>> 1) A Linux kernel is used to boot the VMWare kernel. Yes, through a binary blob loaded by a driver, that does not run on any kernel other than Linux.
>> 2) This makes the VMWare kernel a "derived work." According to Linux logic, and Christoph Hellwig, yes.
>> 3) The Linux kernel source used to boot the VMWare kernel is not available. Er no. You made that up. The article doesn't say that anywhere.
The issue is that the source for vmkernel is not available.
>> 4) Therefore we can't tell if the VMWare kernel is a "derived work" or not. Er no. You made that up too. The article doesn't say that anywhere.
The vmkernel would be considered derived unless it has been ported from another Operating System (proving it does no requite Linux to function). VMware could prove this without giving away the code.
>> 5) And if the Linux kernel source for the kernel used to boot the VMWare kernel is not available, it's a violation of the GPL on its own. Er no. You made that up again. The article doesn't say that anywhere.
The source for the Linux kernel used to start VMware is available from the VMware website. The source for vmkernel is not.
If you need to lie to prove your point, you clearly don't have one.
>> VMkernel does not run on Linux kernel. >> Please, let me reiterate this again: VMKernel DOES NOT RUN ON Linux.
RTFA. According to everyone, including VMware, vmkernel is started from Linux.
The article has been updated to include a video of an ESX machine booting.
Run 'strings' on the 'vmnix' kernel on your machine (or just watch it boot). It's Linux. Nobody hides this fact. As TFA mentions, vmkernel is started by S90vmware, which insmods vmkmod to load it.
"The vmkernel itself, the underlying OS running everything and managing hardware access, is proprietary, and is not Linux. "
I should point out that during boot time this isn't true (which you, and everyone else who's taken a VMware class, seen ESX boot, probably realize). During boot time, the primary OS is Linux, which is used to start vmkernel (ESX) which is then the primary OS and hypervises its original parent.
"Couldn't they start with a Linux kernel, write a couple of binary blob drivers, and do everything else in userland with their own binary to use instead of init?"
This is exactly what ESX does.
"Would it be based on Linux? Yes."
The blob doesn't necessarily contain Linux code, but according to Linus' logic and Hellwig's complaints is considered a derivative work on the Linux kernel on which it relies.
"The SC runs a modified RHEL 3, and functions on bootup as a bootloader for the vmkernel. Once the vmkernel is loaded, the vmkernel handles all hardware access and virtualization functions, and is a completely separate OS from the service console."
He doesn't have to take a VMware class to understand that, just read the article.
The copyright issue is with the method used to load vmkernel from the Linux kernel (or 'bootload' it as you say).
The article goes into very specific detail about how Linux is used to start vmkernel, and exactly how that related to Linus' stance on licensing for binary blobs, as well as Hellwig's complaints to VMware.
"Dell apparently will be booting a version of ESX from BIOS."
This doesn't prove anything, as Linux would still be used to start vmkernel (unless ESX used in the article was especially modified - which wasn't stated in the article you linked to).
Kexec is quite different in its execution. VMware specifically use a hook to a binary blob, loaded as a module, which Linux has specifically stated he regards as a derived work (unless vmkernel can run on platforms other than Linux - how much of it lives in VMware's other products is know only to VMware).
I really hate that "upper left corner" style of Michael's...where's the rest of the fucking box? Is there some kind of bit shortage where we need to conserve every last byte in transmission? Seriously though, the corner is little more than fluff and does little to enhance readability or make the articles "pop". 93 thumbs down.
Michael's design is the only one Ive seen that doesn't overwhelm the readers with massive chunks of dark green everwhere. There's a header, with enough identity to make it feel like Slashdot, and the articles. No massive shapes drawing your attention away from what you're trying to read. I know those side items are there thank you very much, I don't want them waved in my face.
Jason's design is like using Microsoft Office on a PC. Lots of action, lots of functionality in 150 buttons. Michael's design is like using Pages on a Mac. All the stuff you need is there, but it seems easier as it's not waving every single thing possible in your face.
All of those cartoons are criticisms of political policies of Israel, not the Jewish race or religion. Yes, they draw Jews with big noses. Cartoonists overemphasize everyone's features, including ethnic ones. Asian people have distinct eyes, black people are very black, white people are pale as ghosts, Australian aboriginals wear paint. It's not to make fun of the people, it's to establish their identity in a clear and simple way. Find a cartoonist who doesn't do this and get back to me.
It's not as bad as violence, but calling everybody who criticises Israel racist is a pretty damn pathetic way of silencing those you disagree with.
> Why not? Production machines need to be able to mail their owners about problems. Desktops need to be able to send mail. Both might just not be Windoze machines able to talk to your crappy, virused out Exchange "server".
That's a pretty unusual exchange server you're talking about there.
> Not accepting SMTP requests from desktops is just another workaround to M$'s really shitty security that won't work.
Then why would any well-run Unix shop also use mailhub? Why do Unix MSPs implement that functionality? Why does every well-known figure in Unix mail recommend using that functionality for this purpose?
Digg does spell checking (and doesn't add one persons opinion to every article, and doesn't ban users for modding up controversial articles). It also, as of a few months ago, gets more traffic than Slashdot.
So no, Taco, I don't think people would find something else to complain about.
(I'm reading Slashdot for the first time in about a month, having switched to digg. Before that, I read slashdot religiously every day since 1999).
Slashdot Mirror
Apparently the argument goes as follows:
>> 1) A Linux kernel is used to boot the VMWare kernel.
Yes, through a binary blob loaded by a driver, that does not run on any kernel other than Linux.
>> 2) This makes the VMWare kernel a "derived work."
According to Linux logic, and Christoph Hellwig, yes.
>> 3) The Linux kernel source used to boot the VMWare kernel is not available.
Er no. You made that up. The article doesn't say that anywhere.
The issue is that the source for vmkernel is not available.
>> 4) Therefore we can't tell if the VMWare kernel is a "derived work" or not.
Er no. You made that up too. The article doesn't say that anywhere.
The vmkernel would be considered derived unless it has been ported from another Operating System (proving it does no requite Linux to function). VMware could prove this without giving away the code.
>> 5) And if the Linux kernel source for the kernel used to boot the VMWare kernel is not available, it's a violation of the GPL on its own.
Er no. You made that up again. The article doesn't say that anywhere.
The source for the Linux kernel used to start VMware is available from the VMware website. The source for vmkernel is not.
If you need to lie to prove your point, you clearly don't have one.
>> VMkernel does not run on Linux kernel.
>> Please, let me reiterate this again: VMKernel DOES NOT RUN ON Linux.
RTFA. According to everyone, including VMware, vmkernel is started from Linux.
The article has been updated to include a video of an ESX machine booting.
Run 'strings' on the 'vmnix' kernel on your machine (or just watch it boot). It's Linux. Nobody hides this fact.
As TFA mentions, vmkernel is started by S90vmware, which insmods vmkmod to load it.
The links are not to the source for vmkernel, which is what Hellwig's statements and Linus' logic say are considered to be derived from Linux.
*Sigh*. VMware distribute GPL software, and provide the source for that.
/etc/rc3.d/S90vmware on an ESX box to see how vmkernel is started by Linux.
They do not provide the source to the binary blob loaded by their Linux driver when the ESX product starts.
Look in
This is what Hellwig is complaining about, and this is the kind of software Torvalds objects to.
"The vmkernel itself, the underlying OS running everything and managing hardware access, is proprietary, and is not Linux. "
I should point out that during boot time this isn't true (which you, and everyone else who's taken a VMware class, seen ESX boot, probably realize). During boot time, the primary OS is Linux, which is used to start vmkernel (ESX) which is then the primary OS and hypervises its original parent.
"Couldn't they start with a Linux kernel, write a couple of binary blob drivers, and do everything else in userland with their own binary to use instead of init?"
This is exactly what ESX does.
"Would it be based on Linux? Yes."
The blob doesn't necessarily contain Linux code, but according to Linus' logic and Hellwig's complaints is considered a derivative work on the Linux kernel on which it relies.
"The SC runs a modified RHEL 3, and functions on bootup as a bootloader for the vmkernel. Once the vmkernel is loaded, the vmkernel handles all hardware access and virtualization functions, and is a completely separate OS from the service console."
He doesn't have to take a VMware class to understand that, just read the article.
The copyright issue is with the method used to load vmkernel from the Linux kernel (or 'bootload' it as you say).
The Linux Kernel starts ESX, ESX re-virtualizes Linux (according to both VMware and the article, and shown in the video attached to the article)
The article goes into very specific detail about how Linux is used to start vmkernel, and exactly how that related to Linus' stance on licensing for binary blobs, as well as Hellwig's complaints to VMware.
Could you provide technical information? Many companies have great ethics, that doesn't prove whether or not they're violating a software license.
Is it possible to run the binary vmkernel blob ESX loads from the Linux kernel, without using Linux to load it?
"Dell apparently will be booting a version of ESX from BIOS."
This doesn't prove anything, as Linux would still be used to start vmkernel (unless ESX used in the article was especially modified - which wasn't stated in the article you linked to).
There's a video of this embedded in the article.
Kexec is quite different in its execution. VMware specifically use a hook to a binary blob, loaded as a module, which Linux has specifically stated he regards as a derived work (unless vmkernel can run on platforms other than Linux - how much of it lives in VMware's other products is know only to VMware).
>> Well let's ignore the "i.e." that I don't think Torvalds actually spelled out and read what this really is saying...
e rsonally+consider+anything+%22&ie=utf-8&oe=utf-8&a q=t&rls=com.ubuntu:en-US:official&client=firefox-a
Actually, that's exactly what he said:
http://www.google.com.au/search?q=torvalds+%22I+p
> Uhh, this is a virtualization system. The ESX kernel provides a hardware abstraction layer which the linux kernel in the service console can access.
Read TFA.
ESX is initially loaded by a module hook by the Linux kernel (later, it virtualizes Linux).
That's the part that's a problem.
As others have mentioned, there's no reason that it should look like this by default.
The package for Windows should use GTK WIMP. Here's what GTK WIMP apps look like:
http://gtk-wimp.sourceforge.net/screenshots/
Got bored and reading Slashdot for the first time in ages.
Open Source doesn't mean you can get the source code. If it did, then Windows, QMail, PocketPC, Pine, etc. would be Open Source. They're not.
Open Source is defined by the Open Source Definition.
You've made an excellent point and I agree with all of it.
/lib/modules
But FYI:
because that is 75 systems I have to boot into text mode, rebuild the Nvidia drivers, rebuild the sata drivers, and reboot back to X windows...
Just compile the driver once, then, on each system:
* copy in into the relevant directory beneath
* run 'depmod -a'
I really hate that "upper left corner" style of Michael's...where's the rest of the fucking box? Is there some kind of bit shortage where we need to conserve every last byte in transmission? Seriously though, the corner is little more than fluff and does little to enhance readability or make the articles "pop". 93 thumbs down.
Michael's design is the only one Ive seen that doesn't overwhelm the readers with massive chunks of dark green everwhere. There's a header, with enough identity to make it feel like Slashdot, and the articles. No massive shapes drawing your attention away from what you're trying to read. I know those side items are there thank you very much, I don't want them waved in my face.
Jason's design is like using Microsoft Office on a PC. Lots of action, lots of functionality in 150 buttons. Michael's design is like using Pages on a Mac. All the stuff you need is there, but it seems easier as it's not waving every single thing possible in your face.
Being forced into a difficult situation IS bullying
Were his parents then bullying him by calling him Ghyslain?
Last week my iPod was a 2G with an 8 hour battery.
Now my new 5G has a 20 hour battery, four times the capacity, video, album artwork, and charges over USB. It was cheaper than my 2G too.
Clearly I'm deluded into thinking this was a significant difference.
> But since the riots in Sydney over the lawful arrest and conviction of some criminals
Huh? The riots happened after some lifeguards were beaten up by a Lebanese gang. Who was rioting about whose arrest?
All of those cartoons are criticisms of political policies of Israel, not the Jewish race or religion. Yes, they draw Jews with big noses. Cartoonists overemphasize everyone's features, including ethnic ones. Asian people have distinct eyes, black people are very black, white people are pale as ghosts, Australian aboriginals wear paint. It's not to make fun of the people, it's to establish their identity in a clear and simple way. Find a cartoonist who doesn't do this and get back to me.
It's not as bad as violence, but calling everybody who criticises Israel racist is a pretty damn pathetic way of silencing those you disagree with.
Subject says all.
> Why not? Production machines need to be able to mail their owners about problems. Desktops need to be able to send mail. Both might just not be Windoze machines able to talk to your crappy, virused out Exchange "server".
That's a pretty unusual exchange server you're talking about there.
> Not accepting SMTP requests from desktops is just another workaround to M$'s really shitty security that won't work.
Then why would any well-run Unix shop also use mailhub? Why do Unix MSPs implement that functionality? Why does every well-known figure in Unix mail recommend using that functionality for this purpose?
Digg does spell checking (and doesn't add one persons opinion to every article, and doesn't ban users for modding up controversial articles). It also, as of a few months ago, gets more traffic than Slashdot.
So no, Taco, I don't think people would find something else to complain about.
(I'm reading Slashdot for the first time in about a month, having switched to digg. Before that, I read slashdot religiously every day since 1999).