11.18.09
Vista 7 Trojans Forecast and Microsoft Hardware Licences
Summary: More concrete problems, some of which artificially introduced, in Windows Vista and possibly its successors
LAST WEEK we saw Vista 7 getting cracked and the ramifications are highlighted as follows:
Trojans likely to follow Win 7 activation hack
[...]
Trojan attacks are likely in the wake of the Windows 7 product activation system cracks developed last week, less than a month after the release of Microsoft’s latest operating system.
The reality behind Vista 7 is not a convenient one and as it turns out, based on one of our readers, a “Microsoft hardware licence” is now required in Vista — an antifeature which was probably inherited by Vista 7.
That 32-bit editions of Windows Vista are limited to 4GB is not because of any physical or technical constraint on 32-bit operating systems. The 32-bit editions of Windows Vista all contain code for using physical memory above 4GB. Microsoft just doesn’t license you to use that code.
[...]
For the question of whether 32-bit Windows Vista will use all your physical memory, the hard-coded limit of 4GB is dominant as the maximum address for the ordinary kernel, which truly cannot form addresses for physical memory above 4GB, but the license limit is dominant for the PAE kernel. If you have physical memory above 4GB and wonder how it can be that the PAE kernel does not use that memory, the answer is licensing. The 32-bit code for using memory beyond 4GB is present in Windows Vista as Microsoft supplies it, but Microsoft prepares license values in the registry so that this code never gets to work with any physical addresses above 4GB.
This is ridiculous. There will probably be more coverage of this in days to come, so a followup is likely. Artificial limitations are a mockery as code is infinitely abundant. █
Yuhong Bao said,
November 19, 2009 at 1:00 am
“as it turns out, based on one of our readers, a “Microsoft hardware licence” is now required in Vista”
That is not what this article by Geoff Chappell meant. What Geoff Chappell was talking about is artificial limits. BTW, Geoff Chappell is good at software reverse engineering and discovering undocumented APIs, dating back to MS-DOS (he wrote DOS Internals). Take a look at his work on the Windows Shell, for example:
http://www.geoffchappell.com/viewer.htm?doc=studies/windows/shell/index.htm
“There will probably be more coverage of this in days to come, so a followup is likely.”
It already made slashdot and reddit, and there are already software designed to automatically apply the patch described in this article to remove the artifical limits.
your_friend Reply:
November 24th, 2009 at 11:36 pm
What is the difference between an artificial software restriction you must pay to remove and a hardware license you must purchase to be able to use all of your hardware?
The long term nature of this crime, or that others have done it, does not make the crime less offensive. I saw a link here about NT “server” and “workstation” being identical code with a few “flags” thrown in to reduce functionality of the workstation. A software company that sabotages it’s own code will think nothing of sabotaging other people’s code. The computer’s owner is the ultimate victim. How many examples of user sabotage do people need before they escape to software freedom? Reducing the abilities of Windows is like hanging sandbags on a pig, so that it might not fly as well as it could before.
Yuhong Bao said,
November 19, 2009 at 5:18 pm
“an antifeature which was probably inherited by Vista 7.”
This limit dates all the way back to when PAE support was introduced in Windows 2000, read the article for more details.
“so a followup is likely.”
Unfortunately not:
http://geoffchappell.com/viewer.htm?doc=new/index.htm
Roy Schestowitz Reply:
November 19th, 2009 at 5:24 pm
It ought to be disappointing that Microsoft cannot handle RAM sufficiently well. It won’t make it in HPC any time soon.
Yuhong Bao Reply:
November 19th, 2009 at 5:52 pm
“It ought to be disappointing that Microsoft cannot handle RAM sufficiently well. ”
Except that the 64-bit versions can, all this apply only to the 32-bit versions.
Roy Schestowitz Reply:
November 19th, 2009 at 5:58 pm
In many ways, 64-bit Windows has been weaker than GNU/Linux (Flash Player plugin for example).
Yuhong Bao Reply:
November 20th, 2009 at 11:33 pm
Actually, the 64-bit versions do have artificial limit, it is just that it is always higher than 4 GB. If the Home Premium edition’s limit of 16 GB RAM is not enough, you just go up to the Professional or Business edition, then problem solved.
Roy Schestowitz Reply:
November 21st, 2009 at 5:01 am
Artificial limits are always a bad idea. It is indicative of outdated business models.
Yuhong Bao Reply:
November 24th, 2009 at 8:38 pm
“In many ways, 64-bit Windows has been weaker than GNU/Linux (Flash Player plugin for example).”
64-bit Chrome too. Part of it is because MS imposes stricter requirements on 64-bit code in Windows than on Linux. In particular, 64-bit Windows’ table-based SEH imposes restrictions on function prologs and epilogs, as well as requiring unwind tables to be generated by the JIT or compiler. 64-bit Linux has unwind tables too, but it is less important, since processor exceptions are handled by signals.
Yuhong Bao Reply:
November 27th, 2009 at 6:36 pm
“Artificial limits are always a bad idea. It is indicative of outdated business models. ”
Yea, I agree that artificial scarcity is fundamentally flawed. In fact I mentioned the Open Letter to Hobbyists before. The good thing about web applications is that it’s scarcity is not artificial, as the code is stored on the server and clients cannot directly copy it.
Roy Schestowitz Reply:
November 27th, 2009 at 7:22 pm
Yes, the subscription model of very different. Microsoft tried to emulate it with forced upgrades.
Yuhong Bao Reply:
November 28th, 2009 at 4:43 pm
“Yes, the subscription model of very different. Microsoft tried to emulate it with forced upgrades. ”
Not exactly, but yes MS had tried real subscription models several times. But the difference is more fundamental, with proprietary web applications the code is stored on the server and clients cannot copy it at all, so the scarcity is not artificial. In contrast, with proprietary desktop applications the code is stored on the client, where it can easily be copied or reverse engineered.