Slashdot Mirror
The Need for Open Hardware
bwt asks: "With all the talk of DRM lately, it occurs to me that the entire concept depends on limiting the choice for computer hardware. OK, so the proper reaction to the copyright industry's attempts at PC market control is to be able to build a PC that they can't control. I know there have been some discussions on open hardware, but most if it was prior to the emergence of DRM as a real threat. In fact, Richard Stallman wrote an editorial in 1999 and said 'Because copying hardware is so hard, the question of whether we're allowed to do it is not vitally important.' DRM has perhaps changed that. Isn't the need for open hardware becoming critical? What is the status of the open hardware efforts?"
It reminds me of the consequences of the book "The Jungle," which led to the mandatory listing of all ingredients of a food on the label.
This would translate into basically letting you know what components of a product you have, but not necessarily how they work, with each other, or with you. And, you're allowed to test and research the product to make sure they aren't lying. With this, at least you'd know if there's DRM hardware in something you purchase. It could be more of a middle ground, and be some sort of comprimise. Sure, i'd rather have open-everything, and if you comprimise a little, they take a lot, but it's just a possibility.
Perhaps we shouldn't be asking whether or not we should develop a new form of hardware to avoid DRM, but what is currently available that's so bloody weird that they'd not bother. NetBSD and Linux run on practicaly anything. If we all started using say, ARM CPUs, reusing old SPARCs, etc, it'd be alot easier and alot cheeper. Who is going to fund a company dedicated to making open, non-DRMed hardware? Next thing you know, as a VC, your being sued and/or prosecuted for facilitating piracy, terrorism, etc.
There is plent of non-Intel(and friends) stuff out there already. Microsoft doesn't controll it in the slightest, and itd be too much of an undertaking for them to do it. I don't think ARM has much to lose from "just saying no" to microsoft.
Isn't capitalism supposed to solve problems like this? Shouldn't companies who offer non-DRM hardware find favour with the consumer, and thus prosper over crippled-ware sellers? Oh wait, I forgot, the governments of the "Western" world are rapidly abdicating their role of legislating against the most abusive excesses of capitalism, in favour of legislation aiding and abetting them... Whoops.
Freedom: "I won't!"
When you look at hardware, the designing isn't the most expensive part, manufacturing is. (just like in software, support is the most expensive part :) So I could see a manufacturing company that was running some ultra cheap process try to make money - but there isn't much there. Plus, you have to do literally months of verification on each design before sending it to fab - I don't think most Open Source projects do that amount of testing...
The reality is that it still costs $1/4 million dollars to send a chip to Fab (rumored to cost a cool million for 0.1 micron). I don't know who is willing to put up that kinda money without some assurance the government isn't going to shoot them down half way through production.
Not living in the USA i couldnt give a crap if you get Fritz'ed or not. Im more worried that the just-as-evil governments around the world will decide they want this too. Whats worse, is that even if they don't, allot of important hardware comes from the US and the hardware companies there might decide that its easier to just make locked products and sell the same thing to everyone rather than have the extra over-head of building to versions of something. You never know, the government may decide that its illigal to even build unlocked devices for export.
DVD, Tivo, and modern games consoles have proven that no-one really cares if they have restricted control of a device in their own home, or if its proprietry. Just as long as they can see pretty colours, and drink their starbucks its all good.
This comment does not represent the views or opinions of the user.
It is actually kind of depressing that even though we were all so well aware of what was coming we are still here, right up against the wall with so little progress to show.
P.S. Yes, I am aware of how the "M$" makes me look :-) the sad thing is I am a lot like that guy, except until I got my well paying IT job it was my parent's garage, not basement.
All the creatures will die, And all the things will be broken. That's the law of samurai. (Jubai, 1605)
I, by no means have exhausted the possible reading available on the subject of DRM, but it seems to me that there's one way to trivially circumvent it - trivial, at least, compared to the prospect of building and running a multi-billion dollar chip fab plant or a multi-million dollar mobo fab plant with volunteers. :)
What if you created a VM that bootstrapped your favourite OS into its own memory space? All the code hardware translations would be minimal, so there's not much overhead there, but the VM environment would require that you boot from a virtual filesystem - and with my experience of using virtual volumes in Mac OS X, there's not much of a performance hit there, either.
It seems to me that from the perspective of DRM hardware, it's seeing one really big piece of software manipulating one really big file. Want to share a file with someone? If that person has the same VM environment, then it's simple - copy the file within that VM environment. What if they don't have the VM environment? Guess we'll have to implement some VM 'hardware' to implement DRM 'our way': just enough to make it easy to copy the file transparently, but not so much as to cripple our use of our own OS.
What makes this idea even more interesting is that the VM and VFS can both be DRM approved, because with respect to the environment we're running in, it's irrevalent.
Are there any holes in this strategy?
Was it DeTocqueville who said something to the effect
that democracy in America would last until the people
realized that they could vote themselves bread?
I guess now we know how long that is. About
150 years from 1783 to 1933.
Or you could argue that the union system broke
down 70 years before that, when Lincoln established
the American empire.
-I like my women like I like my tea: green-
But see, that's exactly the kind of conversation about Palladium that I'm not interested in having.
The Windows operating system is extremely insecure. UNIX is a little better, because security was designed in at a much lower level, but it's still not perfect. Of course, in most environments security simply isn't necessary, but if you need to put a computer on the Internet without benefit of a firewall, it's suddenly very important.
Palladium is one proposal to improve the security of PCs by implementing cryptographic technology at the hardware layer. On its face, it's actually a pretty neat idea. But to see that, you have to think of it as a feature, and not a set of handcuffs.
I just wish we could have meaningful conversations about the pros and cons of the technology proposal itself, without immediately collapsing into "it's about control" and "it's about freedom." Because, contrary to popular Slashdot opinion, it's not always about freedom. Sometimes it's just about technology. Technology-- specifically, trying to be a Monday-morning quarterback on technology matters-- is interesting and fun. Politics is not. Is it too much to ask that Slashdot be a place where we can have conversations about technology that don't always become conversations about politics?
I believe Neal Stephenson made a pretty good argument for "open hardware" being a significant contributor to the existence of linux. He mentions the unlikely trinity of Bill Gates, IBM clones, and Linus as being the combination of things needed for linux to be created. Read it here
There are a lot of parallels between this and the introduction of the PS/2 microchannel architecture. IBM ultimately decided that it was more important to have central control over the bus and be in charge of who is allowed where on the bus and, of course, collecting royalties for allowing people on the bus. This was to recapture the "clone" market that had gone awry.
In the same way, M$ wants to reclaim control over the operating system where it feels threatened by other companies that are offering similar/competing products for cheaper.
In the former case, the clone manufacturers got together and invented EISA as a way to get at the real problem - the fact that users had a nightmarish time trying to configure multiple hardware cards that were pretty inflexible. That worked as a stopgap method until Intel came out with the PCI bus that pretty much eliminated that problem.
So, we have this problem that "unauthorized" code can run on a computer without needing to authenticate with the user. Seems as though we could build some hooks into Linux to control that execution in a similar way to how MS wants to do it, but have it decentralized instead of centralized.
One way to do this is to add an authentication method to the program loader. For example, if you want to run mozilla, the OS checks to make sure you have allowed mozilla to run. That authentication would have an md5sum attached to it, so it would check the md5sum of the executable against what is stored in the authentication. When the program changes (due to virus or upgrade) the md5sum is no longer valid and the kernel would return a message (this could be a loadable kernel module with some API hooks) that the program is not authorized to run. You could have it set to "authorize it" as an option.
The next thing would be to create programs that interface to the module and look up a repository (or repositories) of md5sums of known viruses (virii?), thus you could have the following sequence. Open mail. Click on "trojan.sh" -> Dialog box comes up and says, "The program you are attempting to run is not authorized to run on this computer. Click OK to run, Check to do an online virus scan, Cancel to abort execution." (Check) "The following code is recognized as the "trojan" virus.... blah blah"
So basically, if Linux solves the real problem and doesn't give MS a corner on the market in terms of licensing all new apps that run on a computer, then vendors are going to switch to the operating system that has the best licensing terms. It happened for CP/M. It happened for ISA->EISA. It happened for RAMBUS. Build it and they will come.
Yes. 'People are scum.'
That's why a vanguard elite is needed to show them the folly in their ways and lead them to a victorious revolution.
And clearly you're that vanguard elite. You've got it all figured out, eh?
FOAD.