Re:My take on the death of the Trek series...
on
Voyager Eulogy
·
· Score: 1
No, it was a smart move. The show had moved so far to the watered-down mainstream that it had essentially lost the hardcore trekie audience. That doesn't bode well for the future of the franchise.
T&A got the core fanbase back. They might not have liked the plots, but they were watching with some interest at least. That puts them in an ideal position for "Enterprise" which, on the face of it, is going to be a more Trekie-type show. The idea appears to be to bring the librarian crowd closer to the spock ears, rather than visa-versa.
(And ironically, the non-trekie females I know that watch the show didn't dislike the 7of9 character) --
I've noticed the same thing, both on/. and in newsgroups I've been reading over the years. The problem is there's only a finite number of things to say about Topic X, and once the flamewar has been had and everyone knows what the positions are, it's pretty much over and the interesting people drift off. You have to have a dedicated group of flamebots to keep the same argument rolling (see abortion arguments or OS advocacy groups).
To keep things going, you either need a steady supply of news (which is how/. does it), or a willingness to go waaaay off-topic in interesting ways (see alt.folklore.computers for example). --
I admit the situation looks bleak. Most ISPs aren't in exactly the best financial state and text-only Usenet appeals to such a small number of users that it's almost inevitable it will go away unless folks call and complain when it does. (My news server went down for a week, which at least was an opportunity to go on record with my ISP).
Furthermore, there's thousands of groups which get 0 traffic (.bork.bork groups which got a 5 second laugh from someone in 1993 and the like are still there!) and there's apparently no cabal going around and voting those groups out of existance. From the harried ISP SA point-of-view, the entire thing must seem decrepit and unmanaged.
Maybe the easy solution for the binaries problem is to just press the Off button on the news server, but hopefully they'll have more insight than that. Binary distribution via news can work, but only if it's limited to a very small number of servers with large interconnects. Unfortunately, huge ISPs such as @home don't want to make a special case of binary news, probably because it would draw attention to the fact that almost everything there is illegal. --
Too bad OSI applied for a trademark on "Open Source" and got shot down because the term is too generic and has a plain english meaning.
You can have an "Open Source Definition", but there's nothing preventing Microsoft or Sun or anyone else from using the term "Open Source" to describe whatever they want. In fact the term runs counter to a long usage tradition of the word "Open" in the industry for software/hardware that has documented interfaces and behavior.
Which why "Open Source" is a lousy term for propaganda value.
There was an opportunity to invent a real identifable 'brand' a few years ago, but it was missed ("OSI-approved" comes close, but is dull), which is exactly why most slashdotters will spend their life arguing about what is or isn't "open source". --
Back in the early-90s, "in the know" people would tell their AOL using friends "The Internet is sooo cool! Check out Usenet - there's everything!". Usenet was the premier Internet application (next to e-mail), and therefore every luser ended up there.
Now, nobody's on Usenet execept the people who know about Usenet. Not even the spammers are on usenet anymore (well, maybe in the porn groups still). J.Random AOL doesn't even know what Google is, much less groups.google.com.
I dunno. Maybe you are just reading different groups than me. Usenet seems in the best shape in ages - the endless september seems to have ended. --
Re:My take on the death of the Trek series...
on
Voyager Eulogy
·
· Score: 2
Trek is a massive cash cow for Paramount, a major franchise for them. Because of this, they were far less likely to do anything that would cut off that revenue flow.
Since we live in a capitalist society, the key question for the Star Trek franchise is not "How big is it?" but instead "How fast is it growing?"
And, unfortunately, the hardcore (90% male) nerd market has been pretty much saturated. Starting with about the 3rd season of TNG, the francise made radical adjustments to gain more female viewership. This continued with Voyager (which after all was a "network" show), by trying to appeal to a much broader and less science fiction-oriented mainstream audience.
Unfortunately, it was full of flat, boring, motavationless characters, no soap operatics, very little space operatics, and the over-simplified particle-of-the-week plots failed to appeal to both the Technical Manual crowd and their mothers. So, the show failed, both critically and more importantly in the ratings. If it didn't have "Star Trek" in the title, it would have been cancelled after a year or two, and the only reason it lasted as long as it did was the wise addition of some T&A to get back the spock ears group. --
Partially Correct: In ME it was not done so much to isolate the OS from the user,
as to remove all the legacy DOS cruft that was a large source of driver problems
for Windows 9x systems.
The irony is that the most significant reason that Win9x was ever created was because the users demanded support for "legacy DOS cruft". And, all things considered, Win9x did an excellent job of supporting ancient drivers while presenting an WinNT-like modern API to applications. It was a compatibilty-hack to transition users to a modern system, nothing more.
Since they removed the old driver support, there is no real reason for the OS to even be on the market. Resource requirements aren't significantly less than the real thing (NT) either. So ME was just really a pointless upgrade to fatten Microsoft's wallet and send one final parting shot to users that they better upgrade their stuff. --
Windows 95 contains a substantial amount of code which "integrates" the MSN client into the desktop. AOL complained about it to the feds, but that actually went nowhere.
The reason that AOL got an icon in later revs of Win95 is because they chose to excusively use IE instead of Netscape. In fact they signed a deal with Netscape (then independant) and broke it the very next day when the found they could bargin desktop real-estate out of MS. --
You're right on the money. Windows XP is yet another attempt at wiring the Windows interface for MSN access.
The last time MS tried this (with Windows 95), they had to face a fairly long federal trade commission investigation, and while no charges were filed, it started the log-rolling towards the anti-trust case.
By including the AOL software, they've basically bought off the only business with stature to file suit against them for extending their monopoly.
Besides, anyone remember the Bill Gates comment about including a can of Pepsi in every six pack of Coke/ That's exactly what they are doing now!
The *real* problem is that there is only ONE independant implementation of Intel's IA32 architecture, and it's Transmeta's.
Yes, believe it or not, AMD is a licencee of Intel technology, and therefore the Athlon is somewhat dependant on Intel's goodwill.* Fortunately for them, Intel has a large number of contracts (government and otherwise) which require multiple sources for their tech.
Now, if Intel decides to be a bad boy and pulls AMD's x86 licence, Transmeta's IP portfolio starts to become very helpful. Namely, AMD will be able to stay in business.
*Source: AMD quartery report:
RECENT DEVELOPMENTS
On May 4, 2001, we, along with Intel Corporation, announced the renewal of the
patent cross-license agreement between the companies. --
Well, Microsoft is very interested in feature checkbox marketing. They obviously don't like to see something like:
IPv6
--------
Linux: Yes
Solaris: Yes
Windows 2000: No
64-bit support
-------------
Linux: Yes
Solaris: Yes
Windows: Pending
BTW, did anyone else notice, that despite 6+ YEARS of promises that "64-bit Windows will ship on the same day as IA64", that Itanium will roll out with only beta-level Windows support? Somebody submit it to Slashdot so we can have a 600 post flamefest. Story on ZDNet. --
Still, a GPL program that says "obtain MSVCP70.DLL somehow" on it's face seems to defy the licence. As does a GPL program that requires the Borland compiler, or a Java 2 program on a platform which doesn't ship with Java 2.
Generally the way the GPL has dealt with this has been to require a specific exception from the author in addition to the licence. If this exception is missing (see KDE), people freak, and it also prevents the author from using standard GPL code without the exception.
And if an implied or explicit compiler library exception is OK, why isn't one for "Money Guzzler Inc."'s libraries?
Likewise, does it really help the cause of Free Software to (say) prohibit GPL plug-ins for Netscape?
Saying "this is effectively that" probably doesn't hold contract law water, so hopefully GPL3 will deal with these issues in a flexible manner. --
Re:Lawyer: Keep in mind . . .
on
GPL FAQ
·
· Score: 2
Thanks for your response, which as usual is very well stated.
IANAL, but as Bruce Perens has repeatedly pointed out, intent goes a long way into making the linking situation more clear, at least until you hit the network interface.
But it's entirely unclear what the intent of the "major component of the operating system" clause is once you get past a typical 1980s-style Unix system. Having built in object brokers (COM+) and virtual machines (Java) and add-on compilers (VC) and API personalities (Interix) don't help. And I'm really just throwing logs into the GPL v3 fire on that issue alone. --
I'm writting a Windows application with Microsoft Visual C++ and I will be releasing it under the GPL. Is dynamically linking my program with the Visual C++ run-time library permitted under the GPL?
Yes, because that run-time library normally accompanies the compiler you are using
So even though the VC runtimes aren't "a major component of the operating system", because they are distributed seperately, according to this I can still use them.
Can I use the GPL for a plug-in for a non-free program?
If the program uses fork and exec to invoke plug-ins, then the plug-ins are separate programs, so the license for the main program makes no requirements for them. So you can use the GPL for a plug-in, and there are no special requirements.
OK, so I can't turn VirtuaDub into COM component. But if I register it with COM+ and run it in a seperate process, then it's OK?
I'd like to modify GPL-covered programs and link them with the portability libraries from Money Guzzler Inc. I cannot distribute the source code for these libraries, so any user who wanted to change these versions would have to buy those libraries. Why doesn't the GPL permit this?
So I can use operating system runtimes, and Visual C++ runtimes, but I can't use 3rd party runtimes that don't happen to ship with a compiler. I take from this I probably could write a J2SE GPL application (using included libraries) but not a J2EE version (which might be from a 3rd party). So, the question is -- Is Microsoft's GPLed software shipping with Interix legal?!? --
Digging on google came up with the year of 1995 as the "first" year that the gaming industry passed up Hollywood. Take it as more of an order of magnitude than a great claim. --
COM) It all depends on your interpretation of "derived work". It's true that if significant functionality was derived from a GPL COM object, you'd be in trouble. (I don't know if Virtua Dub actually wrapped the DLLs in COM or not, but it shouldn't matter.)
For example, if I write a GPL media player COM control, does Word and IE have to be open sourced because someone could embed the control there?
DCOM) MS DCOM/COM+ eventually turns into MS-RPC, so your distinction doesn't make any sense.
I can write open source server component (OK - talks to the OS DCOM libraries) and talk to it with a closed source client (talks through the OS DCOM libraries and RPC to get the unrestricted 'output' of the server). --
Laptop support is problematic even on Windows because laptop hardware sucks. Ever notice how IBM and Dell can charge a sigificant premium for their gear just because it "works as advertised"?
I was a little dismayed about Alan Cox's recent comments about ACPI, namely that he doesn't care. That's all fine and good until systems drop APM and 'PnP BIOS' support and Linux needs to start spending a year catching up. (And ACPI isn't just about power management -- it's the whole device discovery routine.)
But it's understandable. Apparently Win98 shipped using a subset of ACPI which lead the BIOS folks to not produce compatible implementations. Win98SE installed with ACPI disabled, so nobody worked on improving it. When Windows 2000 shipped, lots of folks were caught with their pants down -- I still see posts about people complaining about ACPI issues on Via boards for example. --
OS/2's desktop had terrible usability faults and was ugly as shit. (By the time this was largely resolved in Warp 4, it was too late.) Sure, there was some nice powerful features there, but the default organization of it all was a complete and total obscured mess.
Now, that might be acceptable if (as with Linux), the desktop was developed as an add-on for a server operating system. But OS/2 was designed and marketed almost purely as desktop operating system! --
Just a note so that you feel better qualified to discuss GUI issues in the future. The reason that MacOS uses the popup effect and the watch cursor and the grayed icon is to provide direct feedback to the users actions. Don't forget that their eyeballs are pointing at the mouse cursor, not reading system monitor tea leaves. ('Hmm, I know that disk read load pattern. Must be mozilla starting up.')
Now, I'm all for you finding a solution that works for you, but the entire discussion is about the "not good enough" failures of the Unix desktop. And trying to bullshit basic failures such as this by pointing at top or whatever is really pathetic and one big reason the whole thing feels half-assed.
Whoops -- disregard some of the last post. I was conflating your post with grammar fascist's. Apologies.
However, contrary to what you say, I don't think the system call numbers are standardized at all between unixes (although the names are), and I also doubt very much that you could build a libc for Linux without including the kernel headers. (And don't some programs like Oracle bypass glibc all together?)
That's why it's essentially an exception. The GPL doesn't contain special language pertaining to kernels or components which follow documented Single Unix Specificaiton interfaces.
(And there are people that have squinted at the GPL long enough to believe that it prohibits runtime binary loading or 'derived' works that don't actually link to the headers.) --
So you are saying that anyone can defeat the GPL by writing LGPL wrapper code around the headers. Somehow I don't think so.
It's true that user space developers don't have to worry about licence issues because libc is LGPL. However, the only reason LIBC can be LGPL is because of the Linus exception.
Anyway, time to start getting chapter and verse on your ass. From the kernel's COPYING file:
NOTE! This copyright does *not* cover user programs that use kernel
services by normal system calls - this is merely considered normal use
of the kernel, and does *not* fall under the heading of "derived work".
Linus Torvalds
Now, what's "normal use" versus GNU's idea of a "derived work"? AFAIK, the kernel is one of the few works that makes this distinciton, which most GNUites would reject.
Linux is actually a special case because Linus has modified or clarified the GPL to allow certain things:
1) Proprietary and non-GPL applications CAN include the kernel headers. Otherwise the entire system would have to be GPL.
2) You can even load closed source modules directly into the kernel, something the Stallmanites would oppose, although the GPL totally unclear here.
Like you said, this all hinges on the interpretation of "link". However, Linus has offered his opinion, and the developer community has gone along with it.
(In my view, the "link" and "major components operating system" issues are the things in most dire need of clarification in the GPL. Both for binary runtime loading (plug-ins or modules or components) and also to accomodate optional runtime enviornments like Java.) --
Actually, many of those Apple ][ and C64 games did bring in millions of dollars. The videogame industry first surpassed Hollywood in revenue in around 1982, the good times lasting to the 1984 crash. --
Slashdot Mirror
No, it was a smart move. The show had moved so far to the watered-down mainstream that it had essentially lost the hardcore trekie audience. That doesn't bode well for the future of the franchise.
T&A got the core fanbase back. They might not have liked the plots, but they were watching with some interest at least. That puts them in an ideal position for "Enterprise" which, on the face of it, is going to be a more Trekie-type show. The idea appears to be to bring the librarian crowd closer to the spock ears, rather than visa-versa.
(And ironically, the non-trekie females I know that watch the show didn't dislike the 7of9 character)
--
I've noticed the same thing, both on /. and in newsgroups I've been reading over the years. The problem is there's only a finite number of things to say about Topic X, and once the flamewar has been had and everyone knows what the positions are, it's pretty much over and the interesting people drift off. You have to have a dedicated group of flamebots to keep the same argument rolling (see abortion arguments or OS advocacy groups).
/. does it), or a willingness to go waaaay off-topic in interesting ways (see alt.folklore.computers for example).
To keep things going, you either need a steady supply of news (which is how
--
I admit the situation looks bleak. Most ISPs aren't in exactly the best financial state and text-only Usenet appeals to such a small number of users that it's almost inevitable it will go away unless folks call and complain when it does. (My news server went down for a week, which at least was an opportunity to go on record with my ISP).
Furthermore, there's thousands of groups which get 0 traffic (.bork.bork groups which got a 5 second laugh from someone in 1993 and the like are still there!) and there's apparently no cabal going around and voting those groups out of existance. From the harried ISP SA point-of-view, the entire thing must seem decrepit and unmanaged.
Maybe the easy solution for the binaries problem is to just press the Off button on the news server, but hopefully they'll have more insight than that. Binary distribution via news can work, but only if it's limited to a very small number of servers with large interconnects. Unfortunately, huge ISPs such as @home don't want to make a special case of binary news, probably because it would draw attention to the fact that almost everything there is illegal.
--
Too bad OSI applied for a trademark on "Open Source" and got shot down because the term is too generic and has a plain english meaning.
You can have an "Open Source Definition", but there's nothing preventing Microsoft or Sun or anyone else from using the term "Open Source" to describe whatever they want. In fact the term runs counter to a long usage tradition of the word "Open" in the industry for software/hardware that has documented interfaces and behavior.
Which why "Open Source" is a lousy term for propaganda value.
There was an opportunity to invent a real identifable 'brand' a few years ago, but it was missed ("OSI-approved" comes close, but is dull), which is exactly why most slashdotters will spend their life arguing about what is or isn't "open source".
--
Back in the early-90s, "in the know" people would tell their AOL using friends "The Internet is sooo cool! Check out Usenet - there's everything!". Usenet was the premier Internet application (next to e-mail), and therefore every luser ended up there.
Now, nobody's on Usenet execept the people who know about Usenet. Not even the spammers are on usenet anymore (well, maybe in the porn groups still). J.Random AOL doesn't even know what Google is, much less groups.google.com.
I dunno. Maybe you are just reading different groups than me. Usenet seems in the best shape in ages - the endless september seems to have ended.
--
Trek is a massive cash cow for Paramount, a major franchise for them. Because of this, they were far less likely to do anything that would cut off that revenue flow.
Since we live in a capitalist society, the key question for the Star Trek franchise is not "How big is it?" but instead "How fast is it growing?"
And, unfortunately, the hardcore (90% male) nerd market has been pretty much saturated. Starting with about the 3rd season of TNG, the francise made radical adjustments to gain more female viewership. This continued with Voyager (which after all was a "network" show), by trying to appeal to a much broader and less science fiction-oriented mainstream audience.
Unfortunately, it was full of flat, boring, motavationless characters, no soap operatics, very little space operatics, and the over-simplified particle-of-the-week plots failed to appeal to both the Technical Manual crowd and their mothers. So, the show failed, both critically and more importantly in the ratings. If it didn't have "Star Trek" in the title, it would have been cancelled after a year or two, and the only reason it lasted as long as it did was the wise addition of some T&A to get back the spock ears group.
--
Partially Correct: In ME it was not done so much to isolate the OS from the user,
as to remove all the legacy DOS cruft that was a large source of driver problems
for Windows 9x systems.
The irony is that the most significant reason that Win9x was ever created was because the users demanded support for "legacy DOS cruft". And, all things considered, Win9x did an excellent job of supporting ancient drivers while presenting an WinNT-like modern API to applications. It was a compatibilty-hack to transition users to a modern system, nothing more.
Since they removed the old driver support, there is no real reason for the OS to even be on the market. Resource requirements aren't significantly less than the real thing (NT) either. So ME was just really a pointless upgrade to fatten Microsoft's wallet and send one final parting shot to users that they better upgrade their stuff.
--
Did anyone know when Itanium would actually ship back in November? (Even now, launch date is soft.)
--
Windows 95 contains a substantial amount of code which "integrates" the MSN client into the desktop. AOL complained about it to the feds, but that actually went nowhere.
The reason that AOL got an icon in later revs of Win95 is because they chose to excusively use IE instead of Netscape. In fact they signed a deal with Netscape (then independant) and broke it the very next day when the found they could bargin desktop real-estate out of MS.
--
You're right on the money. Windows XP is yet another attempt at wiring the Windows interface for MSN access.
The last time MS tried this (with Windows 95), they had to face a fairly long federal trade commission investigation, and while no charges were filed, it started the log-rolling towards the anti-trust case.
By including the AOL software, they've basically bought off the only business with stature to file suit against them for extending their monopoly.
Besides, anyone remember the Bill Gates comment about including a can of Pepsi in every six pack of Coke/ That's exactly what they are doing now!
--
The *real* problem is that there is only ONE independant implementation of Intel's IA32 architecture, and it's Transmeta's.
Yes, believe it or not, AMD is a licencee of Intel technology, and therefore the Athlon is somewhat dependant on Intel's goodwill.* Fortunately for them, Intel has a large number of contracts (government and otherwise) which require multiple sources for their tech.
Now, if Intel decides to be a bad boy and pulls AMD's x86 licence, Transmeta's IP portfolio starts to become very helpful. Namely, AMD will be able to stay in business.
*Source: AMD quartery report:
RECENT DEVELOPMENTS
On May 4, 2001, we, along with Intel Corporation, announced the renewal of the
patent cross-license agreement between the companies.
--
Well, Microsoft is very interested in feature checkbox marketing. They obviously don't like to see something like:
IPv6
--------
Linux: Yes
Solaris: Yes
Windows 2000: No
64-bit support
-------------
Linux: Yes
Solaris: Yes
Windows: Pending
BTW, did anyone else notice, that despite 6+ YEARS of promises that "64-bit Windows will ship on the same day as IA64", that Itanium will roll out with only beta-level Windows support? Somebody submit it to Slashdot so we can have a 600 post flamefest. Story on ZDNet.
--
Still, a GPL program that says "obtain MSVCP70.DLL somehow" on it's face seems to defy the licence. As does a GPL program that requires the Borland compiler, or a Java 2 program on a platform which doesn't ship with Java 2.
Generally the way the GPL has dealt with this has been to require a specific exception from the author in addition to the licence. If this exception is missing (see KDE), people freak, and it also prevents the author from using standard GPL code without the exception.
And if an implied or explicit compiler library exception is OK, why isn't one for "Money Guzzler Inc."'s libraries?
Likewise, does it really help the cause of Free Software to (say) prohibit GPL plug-ins for Netscape?
Saying "this is effectively that" probably doesn't hold contract law water, so hopefully GPL3 will deal with these issues in a flexible manner.
--
Thanks for your response, which as usual is very well stated.
IANAL, but as Bruce Perens has repeatedly pointed out, intent goes a long way into making the linking situation more clear, at least until you hit the network interface.
But it's entirely unclear what the intent of the "major component of the operating system" clause is once you get past a typical 1980s-style Unix system. Having built in object brokers (COM+) and virtual machines (Java) and add-on compilers (VC) and API personalities (Interix) don't help. And I'm really just throwing logs into the GPL v3 fire on that issue alone.
--
COM+ is a "major component of the operating system", so therefore you can write GPL programs that link to it.
--
I'm writting a Windows application with Microsoft Visual C++ and I will be releasing it under the GPL. Is dynamically linking my program with the Visual C++ run-time library permitted under the GPL?
Yes, because that run-time library normally accompanies the compiler you are using
So even though the VC runtimes aren't "a major component of the operating system", because they are distributed seperately, according to this I can still use them.
Can I use the GPL for a plug-in for a non-free program?
If the program uses fork and exec to invoke plug-ins, then the plug-ins are separate programs, so the license for the main program makes no requirements for them. So you can use the GPL for a plug-in, and there are no special requirements.
OK, so I can't turn VirtuaDub into COM component. But if I register it with COM+ and run it in a seperate process, then it's OK?
I'd like to modify GPL-covered programs and link them with the portability libraries from Money Guzzler Inc. I cannot distribute the source code for these libraries, so any user who wanted to change these versions would have to buy those libraries. Why doesn't the GPL permit this?
So I can use operating system runtimes, and Visual C++ runtimes, but I can't use 3rd party runtimes that don't happen to ship with a compiler. I take from this I probably could write a J2SE GPL application (using included libraries) but not a J2EE version (which might be from a 3rd party). So, the question is -- Is Microsoft's GPLed software shipping with Interix legal?!?
--
Digging on google came up with the year of 1995 as the "first" year that the gaming industry passed up Hollywood. Take it as more of an order of magnitude than a great claim.
--
COM) It all depends on your interpretation of "derived work". It's true that if significant functionality was derived from a GPL COM object, you'd be in trouble. (I don't know if Virtua Dub actually wrapped the DLLs in COM or not, but it shouldn't matter.)
For example, if I write a GPL media player COM control, does Word and IE have to be open sourced because someone could embed the control there?
DCOM) MS DCOM/COM+ eventually turns into MS-RPC, so your distinction doesn't make any sense.
I can write open source server component (OK - talks to the OS DCOM libraries) and talk to it with a closed source client (talks through the OS DCOM libraries and RPC to get the unrestricted 'output' of the server).
--
Laptop support is problematic even on Windows because laptop hardware sucks. Ever notice how IBM and Dell can charge a sigificant premium for their gear just because it "works as advertised"?
I was a little dismayed about Alan Cox's recent comments about ACPI, namely that he doesn't care. That's all fine and good until systems drop APM and 'PnP BIOS' support and Linux needs to start spending a year catching up. (And ACPI isn't just about power management -- it's the whole device discovery routine.)
But it's understandable. Apparently Win98 shipped using a subset of ACPI which lead the BIOS folks to not produce compatible implementations. Win98SE installed with ACPI disabled, so nobody worked on improving it. When Windows 2000 shipped, lots of folks were caught with their pants down -- I still see posts about people complaining about ACPI issues on Via boards for example.
--
OS/2's desktop had terrible usability faults and was ugly as shit. (By the time this was largely resolved in Warp 4, it was too late.) Sure, there was some nice powerful features there, but the default organization of it all was a complete and total obscured mess.
Now, that might be acceptable if (as with Linux), the desktop was developed as an add-on for a server operating system. But OS/2 was designed and marketed almost purely as desktop operating system!
--
Just a note so that you feel better qualified to discuss GUI issues in the future. The reason that MacOS uses the popup effect and the watch cursor and the grayed icon is to provide direct feedback to the users actions. Don't forget that their eyeballs are pointing at the mouse cursor, not reading system monitor tea leaves. ('Hmm, I know that disk read load pattern. Must be mozilla starting up.')
Now, I'm all for you finding a solution that works for you, but the entire discussion is about the "not good enough" failures of the Unix desktop. And trying to bullshit basic failures such as this by pointing at top or whatever is really pathetic and one big reason the whole thing feels half-assed.
--
Whoops -- disregard some of the last post. I was conflating your post with grammar fascist's. Apologies.
However, contrary to what you say, I don't think the system call numbers are standardized at all between unixes (although the names are), and I also doubt very much that you could build a libc for Linux without including the kernel headers. (And don't some programs like Oracle bypass glibc all together?)
That's why it's essentially an exception. The GPL doesn't contain special language pertaining to kernels or components which follow documented Single Unix Specificaiton interfaces.
(And there are people that have squinted at the GPL long enough to believe that it prohibits runtime binary loading or 'derived' works that don't actually link to the headers.)
--
So you are saying that anyone can defeat the GPL by writing LGPL wrapper code around the headers. Somehow I don't think so.
It's true that user space developers don't have to worry about licence issues because libc is LGPL. However, the only reason LIBC can be LGPL is because of the Linus exception.
Anyway, time to start getting chapter and verse on your ass. From the kernel's COPYING file:
NOTE! This copyright does *not* cover user programs that use kernel
services by normal system calls - this is merely considered normal use
of the kernel, and does *not* fall under the heading of "derived work".
Linus Torvalds
Now, what's "normal use" versus GNU's idea of a "derived work"? AFAIK, the kernel is one of the few works that makes this distinciton, which most GNUites would reject.
--
Linux is actually a special case because Linus has modified or clarified the GPL to allow certain things:
1) Proprietary and non-GPL applications CAN include the kernel headers. Otherwise the entire system would have to be GPL.
2) You can even load closed source modules directly into the kernel, something the Stallmanites would oppose, although the GPL totally unclear here.
Like you said, this all hinges on the interpretation of "link". However, Linus has offered his opinion, and the developer community has gone along with it.
(In my view, the "link" and "major components operating system" issues are the things in most dire need of clarification in the GPL. Both for binary runtime loading (plug-ins or modules or components) and also to accomodate optional runtime enviornments like Java.)
--
Actually, many of those Apple ][ and C64 games did bring in millions of dollars. The videogame industry first surpassed Hollywood in revenue in around 1982, the good times lasting to the 1984 crash.
--