Jedi" released in 1983? Not the height of Star Wars' popularity (that would be around 1977), but "Jedi" was the most popular movie released that year, so Star Wars was still a huge name in popular culture.
You'd be sadly surprised of how many IIS developers hardcode the SQL Server SA password into their scripts.
"This isn't true cryptographically."
It doesn't need to be true cryptographically -- it's an ACL issue, that's all. If you want to grant author rights to your website using clear HTTP authentication, there's nothing stopping you with IIS or Apache or anything else.
Come on -- if the "public as a whole" was ready to worry about security problems, Linux would have about seven users. Every major Linux component has been rootkitted about nine times, here a competant NT admin only needs to delete one obsolete DLL.
The "public" doesn't give a rats ass about this stuff. Nerd politics at it's finest.
The only thing that's weird is that the NT community is so willing to believe that Microsoft backdoored their software. Even to the point of making hay about the whole matter before it was verified.
Small Dick is right -- this bug was reported by none other than Russ Cooper, who runs the "NTBugTraq" list. That makes it "news". (However, notice how Slashdot fails to report root holes in popular Unix software like Bind and WUFTP....)
Cooper shot off his mouth to none other than the Wall Street Journal, before posting to his own list. Slashdot even had the story several hours before I got it in the mail. Once the folks on NTBugTraq looked at it, they determined that "netscapeengineersareweenies" was not a password, but did find another exploit concerning this DLL.
While Cooper's list is an important source for NT admins, right now Russ is definately wearing a boob on his head.
As for the facts of the matter, the DLL in question is only needed to support Visual InterDev 1.0 (obsolete), so by all means it should be deleted. Shame on Microsoft for including it in the base IIS product to begin with.
So first you say Reed is not working for Bush.
Then number of people correct you.
Then you change the subject and make fun of RMS.
Did you forget to take your Ritalin today, Rombuu?
Yes, everyone knows a senior citizen who is angry that Medicare doesn't cover their expensive drug bills. Everyone has ridden on a crowded airline and eaten their crappy food. Everyone knows someone who has died from cancer.
On the other hand everyone sees Microsoft's name all over their computer. Computer good, Internet good, Microsoft good. Or as good as they understand. Almost nobody has seen their real bastardly tactics, unless you are in the IT business, and even then your career probably depends on Microsoft, so be careful what you think.
Uhhh, Clinton and the Democrats (and most Republicans) supported the capital gains cut on it's political and economic merits, not some arguable moral benefits.
Your argument embodies the hypocracy of these clowns. If a capital gains cut is good for Americans and American families, then argue the merits of that. The government has no mandate to uphold "Christian virtue", and by buying into a policy argument centered on that you have exposed yourself as one of the theocratic droids that have made Reed so successful.
Nobody's said it yet, but does it uphold "Christian virtue" to scotch the anti-trust trial against Microsoft? Maybe now that Reed's pockets have been lined it does. (He sure didn't worry about implying that John McCain was pro-choice - backstabbing one of his own for the money.) I'm sure that you are waiting for your orders from on-high, just as soon as God sends a messenger to articulate these important details of Christian public policy, you'll come out with an argument one way or another.
What you should understand is that the Republican party is made up of two major wings -- the majority of people active in the party are from the Christian Right, working to promote moral values in government (specifically the elimination of Abortion). However, the majority of the party's funders are from the upper class of society and have a quite different pro-business, lower taxes agenda.
Consequentally, these two wings are always making deals with one another, leading to all sorts of interesting contridictions. The Christian folks back the immoral but pro-business of Newt Gingrich. The moderate Doles and Bushes promise to uphold the anti-abortion cause. And so on and so on. If it's beyond your imagination to figure out why Reed would claim that a capital gains cut was a "Christian" policy, you should really start praying, this time for more brains.
(Sure, the democrats have the same sort of divergant factions in their party. But the contradictions aren't so great: Civil Rights folks, labor unions, trial lawyers, and Hollywood types have more in agreement than the faultering "Reagan" coalition of the Republicans.)
"A driver version should not be tied to a kernel version in the first place. "
Linus and everyone else on linux-kernel disagrees with you.
But don't worry -- someone (probably Intel) will donate a UDI implementation for Linux. Just don't expect any of the normal Linux device driver authors to write to it.
Reed is on the Bush payroll right now.
This guy is seriously whacked. He once told his legion that changing the capital gains tax laws was the "christian" and "moral" thing to do. If that doesn't make him a dancing monkey for the country club set, I don't know what does.
In the original series, they alluded to a huge war with the Romulans where they were using Nuclear weapons against each other. That might make an interesting series because the 'tech' aspect would be much more primitive and less able to drive the plot.
Oh, the flaps, the velcro straps, the guts pushing out. Ugh, those uniforms were bad. The fact that the actors were fiddling with them all of the time on screen only drew attention to the fact that they were bad.