Small point, but NT is most definitely not available for Sparc (thank GOD). There is the Alpha port, but that's as non-x86 as any M$ operating system gets.
The only OS that'll run on anything you want is NetBSD. (but Linux is catching up fast!)
Playing devil's advocate here, but I was wondering . . .
Assuming that implementing a secure-music system in hardware is completely unfeasible, it might be possible to devise an audio processing scheme that would subvert the next logical step after/dev/audio redirection, i.e. MP3 compression.
It seems plausible that a secure-music system could add certain artifacts to raw digital audio which would wreak havoc with perceptual encoding schemes. So the raw data obtained from/dev/audio would sound all right, but the finished MP3 would be hopelessly mangled. Macrovision for MPEG, if you will. (might even be doable with VQF, AAC/MP4 et. al.)
Anyway, this is observing how some CD passages can point out annoying idiosyncracies of the L3 codec. Anyone know if such "flukes" could be engineered in this way?
Well, small confession here, but I never did notice GNU was in there:-] Gotta give 'em big points for subtlety... too bad they can't stick with such a clever name.
P.S.: I was going to suggest OpenWorks, something with some built-in allegory <g> But alas, it's taken. (You know, I worry sometimes, when it takes a good fifteen minutes to think up of something that isn't already in the InterNIC...)
First off, let me say that I'm not here to flame. Your case looks to be one in favor of the patent system; of course, there are others that slight it equally well. IMHO, I think a change is needed for a more equitable/sensible system, but the jury's still out on what that would be.
My question is, from what you know, how easy/hard was it for the average Joe with a disassembler and some time to kill to decompile your product, to the point where he could make it into another?
For OSS-goers, binary code is assumed to be a black box, but as your case shows, it isn't quite as black as it might be.
More to the point, would it be possible to build some sort of compiler add-on which would, in a sense, obfuscate the compiled code? i.e. add jumps all over the place, encrypt whole blocks of instructions, etc. etc., in a way that would make binary disassembly more difficult by several orders of magnitude? to the point of sheer impossibility?
(in a way that would not adversely affect performance, of course. I think it can be assumed that a compiler-level tool could generate better spaghetti code than a human programmer).
If binaries can really be made as impenetrable as everyone assumes they are, it would then be sufficient for an enterprising company not to publish their algorithms, nor other details of their implementation. At least then IP laws could regain the "independent re-creation" provision (as similar to that of copyright) without adverse effect, and all the worry of patents stifling innovation would be laid to rest.
In your case, some joe stole your code, which I can agree is wrong. But if some other joe were to perform similar research as your company has, and completely independently develop his own approach which may by chance have a few similarities to yours, current IP law would be very unkind to him. And that, I think, is also wrong.
I haven't followed the Blackdown JDK proceedings too closely, but wasn't there an issue with the Sun SCSL that disallowed any release prior to passing the full test suite?
(Someone posted here not long ago, re Sun's semi-open licensing: "...if they would change the SCSL to allow Blackdown to make beta releases..." or WTTE)
I hope you're not looking to use XFree under CGA... cyan background, white xterms...... it'd be just like (dare I say it?) OpenLook!
Heh. All kidding aside (yes, we all know CGA is much nicer than that) the time for such a driver is long past, IMHO. Really, are there any CGA-equipped machines out there that can handle XFree86 to begin with?
(The one 486+CGA I know of is console-only. X? Not a chance)
Might want to mention the new command convention in there... with the package I have (PGP 5.0), instead of pgp -kblah, pgp -vblah, it's pgpk -blah, pgpv -blah, etc.
Still, though, very helpful! We need docs like this to get strong crypto to the masses. Now if only there were as good a HOWTO for pgp4pine....
The difference, IMHO, is that this can bring digital [node] identification into the mainstream. Sure, SGIs, Suns, NICs, etc. all numbered, but how many advertisers/data miners/consumer-level programs (read: Windows apps) have made use of those? If the PIII does catch on, it stands to change the rules of the game.
I really wonder what AMD et. al. are going to do. It could go one way ("we don't make insecure CPUs") or the other ("AMD: Because We're Not Your Big Brother(tm)"). It could be a big opportunity, esp. with the PIII's unimpressive performance specs.
More importantly, it would mean no major software vendor (hint: Redmond) could explicitly require a CPUID for their products to run, without (quoting some poster above) alienating a large portion of the market. As long as we have the choice, ah?
Well, though I can say little about crashing, before Mozilla 1.0 is released, some time ago I too became dissatisfied with the rather lousy X standard fonts just trying to become involved in HTML page rendering.
So what did I do? I converted a bunch of TTF's, and to make a long story short, NS/X11 on my workstation looks reeeeeeally nice.
Check out this page, toward the bottom end, for a tarball of.bdf's and some before/after screenshots. One of them's the ZDNet home page; you've got to admit, it's a noticeable difference }:)
Hope it can make non-W98 browsing a bit more palatable....
This thing loaded up in less than a second. Scrolls smooth and nice too. This is stress?
(for Netscape 4.08, anyway. BTW, that is some pretty nifty-looking HTML)
IMO, if you really want to load down a browser, go to freshmeat. I love how NS chokes up for ~5 seconds "thinking" over the page. Or better yet, load up any page (dern if I could remember one now) that has a multi-column full-width table with 500+ rows. Use your thumbwheel, too. Doth I hear a hard drive page?:-)
> They will do the same to Microsoft Linux. It's coming. Believe it.
BG's ego alone would never allow that. This is the man who chewed out the higher-minded managers within his own company that could see the writing on the wall (the eventual decline of Windows). "Microsoft Linux" would basically mean swallowing every ounce of pride he has. (*choke*) It would mean he has lost the war, and even more, his ideology-- "There is only Windows."
If I understand it correctly, the whole Free-PC concept is based on a border surrounding an 800x600 user area on the screen, right? (whereas the entire thing is 1024x768). And ads are displayed in this border area, right?
So-- instead of cracking the system, reformatting the HD, booting off a floppy, etc.-- couldn't you just put a paper cutout over the screen, and literally block out the advertising?
(Well, granted, it still won't be running our favorite OS, but that does kind of short out the entire principle-- doesn't it?)
At "60,000 times the speed of a home PC," this thing should be able to raytrace in REALTIME (30 fps) what would take my humble K6-2 nearly half an hour per frame.
*shudder*
All we would need then is a neural VR interface, and then things could get really interesting...
Slashdot Mirror
"A Computer On Every Desktop (Running Windows)"
Small point, but NT is most definitely not available for Sparc (thank GOD). There is the Alpha port, but that's as non-x86 as any M$ operating system gets.
The only OS that'll run on anything you want is NetBSD. (but Linux is catching up fast!)
Playing devil's advocate here, but I was wondering . . .
/dev/audio redirection, i.e. MP3 compression.
/dev/audio would sound all right, but the finished MP3 would be hopelessly mangled. Macrovision for MPEG, if you will. (might even be doable with VQF, AAC/MP4 et. al.)
Assuming that implementing a secure-music system in hardware is completely unfeasible, it might be possible to devise an audio processing scheme that would subvert the next logical step after
It seems plausible that a secure-music system could add certain artifacts to raw digital audio which would wreak havoc with perceptual encoding schemes. So the raw data obtained from
Anyway, this is observing how some CD passages can point out annoying idiosyncracies of the L3 codec. Anyone know if such "flukes" could be engineered in this way?
Well, small confession here, but I never did notice GNU was in there :-] Gotta give 'em big points for subtlety... too bad they can't stick with such a clever name.
P.S.: I was going to suggest OpenWorks, something with some built-in allegory <g> But alas, it's taken. (You know, I worry sometimes, when it takes a good fifteen minutes to think up of something that isn't already in the InterNIC...)
First off, let me say that I'm not here to flame. Your case looks to be one in favor of the patent system; of course, there are others that slight it equally well. IMHO, I think a change is needed for a more equitable/sensible system, but the jury's still out on what that would be.
My question is, from what you know, how easy/hard was it for the average Joe with a disassembler and some time to kill to decompile your product, to the point where he could make it into another?
For OSS-goers, binary code is assumed to be a black box, but as your case shows, it isn't quite as black as it might be.
More to the point, would it be possible to build some sort of compiler add-on which would, in a sense, obfuscate the compiled code? i.e. add jumps all over the place, encrypt whole blocks of instructions, etc. etc., in a way that would make binary disassembly more difficult by several orders of magnitude? to the point of sheer impossibility?
(in a way that would not adversely affect performance, of course. I think it can be assumed that a compiler-level tool could generate better spaghetti code than a human programmer).
If binaries can really be made as impenetrable as everyone assumes they are, it would then be sufficient for an enterprising company not to publish their algorithms, nor other details of their implementation. At least then IP laws could regain the "independent re-creation" provision (as similar to that of copyright) without adverse effect, and all the worry of patents stifling innovation would be laid to rest.
In your case, some joe stole your code, which I can agree is wrong. But if some other joe were to perform similar research as your company has, and completely independently develop his own approach which may by chance have a few similarities to yours, current IP law would be very unkind to him. And that, I think, is also wrong.
We need a change.
I haven't followed the Blackdown JDK proceedings too closely, but wasn't there an issue with the Sun SCSL that disallowed any release prior to passing the full test suite?
(Someone posted here not long ago, re Sun's semi-open licensing: "...if they would change the SCSL to allow Blackdown to make beta releases..." or WTTE)
I resent that!
:-)
(j/k
Let's see what the PHB's have to say NOW!
:'-)
Gosh. It's only Tuesday, and this is the second high I've had this week. (first was SGI going open-source)
Our little OS is growing up so fast... *snif*
I hope you're not looking to use XFree under CGA... cyan background, white xterms...... it'd be just like (dare I say it?) OpenLook!
Heh. All kidding aside (yes, we all know CGA is much nicer than that) the time for such a driver is long past, IMHO. Really, are there any CGA-equipped machines out there that can handle XFree86 to begin with?
(The one 486+CGA I know of is console-only. X? Not a chance)
Might want to mention the new command convention in there... with the package I have (PGP 5.0), instead of pgp -kblah, pgp -vblah, it's pgpk -blah, pgpv -blah, etc.
Still, though, very helpful! We need docs like this to get strong crypto to the masses. Now if only there were as good a HOWTO for pgp4pine....
Most cool! Now the linux.com site might finally get a decent webmaster!
;)
Oh, and that final quote from the article is golden
My memory might be off here, but don't Hummers normally sell for over $700,000?
(Don't forget-- built by Lamborghini)
(IIRC)
The difference, IMHO, is that this can bring digital [node] identification into the mainstream. Sure, SGIs, Suns, NICs, etc. all numbered, but how many advertisers/data miners/consumer-level programs (read: Windows apps) have made use of those? If the PIII does catch on, it stands to change the rules of the game.
I really wonder what AMD et. al. are going to do. It could go one way ("we don't make insecure CPUs") or the other ("AMD: Because We're Not Your Big Brother(tm)"). It could be a big opportunity, esp. with the PIII's unimpressive performance specs.
More importantly, it would mean no major software vendor (hint: Redmond) could explicitly require a CPUID for their products to run, without (quoting some poster above) alienating a large portion of the market. As long as we have the choice, ah?
Well, though I can say little about crashing, before Mozilla 1.0 is released, some time ago I too became dissatisfied with the rather lousy X standard fonts just trying to become involved in HTML page rendering.
.bdf's and some before/after screenshots. One of them's the ZDNet home page; you've got to admit, it's a noticeable difference }:)
So what did I do? I converted a bunch of TTF's, and to make a long story short, NS/X11 on my workstation looks reeeeeeally nice.
Check out this page, toward the bottom end, for a tarball of
Hope it can make non-W98 browsing a bit more palatable....
This thing loaded up in less than a second. Scrolls smooth and nice too. This is stress?
:-)
(for Netscape 4.08, anyway. BTW, that is some pretty nifty-looking HTML)
IMO, if you really want to load down a browser, go to freshmeat. I love how NS chokes up for ~5 seconds "thinking" over the page. Or better yet, load up any page (dern if I could remember one now) that has a multi-column full-width table with 500+ rows. Use your thumbwheel, too. Doth I hear a hard drive page?
> They will do the same to Microsoft Linux. It's coming. Believe it.
BG's ego alone would never allow that. This is the man who chewed out the higher-minded managers within his own company that could see the writing on the wall (the eventual decline of Windows). "Microsoft Linux" would basically mean swallowing every ounce of pride he has. (*choke*) It would mean he has lost the war, and even more, his ideology-- "There is only Windows."
Say, here's a thought...
If I understand it correctly, the whole Free-PC concept is based on a border surrounding an 800x600 user area on the screen, right? (whereas the entire thing is 1024x768). And ads are displayed in this border area, right?
So-- instead of cracking the system, reformatting the HD, booting off a floppy, etc.-- couldn't you just put a paper cutout over the screen, and literally block out the advertising?
(Well, granted, it still won't be running our favorite OS, but that does kind of short out the entire principle-- doesn't it?)
At "60,000 times the speed of a home PC," this thing should be able to raytrace in REALTIME (30 fps) what would take my humble K6-2 nearly half an hour per frame.
*shudder*
All we would need then is a neural VR interface, and then things could get really interesting...