Hmm. Does it work on other OSes? I suspect that some monitors have screwed up EDID information. If not, you could file a bug report (though if you're using the proprietary nVidia binary drivers, you might not have much luck with that).
Well, then you appear to be wrong on two counts: 1. X does support reading monitor specifications from the monitor itself (assuming the video driver supports it -- Mine does). 2. DPMS has nothing to do with it. You're probably referring to EDID or DDC.
Do you know anything about how a KVM switch works? X can't detect your monitor's parameters if the monitor isn't connected to the video card when X starts.
You missed the point, being that in order to legitimately issue licenses for the performance of certain actions, you must necessarily have, in law, a general prohibition of the performance of those actions by unlicensed parties, and you must have been granted the authority to issue licenses.
Microsoft is selling "licenses" under the false pretense that those actions are otherwise prohibited by law.
Wow, Microsoft is offering 4 licenses of Windows to run in virtualization. Apparently in Ballmer's world, people need to pay more than once to run as many instances of a program on their computer as they want.
I guess, by that logic, running multiple instances of MS Word, Internet Explorer or any other program on my computer is copyright infringement.
I don't truly know, but I suspect RAID drivers must behave more like SCSI drivers than what we normally think of as a "Windows driver" -- that is, a much more direct hardware interaction than is used with NT's HAL.
I'm not sure what you mean by this -- maybe it's because I'm not familiar with Windows NT driver development -- but aren't all "Windows drivers" the same in that they run in ring 0? Different drivers might use different programming interfaces (much like Linux drivers can all use the PCI interface to access PCI devices), and certain interfaces (like those provided by network or video drivers) might not be available early in the boot process, but I would expect that they would all be running above or at the same level as the NT kernel.
Think about it -- the RAID and SCSI HD drivers would have to load *before* Windows-proper, otherwise how is Windows going to *see* the HD to load the rest of itself?
An interesting point, but I imagine it would work in a similar way as Linux does it with its initial ramdisk: the drivers would be loaded into RAM alongside the kernel by the boot loader (which uses the BIOS interface to access the disk).
Also consider that NT had non-x86 versions, so Microsoft would have had to have solved these problems without resorting to DOS on those architectures. I can't see the advantage of solving the problem in a relatively clean way on those architectures, but using an MSDOS-based hack on x86.
As to the issue of working around other buggy programs, that may happen too (tho more likely a matter of M$'s choice of package prep), but 99% of Windows is never spoken to by other apps *by filename*.
Heh. Maybe it's to keep old 16-bit versions of InstallShield from breaking things.
(Side note: I'm highly suspicious the "M$DOS.SYS must be greater than 1024 bytes" thing is needed only by one of the M$ bootup critters, not by anything else.:)
Is that an NT thing? I remember seeing that on Win9x, but the IO.SYS, MSDOS.SYS, AUTOEXEC.BAT, and CONFIG.SYS files on a fresh install of W2K are all 0 bytes.
BTW at that "F6 to load RAID driver" prompt, you can load any driver or even a program that lies and merely claims to be a driver; frex, NT password breakers.
By your own arguments, wouldn't those programs have to be MSDOS-based, then? I strongly suspect that they are not.
Almost as ridiculous as a system that says "Do this or we'll fine you", then after you take measures that you say bring you into compliance, says "we're going to take a long time to decide if you're really in compliance, and if at the end of that we decide you haven't been, we'll fine you retroactively."
So you think it would be better if the overall fines MS had to pay were less if the company's execs lie than if they tell the truth that they are still not in compliance?
Honestly, it's ridiculous that Microsoft needs to be babysat in the first place.
This already exists in mobile phones. Some phones provide a JVM that lets you run code in a sandbox, but their bootloaders check an RSA signature before executing the operating system, and the operating system checks signatures on the Java classes before giving them privileges. It's how the phone companies get away with charging you an arm and a leg for ring tones and wallpapers.
And I'm telling you that they don't. When we used to install our NT4 servers, we had to press F6 at a certain point in the install process to load the NT RAID drivers from a floppy. How could you load NT drivers if it was DOS that was running?
My guess is that Microsoft uses 8.3 filenames for the same reason that many people avoid spaces in directory and file names on Unix: to avoid triggering bugs in poorly-written programs.
Slashdot Mirror
Hmm. Does it work on other OSes? I suspect that some monitors have screwed up EDID information. If not, you could file a bug report (though if you're using the proprietary nVidia binary drivers, you might not have much luck with that).
AHAHAHA HAHA HAH
... is what I would be saying if I were 12-15 years old.
Well, then you appear to be wrong on two counts: 1. X does support reading monitor specifications from the monitor itself (assuming the video driver supports it -- Mine does). 2. DPMS has nothing to do with it. You're probably referring to EDID or DDC.
We're now a little bit closer to Star Trek-style touch screens...
Never mind. Apparently X supports that too.
Do you know anything about how a KVM switch works? X can't detect your monitor's parameters if the monitor isn't connected to the video card when X starts.
Read the post. He wrote more than 3 buttons.
I'm not so sure, but it should be Microsoft's responsibility to ensure that they're compliant, just as with any other court order.
You missed the point, being that in order to legitimately issue licenses for the performance of certain actions, you must necessarily have, in law, a general prohibition of the performance of those actions by unlicensed parties, and you must have been granted the authority to issue licenses.
Microsoft is selling "licenses" under the false pretense that those actions are otherwise prohibited by law.
It isn't necessarily bad, but it could result in a massive population reduction of other species...
... including Homo sapiens.
Maintaining what has been the status quo for thousands of years is seen as being much less likely to have that kind of effect.
The goggles do nothing!
Way to stick it to The Man!
Oh wait...
Somebody just Vorbis-encoded all my backup tapes!
It wasn't the repository that was compromised...
Wow, Microsoft is offering 4 licenses of Windows to run in virtualization. Apparently in Ballmer's world, people need to pay more than once to run as many instances of a program on their computer as they want.
I guess, by that logic, running multiple instances of MS Word, Internet Explorer or any other program on my computer is copyright infringement.
What a load of crap.
I'm not sure what you mean by this -- maybe it's because I'm not familiar with Windows NT driver development -- but aren't all "Windows drivers" the same in that they run in ring 0? Different drivers might use different programming interfaces (much like Linux drivers can all use the PCI interface to access PCI devices), and certain interfaces (like those provided by network or video drivers) might not be available early in the boot process, but I would expect that they would all be running above or at the same level as the NT kernel.
An interesting point, but I imagine it would work in a similar way as Linux does it with its initial ramdisk: the drivers would be loaded into RAM alongside the kernel by the boot loader (which uses the BIOS interface to access the disk).
Also consider that NT had non-x86 versions, so Microsoft would have had to have solved these problems without resorting to DOS on those architectures. I can't see the advantage of solving the problem in a relatively clean way on those architectures, but using an MSDOS-based hack on x86.
Heh. Maybe it's to keep old 16-bit versions of InstallShield from breaking things.
Is that an NT thing? I remember seeing that on Win9x, but the IO.SYS, MSDOS.SYS, AUTOEXEC.BAT, and CONFIG.SYS files on a fresh install of W2K are all 0 bytes.
By your own arguments, wouldn't those programs have to be MSDOS-based, then? I strongly suspect that they are not.
"I quit slashdot about a month ago"
-- seen on slashdot
Just because proprietary software can be worse doesn't mean that free/open-source software is secure.
Does pointing IE at a website that exploits its bugs and thereby installs rootkits/spyware/other browsers count?
"Lynx/links" is clearer and takes up fewer characters.
No, but TiVo probably can.
So you think it would be better if the overall fines MS had to pay were less if the company's execs lie than if they tell the truth that they are still not in compliance?
Honestly, it's ridiculous that Microsoft needs to be babysat in the first place.
This already exists in mobile phones. Some phones provide a JVM that lets you run code in a sandbox, but their bootloaders check an RSA signature before executing the operating system, and the operating system checks signatures on the Java classes before giving them privileges. It's how the phone companies get away with charging you an arm and a leg for ring tones and wallpapers.
No text.
And I'm telling you that they don't. When we used to install our NT4 servers, we had to press F6 at a certain point in the install process to load the NT RAID drivers from a floppy. How could you load NT drivers if it was DOS that was running?
My guess is that Microsoft uses 8.3 filenames for the same reason that many people avoid spaces in directory and file names on Unix: to avoid triggering bugs in poorly-written programs.