Hmm, didn't Microsoft's lawyers patent innovation? Or will this, since it's made with GPLed software that obviously (by Bill's standards) hinders innovation, force a rip in the space time continuum and send us all to an alternate Universe where up is down, down is up, cats and dogs live together, and Microsoft actually invents new technology instead of buying it?
Wasn't the XBox supposed to crush Sony like a grape because it used commodity parts while silly Sony used specialized ones, therefore much more expensive?
SCO is getting worse than the crazy homeless people in San Francisco that scream Bible passages at you as you're walking by..
Is the weird asian guy with the sunglasses and signs still walking Market? I took a pic of him once, and he spooked, I think he thought I was part of some Clintonian conspiracy. Was glad to be part of his tinfoil-hat paranoia for a day.
Ahh, the "hooker trading cards" as one friend called them.
There's a reason for them handing them out in the weird way they do, silently just holding them out. I think the rule in Vegas is if someone just "picks them up" it's not illegal. But if they say anything and try to entice you into grabbing any, it's solicitation.
Benefits to users and to the Debian Project include <FLAMMABLE>Benefits to RMS include stroking his ego (it's really the GNU in GNU/Linux that matter, trust me!!) even if it comes at the expense (to the user) of what makes the *BSDs great - their single source cohesiveness.</FLAMMABLE>
It's not cross referencing between Linux and FreeBSD source, it's just a cross reference within the software itself. Fr'instance, I have a driver, and it talks about some function in the DDK, you click on a link and it takes you there. Very confusing title I admit.
Closable BSDs? Yes someone could make a fork of FreeBSD and close it, but the original FreeBSD source would still exist.
1) Not a single vendor ban, they want to raise OpenSource, not kill Microsoft. Since Microsoft is a monopoly, it just seems like a single vendor ban. 2) Essential becuase of network effects and inertia. If Microsoft and Linux have the same effective purchase cost (because of piracy) why would you overcome the inertia to switch? 3) Necessary because of trade regulations that require piracy numbers to go lower. No license to purchase, no piracy. Thats why it needs to be a law.
Publicly accessible doesn't mean it's in Public Domain. Public Domain has a very specific meaning in copyright.
GPL is copyright (err, copyleft) which puts constraints on how you can use the code. Public Domain means no copyright at all, do anything you want to it whatsover.
No one looks forward to going to Court unless they are sure they are going to win HUGE. But most folks say they look forward to it, need the poker face whether or not you think you'll win.
Going to court is a PITA, no matter how strong you think your case is. Lawyers cost. Getting info together costs. Pulling your developer off a project so he can answer some bullshit about JFS when it wasn't even first developed for AIX or Linux costs.
this has been (kinda) answered by others, but I'll try to make mine a bit clearer.
The above statement is talking relative to release of 4.8, as cut to CD. This doesn't mean it hasn't been fixed, it just means they can't go back in time to fix it on 4.8 as it was on release day (think what was cut to CD).
That said, FreeBSD users don't have to stay on the "as cut to CD version". Once you get a release, a good FreeBSD user can update his system, tracking one of a few cvs branches, such as STABLE (which will get you this whole release), the 4.8-RELENG "security fixes only" branch, or CURRENT, which would put you in 5.x world. All security problems are fixed in all releases.
If this is the case, why isn't the/. community all over them like they are Microsoft? Because all critical vulnerabilities are on a mailing list, all versions of FreeBSD affected are updated, even old ones (you'll occasionally see updates for 3.x and even 2.x sometimes) instead of forcing people to upgrade off NT 4 so they can sell more XP licenses.
The APIs for me seem hacked together. Look at all the filename overloads for OpenFile, yet it still can't open a named pipe. How are they consistent? UNIX got this right (though many things wrong in other areas) by realizing the kernel should know what you should do with the open, and not have to hack the API, adding kludges to code and have to explain these special cases in the documentation. Then all of the API calls with named...Ex, because they didn't get the first implementation right.
In some ways, this is not me being harsh at microsoft. Linux has a lot less baggage because of it's youth, and a lot of things are inconsistent in the kernel DDI, and the windowing libraries tend to change often. But the system apis are stable becuase they are all based on standards.
Reasonably consistent consistent APIs across products
Talk to the Samba guys about how inconsistent they are about protocols. They are a huge company, and many things are inconsistent. You do raise valid points tho, and many in the OS community don't want to hear anything negative.
There was a MS funded benchmark a while back, where Windows came out on top of Linux when it came to webserver performance. The great sea of Slashdotters were up in arms, They shilled for MS!!! A few people actually decided to think "maybe they're right" and looked for improvements in Linux networking code. And Linux got faster, and has beat beating Windows IIS's ass ever since. There are advantages to listening to bad news sometimes.
We do this - they backported a bunch of things from Linux 2.6 into it. The one thing we needed was the 4G user/4G kernel option. We were porting some things from solaris and they couldn't allocate enough memory with the 3G kernel split. Works pretty well.
Serial ports may lessen in importance, but they can never be completely replaced by USB. Certain RT events require hard timing, which USB can never provide. My Brother In Law needed to buy a PCMCIA serial port card for his laptop to talk to a ladder programmer controller.
They may disappear as defaults from machines though, like the floppy has been eliminated as standard from Macs and now Dells.
Linus wrote the kernel, unsatisfied with Minix and its licensing restrictions, and really wanting to become a tinkerer. He never really cared about userland and he took the freely available GNU tools and got his kernel running. Because a huge part (essentially all) of the early Linux userland was GNU tools, RMS felt that the FSF was justified for taking part credit for Linux. When Stallman first proposed GNU/Linux, Linus thought he was talking about Debian Linux, which was at that time the most FSF friendly distro. RMS actually meant all Linux distros. There's no real way to settle this, since most anything past this is opinon (what level of importance is the kernel experience vs. userland, is GNU/ an "advertising clause" that pissed Stallman off about BSD, the percentage of non-GNU tools and the lessening importance of the GNU command line tools, yadda yadda) and it just becomes a FlameWar. Linus is usually pragmatic about it, and he usually stays out of the whole naming mess. He even doesn't mind that there's a Linux variant that doesn't even bear the Linux name (the name escapes me) yet uses the Linux kernel. Evidently having GNU credit is vital for the safety of the world, but having credit for the kernel is somewhat less critical.
In general, the stock FreeBSD system has only one big chunk of FSF code, and that's GCC (there is a FSF soft-FPU available, but since that only is important to 386 and FPU-less 486 users, its almost never used). The userland tools are, unsurprisingly, BSD variants. There are a lot of FSF userland tools available in ports, but the base system is BSD. There's a few folks who are license zealots that talk about suporting a non-GPL compiler in some of the BSDs, but most folks see that as a pisswar and a waste of effort, especially since significant chunks of the base system in now C++ and writing an ANSI and GCC compatible compiler would be REAL hard.
There actually is a GNU/NetBSD variant, with the NetBSD kernel and a FSF/GNU userland. My personal opinion is this is just a proof of concept to show that its the userland more than the kernel that's important, somewhat of a self-congratulatory exercise with no real use. The strength of the BSD systems are that they are cohesive, with a single source. Adding a lot of GNU tools all with different releases, all requiring different downloads kind of destroys the current BSD gestalt.
This is the first time I read this, but it kind of pisses me off (though I really am not trying to start a flamewar).
The BSD developers were inspired to make their code free software by the example of the GNU Project, and explicit appeals from GNU activists helped persuade them,....
This is only true if you believe in the RMS/FSF conceit that only things defined by the FSF are truly free. BSD was releasing code to Universities the only places their AT&T license would let them) for many years before stallman even started at MIT. They spent the great effort to rewrite all of BSD (to get rid of AT&T encumbered code) and release it to the world. They even went to court (the Great Lawsuit, which even Linus admits he probably wouldn't have written Linux if FreeBSD wasn't stuck in legal crosshairs) to allow people to use BSD code. Just because he had a problem with the original license (ironically, because of the BSD License's advertising clause, yet he insists on GNU/Linux, a sort of GNU advertising clause) doesn't mean it wasn't free software.
I think Stallman has done a lot for computing, but as a zealot and a purist, he tends to focus only on his agenda, and tends to be a bit revisionist for things that don't follow his vision. I think this, and even the whole GNU/Linux naming thing show that.
I did some small work on Project UDI, the Uniform Driver Interface. It attempts to make a consistent DDI across all platforms. It aims to make device drivers source compatible across all platforms, and binary compatible across architectures with the same C ABI. Pretty slick, you don't need to worry about synchronization primitives, the environment handles all that for you, giving you ways of handling interupts, getting memory... etc.
Caldera was one of the big supporters actually, back in the "lets, Idunno, actually ENGINEER something that people would want to buy" instead of getting the lawyers involved. Never really went anywhere unfortunately. The big OSes don't need it because people already make drivers for them. The smaller ones tended to have philosophical differences (RMS hated it for Linux, made it easier for binary only drivers he thought). I'ts been pretty much dead since 2001, being in the odd place of having drivers but no OSes.
Slashdot Mirror
Hmm, didn't Microsoft's lawyers patent innovation? Or will this, since it's made with GPLed software that obviously (by Bill's standards) hinders innovation, force a rip in the space time continuum and send us all to an alternate Universe where up is down, down is up, cats and dogs live together, and Microsoft actually invents new technology instead of buying it?
Wouldn't "regenerative", like regenerative braking on most electrics/hybrids been a better term?
Wasn't the XBox supposed to crush Sony like a grape because it used commodity parts while silly Sony used specialized ones, therefore much more expensive?
SCO is getting worse than the crazy homeless people in San Francisco that scream Bible passages at you as you're walking by..
Is the weird asian guy with the sunglasses and signs still walking Market? I took a pic of him once, and he spooked, I think he thought I was part of some Clintonian conspiracy. Was glad to be part of his tinfoil-hat paranoia for a day.
Ahh, the "hooker trading cards" as one friend called them.
There's a reason for them handing them out in the weird way they do, silently just holding them out. I think the rule in Vegas is if someone just "picks them up" it's not illegal. But if they say anything and try to entice you into grabbing any, it's solicitation.
Benefits to users and to the Debian Project include
<FLAMMABLE>Benefits to RMS include stroking his ego (it's really the GNU in GNU/Linux that matter, trust me!!) even if it comes at the expense (to the user) of what makes the *BSDs great - their single source cohesiveness.</FLAMMABLE>
Unfortunately, utilization efficiency is bounded -- it's hard to use more than 100% of your allocated IP addresses.
The article actually lists two technologies to do just this; NAT and DHCP.
It's not cross referencing between Linux and FreeBSD source, it's just a cross reference within the software itself. Fr'instance, I have a driver, and it talks about some function in the DDK, you click on a link and it takes you there. Very confusing title I admit.
Closable BSDs? Yes someone could make a fork of FreeBSD and close it, but the original FreeBSD source would still exist.
1) Not a single vendor ban, they want to raise OpenSource, not kill Microsoft. Since Microsoft is a monopoly, it just seems like a single vendor ban.
2) Essential becuase of network effects and inertia. If Microsoft and Linux have the same effective purchase cost (because of piracy) why would you overcome the inertia to switch?
3) Necessary because of trade regulations that require piracy numbers to go lower. No license to purchase, no piracy. Thats why it needs to be a law.
Is this Calvin's new G.R.O.S.S?
Get Rid Of Slimey non-hackerS?
Publicly accessible doesn't mean it's in Public Domain. Public Domain has a very specific meaning in copyright.
GPL is copyright (err, copyleft) which puts constraints on how you can use the code. Public Domain means no copyright at all, do anything you want to it whatsover.
No one looks forward to going to Court unless they are sure they are going to win HUGE.
But most folks say they look forward to it, need the poker face whether or not you think you'll win.
Going to court is a PITA, no matter how strong you think your case is. Lawyers cost. Getting info together costs. Pulling your developer off a project so he can answer some bullshit about JFS when it wasn't even first developed for AIX or Linux costs.
this has been (kinda) answered by others, but I'll try to make mine a bit clearer.
/. community all over them like they are Microsoft?
The above statement is talking relative to release of 4.8, as cut to CD. This doesn't mean it hasn't been fixed, it just means they can't go back in time to fix it on 4.8 as it was on release day (think what was cut to CD).
That said, FreeBSD users don't have to stay on the "as cut to CD version". Once you get a release, a good FreeBSD user can update his system, tracking one of a few cvs branches, such as STABLE (which will get you this whole release), the 4.8-RELENG "security fixes only" branch, or CURRENT, which would put you in 5.x world. All security problems are fixed in all releases.
If this is the case, why isn't the
Because all critical vulnerabilities are on a mailing list, all versions of FreeBSD affected are updated, even old ones (you'll occasionally see updates for 3.x and even 2.x sometimes) instead of forcing people to upgrade off NT 4 so they can sell more XP licenses.
The APIs for me seem hacked together. Look at all the filename overloads for OpenFile, yet it still can't open a named pipe. How are they consistent? UNIX got this right (though many things wrong in other areas) by realizing the kernel should know what you should do with the open, and not have to hack the API, adding kludges to code and have to explain these special cases in the documentation. Then all of the API calls with named ...Ex, because they didn't get the first implementation right.
In some ways, this is not me being harsh at microsoft. Linux has a lot less baggage because of it's youth, and a lot of things are inconsistent in the kernel DDI, and the windowing libraries tend to change often. But the system apis are stable becuase they are all based on standards.
=) Damn, can't post and mod in the same story.
Reasonably consistent consistent APIs across products
Talk to the Samba guys about how inconsistent they are about protocols. They are a huge company, and many things are inconsistent. You do raise valid points tho, and many in the OS community don't want to hear anything negative.
There was a MS funded benchmark a while back, where Windows came out on top of Linux when it came to webserver performance. The great sea of Slashdotters were up in arms, They shilled for MS!!! A few people actually decided to think "maybe they're right" and looked for improvements in Linux networking code. And Linux got faster, and has beat beating Windows IIS's ass ever since. There are advantages to listening to bad news sometimes.
If C-level players did an Open Source project, nobody would ever see it.
Cough cough Sourceforge cough cough...
So much stuff there is untouched by human hands its incredible.
We do this - they backported a bunch of things from Linux 2.6 into it. The one thing we needed was the 4G user/4G kernel option. We were porting some things from solaris and they couldn't allocate enough memory with the 3G kernel split. Works pretty well.
Patched much more frequently.
Serial ports may lessen in importance, but they can never be completely replaced by USB. Certain RT events require hard timing, which USB can never provide. My Brother In Law needed to buy a PCMCIA serial port card for his laptop to talk to a ladder programmer controller.
They may disappear as defaults from machines though, like the floppy has been eliminated as standard from Macs and now Dells.
I thought this was the gcc development team? Then they got their butts kicked by egcs so badly it became the new gcc maintaining team.
Rule of Parsimony: Write a big program only when it is clear by demonstration that nothing else will do.
.begin FLAMEWAR
ahem, cough cough emacs cough cough
.end
Linus wrote the kernel, unsatisfied with Minix and its licensing restrictions, and really wanting to become a tinkerer. He never really cared about userland and he took the freely available GNU tools and got his kernel running. Because a huge part (essentially all) of the early Linux userland was GNU tools, RMS felt that the FSF was justified for taking part credit for Linux. When Stallman first proposed GNU/Linux, Linus thought he was talking about Debian Linux, which was at that time the most FSF friendly distro. RMS actually meant all Linux distros. There's no real way to settle this, since most anything past this is opinon (what level of importance is the kernel experience vs. userland, is GNU/ an "advertising clause" that pissed Stallman off about BSD, the percentage of non-GNU tools and the lessening importance of the GNU command line tools, yadda yadda) and it just becomes a FlameWar. Linus is usually pragmatic about it, and he usually stays out of the whole naming mess. He even doesn't mind that there's a Linux variant that doesn't even bear the Linux name (the name escapes me) yet uses the Linux kernel. Evidently having GNU credit is vital for the safety of the world, but having credit for the kernel is somewhat less critical.
In general, the stock FreeBSD system has only one big chunk of FSF code, and that's GCC (there is a FSF soft-FPU available, but since that only is important to 386 and FPU-less 486 users, its almost never used). The userland tools are, unsurprisingly, BSD variants. There are a lot of FSF userland tools available in ports, but the base system is BSD. There's a few folks who are license zealots that talk about suporting a non-GPL compiler in some of the BSDs, but most folks see that as a pisswar and a waste of effort, especially since significant chunks of the base system in now C++ and writing an ANSI and GCC compatible compiler would be REAL hard.
There actually is a GNU/NetBSD variant, with the NetBSD kernel and a FSF/GNU userland. My personal opinion is this is just a proof of concept to show that its the userland more than the kernel that's important, somewhat of a self-congratulatory exercise with no real use. The strength of the BSD systems are that they are cohesive, with a single source. Adding a lot of GNU tools all with different releases, all requiring different downloads kind of destroys the current BSD gestalt.
This is the first time I read this, but it kind of pisses me off (though I really am not trying to start a flamewar).
....
The BSD developers were inspired to make their code free software by the example of the GNU Project, and explicit appeals from GNU activists helped persuade them,
This is only true if you believe in the RMS/FSF conceit that only things defined by the FSF are truly free. BSD was releasing code to Universities the only places their AT&T license would let them) for many years before stallman even started at MIT. They spent the great effort to rewrite all of BSD (to get rid of AT&T encumbered code) and release it to the world. They even went to court (the Great Lawsuit, which even Linus admits he probably wouldn't have written Linux if FreeBSD wasn't stuck in legal crosshairs) to allow people to use BSD code. Just because he had a problem with the original license (ironically, because of the BSD License's advertising clause, yet he insists on GNU/Linux, a sort of GNU advertising clause) doesn't mean it wasn't free software.
I think Stallman has done a lot for computing, but as a zealot and a purist, he tends to focus only on his agenda, and tends to be a bit revisionist for things that don't follow his vision. I think this, and even the whole GNU/Linux naming thing show that.
I did some small work on Project UDI, the Uniform Driver Interface. It attempts to make a consistent DDI across all platforms. It aims to make device drivers source compatible across all platforms, and binary compatible across architectures with the same C ABI. Pretty slick, you don't need to worry about synchronization primitives, the environment handles all that for you, giving you ways of handling interupts, getting memory... etc.
Caldera was one of the big supporters actually, back in the "lets, Idunno, actually ENGINEER something that people would want to buy" instead of getting the lawyers involved. Never really went anywhere unfortunately. The big OSes don't need it because people already make drivers for them. The smaller ones tended to have philosophical differences (RMS hated it for Linux, made it easier for binary only drivers he thought). I'ts been pretty much dead since 2001, being in the odd place of having drivers but no OSes.