The real innovation of Gopher, that continued in the Web browser, was the idea of selecting something (from a menu or from a hyperlink) that specified a host, a port (implied or explicit) and an opaque specifier. That is an incredibly powerful, yet trivially simple, concept.
The two important things the Web introduced were combining that concept with a markup language (as someone already pointed out, SGML was around for a long time before), and using MIME as a type identifier
So, I'd add MIME as an incremental step before the Web browser. Gopher was attempting to migrate to Gopher+, and add more capability of identifying the content type, but it was the weakest part of the protocol. The step of making an index free-form and intermixed with text (and later, images, dropdown menus, controls, etc.) was a fairly small increment.
Don't forget WAIS in their somewhere. Some of the concepts of WAIS, plus things like Veronica, ended up with what various search engines and database access interfaces do now.
(There's an exception that allows *you* to keep doing whatever it is you were doing in this case, but the patent is still valid against everyone else.)
IANAL, but I've spoken with several regarding similar issues. Unless it has changed recently, prior art requires "public use" OR publication. I'm not aware of any such exception, or need for such an exception, could you give me a pointer to the statutory language?
For example, our system uses a specific method of interprocess communication that DEC patented (US patent 4,449,182) and was using as their interface between the OS and peripheral controllers, thus locking out the market for compatible peripherals. Although our code isn't public, our system was in public use, and that was enough to get DEC to drop their suit of a third party. They weren't suing us, so some sort of exemption to allow us to continue using it wouldn't have helped the third party disk manufacturer.
Does DEC still claim that they've never lost a patent case? I guess that's pretty easy to do if you always drop the case before you lose.
BTW, there's a program that is still in existence that was written in 1981 that did on-line voting, including preventing people from voting more than once, allowing anyone to set up a vote. It didn't show percentages, just the raw numbers, and it only offered multiple-choice answers (like Slashdot polls do). The existence of the program in 1981 can be documented, although the file itself was last edited in 1989. I'm not sure how many levels of backups there are, but there are backups from 1991 that contain it, with a relatively easy chain-of-custody that can be established.
In a case involving a bingo program, documentation was provided by a printout of the program from a backup, along with an affidavit describing how backups were done and how we could be sure that the backup was actually from that time period and the dates hadn't been spoofed. An even better method is to print it out, with the date, and sign it. Your testimony later that you personally printed it out and signed it, and that the date was in fact the date when you printed it, should be enough for even the most picky rules of evidence.
It does seem problematic, though. How do you prove that something was well-known and/or obvious in 1975, unless it was a major project run by a company that is still in business? It's going to be just luck to find someone who is an obsessive collector who just happened to hang on to something.
Having said this, some people *have* dunned spammers and collected. See this post.
While I'm sympathetic, such cases present a dangerous precedent. If the City of Champaign passed an ordinance setting a $25 fine per incident for any web site that makes its contents available to a Champaign resident without first registering the site with the City, should they be able to collect on that? Should China be able to make it illegal for US sites to provide certain types of information, or should France be able to block a US auction site?
In fact, the response to the entry in the forum you linked to mentions that there have been problems with the WA law when applied outside of Washington because of the Commerce Clause.
My point was, by giving permission to anyone to decrypt YOUR work, wouldn't that make DeCSS legal? If not, then the same argument would work in reverse: you could prohibit anyone from playing it on an MPAA-authorized DVD player, thus making all such DVD players illegal circumvention devices. To do otherwise is to allow and require only the MPAA to speak for all copyright owners.
You apparently didn't read what I said. Extending the term of copyright after the fact, and after having received adequate compensation for it, is what is stealing. The creator was given certain benefits: an admiring public willing to pay; a legal system willing to enforce rights; the ability to use the entire output of the human race leading up to the time the creator did his or her thing; in exchange, the contract called "copyright" said that the work would enter the public domain after a certain amount of time. The creator priced the work taking that into account. Then the contract was changed, extending the term (and in a way that is making a mockery of the term "for a limited time" from the Constitution).
If the creator didn't want the work to eventually enter the public domain, then it should have remained private. By making it public (as in, "published"), you exchange certain things for others in what should be a mutally beneficial transaction. Part of that transaction is not being honored.
You say that a creator has to be continually paid in order to be able to create new works. Why? Wouldn't you also find that if someone continues to be paid for what they did 20 years ago, they'd find no reason to continue producing new works? If it takes an author a year to write a book, why shouldn't a single payment for that book be sufficient (e.g. a reasonable year's worth of salary)?
Copyrights do provide incentives to create, and some form of protection is probably justified. However, there is also a price to them. Interlocking rights, especially in things like movies, can be a legal nightmare, preventing many interesting projects from being done. Adaptations and extensions of popular stories are attacked, thus preventing new creations (I hear you cry "but they should come up with original work", and I respond "it is original work, based on other work, just as everything is - that's all part of the whole marketplace of ideas that you agreed to participate in when you made your work public").
As far as patronage, you twist the idea if you think that the only reason to do something is to get a "return on investment". You summarily reject the idea of Karma Points, while ignoring all the foundations that rich people set up. Do they expect a "return on investment" from them, other than the general welfare of all people?
(that one is infinitely reproducable at economies of scale is immaterial)
No, that is precisely the material point that is important. If you could reproduce his DVD player so that he lost none of the use of it, he wouldn't have any objection. Sony might, but that's the same argument.
I think most people would agree that a person who puts time and skill into creating something deserves something in return. As long as we live in a scarcity-based economy, where we need to "make a living", that return is most likely going to be in monetary terms. Thus, we grant a monopoly to turn an infinitely-reproducible item into a scarce item, to make it fit into the rest of our economy. At the same time, making that monopoly indefinite is also not fair. I've heard authors (e.g. Stephen King) say "why should you be able to make another copy without paying me?". Yet, isn't the reverse of that also true? Why should he get paid for more than the one copy he wrote? The partial answer is, of course, that the price per copy has been set to account for multiple payments. However, hasn't that price also been set to account for pirating? It is a mostly free market, after all, so it must be taking into account most of the relevant factors when determining the proper price.
However, that should mean that the longer the time span of a copyright, the lower the price per copy, as the payback time has been increased. I haven't seen any evidence of that. In fact, as the copyright term increases, an alternate effect of increasing prices because of less competition (from the public domain) seems to be more dominant.
That's not even taking into account the injustice of changing the term of copyright for existing works after the fact; if the only reason for copyright is to encourage new works, there's absolutely no reason to extend the length for works that have already been created. Those works have already received their payback, with prices set based on a shorter term. It is unfair to steal those works from us.
Perhaps a patronage system would work better. Someone with means pays the artist, author, programmer a salary, giving that person the freedom to create without the pressure of producing something that is financially lucrative. Release the results to the public. Results: the patron gets Karma Points, and a popular creator gets more patrons and/or can get a higher salary.
Wouldn't it make more sense to simply give permission to anyone in the world to decrypt it, using any DVD player (authorized by the MPAA or not)? The only reason that a DVD player is legal at all is because of the "permission of the copyright owner"; if you and the MPAA are both copyright owners, there's no reason your permission is any less valid than the MPAA's.
Why not build an Open Firmware clone? There are free Forth systems readily available, the extensions necessary for OF are not excessive. Doing a byte-code compiler using the OF specs would be interesting, but not particularly difficult (starting with a normal Forth environment).
OF even supports multiple partition schemes and filesystems. Linux has support for using OF as a bootstrap interface. I've always thought that an interesting approach to booting BIOS-based machines would be to use a boot loader that implemented Open Firmware, then use that to boot an OF-aware Linux. Of course, you'd need to keep old-BIOS support if you wanted to boot MS operating systems (and support the new boot ROM "standard" that Intel is introducing for IA-64).
If you're talking about OpenBoot, that turned into Open Firmware, which is alive and well. OF is an excellent solution, and makes a lot more sense than a LinuxBIOS.
Oh My God, you think maybe Apple could implement something like this in MacOS too?
Linux has support for the Apple partition map scheme, which is quite flexible. MacOS and LinuxPPC can both mount FAT partitions on disks with Apple partition maps, or can mount disks with DOS-style partition maps.
There are a few dumb things about Apple partitions, especially with some dumb drivers for drives with block sizes other than 512 bytes (e.g. CD-ROM or some optical drives), but overall it works very well.
I think an ideal first language to implement on your 4000-transistor wire-wrapped 4K machine would be something like Forth; an almost trivial core, a very simple set of primitives, and the rest is effectively self-compiling and very space-conservative. You could hand-assemble your first Forth interpreter, then write an assembler in Forth. I know some people will say LISP would be a better choice, but I just don't like it as much as Forth.
Here's my list of inventions/subject areas that I think would be both possible to implement and most useful with very little base technology behind it (metal working is the only real requirement; you'd have to introduce that as well if it wasn't sufficient). Someone should write a "Future Knowledge for Dummies (1100 A.D. edition)"
Note that some of these might seem a little ambitious, but they are all based on very simple principles (once you know them, that is). For instance, making a transistor wouldn't be possible early on, but with an electric oven and some chemistry developed, you could probably produce crude doped semiconductors, then improve upon the process as you move along.
9. The Free Software Foundation may publish revised and/or new versions
of the General Public License from time to time. Such new versions will
be similar in spirit to the present version, but may differ in detail to
address new problems or concerns.
This clause seems to prevent RMS from "selling out".
We've known for a long time about alternate reading frames (DNA is read in triplets, so if you shift over one before you start reading, you're in a different frame), genes within genes (more in viruses than eukaryotes like us), alternate splicing, post-translational modification of proteins (insulin is a cool example of this)
This sounds a lot like some of the really hacked code people have come up with trying to cram as much functionality into a small amount of memory (e.g. using an instruction as the offset for an overlapping branch instruction, then making sure that the correct code is at that location; self-modifying code; funky arbitrary-looking calculations that produce addresses that just happen to do the right think). This is also the type of code that genetic algorithms tend to come up with.
If there is some evolutionary pressure to keep the genome short (such as increased chance of lethal mutation), I'd expect a really kludged and hacked solution such as what appears to be the actual case.
Although Linus could give his permission for anything he wrote or was assigned to him, all other pieces of the Kernel remain under their own copyright. To the extent that some of those pieces are straight GPL, Linus has no authority to change the terms.
However, I don't see his "NOTE!" as being anything other than an interpretation of the effects of the standard GPL. To the extent that other contributors to the Kernel knew about his interpretation and/or agree with it, it is presumably valid. Where other code was borrowed from other GPLed programs, such an interpretation would have little weight (although I happen to think he is correct).
To an extent, I DO agree with Microsoft. I'd much rather have government sponsored or funded research that produces software be required to be in the public domain than to allow it to be kept proprietary by a private company. If that means that it also can't be GPL, I'd still be for it.
Now, I'd still prefer that it be required to be GPLed, but I can see that some companies and people might not agree with that, and I certainly support their right (even if they are Microsoft) to try to convince the legislators that government policy should be to require it to be public domain. However, it is just as legitimate a point of view to try to convince them that it should be GPLed, or some other "open source" license.
What would be inconsistent would be to allow publically funded software to be made proprietary AND to disallow the GPL.
I'm not sure how you can say that TCP/IP wouldn't have taken off if it had first been made available under the GPL. It's quite possible that we'd not only still be using TCP/IP, but that a whole lot more software would now be available under the GPL instead of being proprietary with fractured standards.
It seems to me (NAL) that it is extremely unlikely that the GPL, or any term of it, would be invalidated. What is more likely to happen is that some portion of it would be interpreted in a way that changes how most people have used it; for example, what "distribution", "source code", "incorporate", "independent and separate works", "normally distributed with the major components" mean, and how dynamic linking is affected by the GPL and LGPL; what about plug-ins, what about source code distributions which use both GPL and restricted libraries (e.g. the RIPEM controversy), etc.
Light reflecting off snow or granite or anything else doesn't depend on the atmosphere. It scatters all by itself. That's why you can easily get a sunburn under your chin on the snow. Sunlight doesn't bounce "straight out" from most surfaces. It scatters all over the place. You need a very smooth surface, like still water or a mirror, for light to behave the way you seem to think it does.
still, i would imagine that in the near future it would be best to install Darwin on these machines. it seems considerably more stable, and can compile just about everything application that compiles for LinuxPPC with minimal effort, especially commandline (i.e. server) tools.
Except that Darwin won't run on older machines, certainly not on Nubus machines. When I got a new Mac (clone), my old desktop machine, a 6100/60, became a mail gateway machine running MkLinux. Part of the decision was actually technical, because the code we were using had originally been on a Sun 3 and had some byte-order dependencies. Other portions of code had been developed on x86 Linux.
That machine is still running, although we upgraded from DR1 to a more recent version of MkLinux when we needed to upgrade sendmail. There is a small Mac partition that is used only to boot into MkLinux (do they have boot code that works without MacOS yet? I know they made it so it could boot from a Mac file system, which makes installing it a lot easier than the original kludge of porting mke2fs and tar into a Mac console-window application).
Other than a very rare glitch with the virtual paging code, the system has been rock solid, with the only downtimes being a power failure longer than our UPS could handle, a hard drive failure, and reloading the OS to upgrade the kernel and sendmail.
I've also tried some of the earlier versions of LinuxPPC; other than the bootstrap process (where I had fun digging into OpenFirmware; I really wish Intel had adopted that instead of inventing Yet Another Bootstrap Protocol for Itanium), it was almost indisinguishable from an Intel box running Linux.
I've ordered the newest LinuxPPC distribution, which should ship tomorrow or so; first thing I'll do is update it to 2.4.0.
Look at the slide shows. They show one slide with the spectrum as seen by a dichromat. There are no gaps in it, just yellow at one end and blue at the other
The problem with Red/Green colorblindness is that you have a large area of the spectrum with only one receptor able to sense it. Once it is beyond the range of the blue receptors, you can't tell what shade it is. The slide is marked as yellow on one end, but could just as well have been marked red or green. Depending on which receptor is missing, a constant-level spectrum would look brighter around one particular frequency, but otherwise look the same hue.
A tetrachromat would see the extra primary color between yellow and red or between yellow and green. The other colors would be better defined. So what most of us see as an indistinct shade of greenish yellow, the tetrachromat would see as a distinct shade
A single-frequency light source would look the same to either. What the tetrachromat would be able to do is distinguish between combinations that look the same to a trichromat. In your example, a mixture of red and green (with a bit more of green than red) would look different from something that was actually greenish-yellow (i.e. with a frequency between red and green) that looks to a trichromat exactly the same.
In particular, color prints and color monitors showing real-world images, being tuned for our RGB eyes, would look different for someone with RYGB eyes. What looks very natural to us would look unnatural to them, since the world is made up of many more frequencies than just RGB. The same thing would happen with someone who had a color shift in their sensors, e.g. RYB, but I don't know how common that might be. From the article, it sounds as if offspring of a tetrachromat might be something like that. I've always wanted to see if having a monitor with 5 colors (Red/Yellow/Green/Cyan/Blue) would look more natural to at least some people (I imagine that a continuous-spectrum monitor would be rather difficult to make).
Much of the prior art in both computer-enabled business methods and computer science will be in the existance of computer programs that were written before the priority date. However, the prior art requested in the BountyQuest postings is exclusively looking for "published documents".
I realize that the easiest way to invalidate an existing patent is to come up with published information that exactly matches the patent. Does this mean that programs that are extremely close, and were in public use, but whose source code was never made public, should be excluded? As an example:
I wrote a program in 1980 that accepted orders for "Think Ahead - Impeach Reagan Now" buttons (hence I can be certain of the date). I still have the source of this program available, and the program was used by "the general public" to order buttons. It was essentially one-click: you ordered the buttons, your identify was stored, and I would deliver the buttons to you and accept payment (because I either knew the person or could look them up in an on-line directory to find out where they lived, and could contact them through the system). It was a reverse auction (actually, more like a dutch auction, as I understand the term) because you specified the price you'd pay and how many buttons you'd want at that price. Later on, I (the vendor) accepted or rejected the price (as it happened, I chose to set the price the same for everyone, at the profit-maximizing level for me; I don't see anything in the priceline patent that requires that the decision as to whether to accept a price or not has to be any particular algorithm, nor that the offered price can't be lowered by the vendor). No, I didn't accept credit cards, and no, I didn't offer services to other "vendors".
Assuming I can prove that the program existed, was used and was publically available in 1980, would that qualify as prior art for these two patents, or be useful in showing that either or both patents are "obvious" (as used in patent law)?
I see no way that BountyQuest is enabling the consolidation of this kind of information. I'd be perfectly willing to submit such information even without the carrot of a bounty.
As noted, the existence of DiVX did weaken the case substantially.
I thought the judge made a significant error with respect to DiVX, though. He said several times that there was significance to the fact that decrypted movies became available on the Internet shortly after DeCSS was created. However, it seems to me that both temporally and causatively a much more likely explanation was due to DiVX becoming widespread, not DeCSS.
Not brought up in the trial record was the fact that "transcoding" will lead to a lower quality than compression from the original digital transfer. There is a fair bit of value to having the DVD in its original compression rather than compressed down to 650MB, and much more value having the original film with no compression. That's why many DVD enthusiasts feel strongly about having an anamorphic transfer rather than a letterboxed widescreen transfer. You've lost a lot of resolution already. Getting a 650MB version of a movie is no substitute for having the DVD - as future compression technologies evolve, you won't be able to take advantage of them as you've already had something like a generational loss. Since that was the whole justification for an abomination like the DMCA in the first place, perhaps Congress should reconsider the whole thing.
Slashdot Mirror
The real innovation of Gopher, that continued in the Web browser, was the idea of selecting something (from a menu or from a hyperlink) that specified a host, a port (implied or explicit) and an opaque specifier. That is an incredibly powerful, yet trivially simple, concept.
The two important things the Web introduced were combining that concept with a markup language (as someone already pointed out, SGML was around for a long time before), and using MIME as a type identifier
So, I'd add MIME as an incremental step before the Web browser. Gopher was attempting to migrate to Gopher+, and add more capability of identifying the content type, but it was the weakest part of the protocol. The step of making an index free-form and intermixed with text (and later, images, dropdown menus, controls, etc.) was a fairly small increment.
Don't forget WAIS in their somewhere. Some of the concepts of WAIS, plus things like Veronica, ended up with what various search engines and database access interfaces do now.
IANAL, but I've spoken with several regarding similar issues. Unless it has changed recently, prior art requires "public use" OR publication. I'm not aware of any such exception, or need for such an exception, could you give me a pointer to the statutory language?
For example, our system uses a specific method of interprocess communication that DEC patented (US patent 4,449,182) and was using as their interface between the OS and peripheral controllers, thus locking out the market for compatible peripherals. Although our code isn't public, our system was in public use, and that was enough to get DEC to drop their suit of a third party. They weren't suing us, so some sort of exemption to allow us to continue using it wouldn't have helped the third party disk manufacturer.
Does DEC still claim that they've never lost a patent case? I guess that's pretty easy to do if you always drop the case before you lose.
BTW, there's a program that is still in existence that was written in 1981 that did on-line voting, including preventing people from voting more than once, allowing anyone to set up a vote. It didn't show percentages, just the raw numbers, and it only offered multiple-choice answers (like Slashdot polls do). The existence of the program in 1981 can be documented, although the file itself was last edited in 1989. I'm not sure how many levels of backups there are, but there are backups from 1991 that contain it, with a relatively easy chain-of-custody that can be established.
In a case involving a bingo program, documentation was provided by a printout of the program from a backup, along with an affidavit describing how backups were done and how we could be sure that the backup was actually from that time period and the dates hadn't been spoofed. An even better method is to print it out, with the date, and sign it. Your testimony later that you personally printed it out and signed it, and that the date was in fact the date when you printed it, should be enough for even the most picky rules of evidence.
It does seem problematic, though. How do you prove that something was well-known and/or obvious in 1975, unless it was a major project run by a company that is still in business? It's going to be just luck to find someone who is an obsessive collector who just happened to hang on to something.
While I'm sympathetic, such cases present a dangerous precedent. If the City of Champaign passed an ordinance setting a $25 fine per incident for any web site that makes its contents available to a Champaign resident without first registering the site with the City, should they be able to collect on that? Should China be able to make it illegal for US sites to provide certain types of information, or should France be able to block a US auction site?
In fact, the response to the entry in the forum you linked to mentions that there have been problems with the WA law when applied outside of Washington because of the Commerce Clause.
My point was, by giving permission to anyone to decrypt YOUR work, wouldn't that make DeCSS legal? If not, then the same argument would work in reverse: you could prohibit anyone from playing it on an MPAA-authorized DVD player, thus making all such DVD players illegal circumvention devices. To do otherwise is to allow and require only the MPAA to speak for all copyright owners.
You apparently didn't read what I said. Extending the term of copyright after the fact, and after having received adequate compensation for it, is what is stealing. The creator was given certain benefits: an admiring public willing to pay; a legal system willing to enforce rights; the ability to use the entire output of the human race leading up to the time the creator did his or her thing; in exchange, the contract called "copyright" said that the work would enter the public domain after a certain amount of time. The creator priced the work taking that into account. Then the contract was changed, extending the term (and in a way that is making a mockery of the term "for a limited time" from the Constitution).
If the creator didn't want the work to eventually enter the public domain, then it should have remained private. By making it public (as in, "published"), you exchange certain things for others in what should be a mutally beneficial transaction. Part of that transaction is not being honored.
You say that a creator has to be continually paid in order to be able to create new works. Why? Wouldn't you also find that if someone continues to be paid for what they did 20 years ago, they'd find no reason to continue producing new works? If it takes an author a year to write a book, why shouldn't a single payment for that book be sufficient (e.g. a reasonable year's worth of salary)?
Copyrights do provide incentives to create, and some form of protection is probably justified. However, there is also a price to them. Interlocking rights, especially in things like movies, can be a legal nightmare, preventing many interesting projects from being done. Adaptations and extensions of popular stories are attacked, thus preventing new creations (I hear you cry "but they should come up with original work", and I respond "it is original work, based on other work, just as everything is - that's all part of the whole marketplace of ideas that you agreed to participate in when you made your work public").
As far as patronage, you twist the idea if you think that the only reason to do something is to get a "return on investment". You summarily reject the idea of Karma Points, while ignoring all the foundations that rich people set up. Do they expect a "return on investment" from them, other than the general welfare of all people?
No, that is precisely the material point that is important. If you could reproduce his DVD player so that he lost none of the use of it, he wouldn't have any objection. Sony might, but that's the same argument.
I think most people would agree that a person who puts time and skill into creating something deserves something in return. As long as we live in a scarcity-based economy, where we need to "make a living", that return is most likely going to be in monetary terms. Thus, we grant a monopoly to turn an infinitely-reproducible item into a scarce item, to make it fit into the rest of our economy. At the same time, making that monopoly indefinite is also not fair. I've heard authors (e.g. Stephen King) say "why should you be able to make another copy without paying me?". Yet, isn't the reverse of that also true? Why should he get paid for more than the one copy he wrote? The partial answer is, of course, that the price per copy has been set to account for multiple payments. However, hasn't that price also been set to account for pirating? It is a mostly free market, after all, so it must be taking into account most of the relevant factors when determining the proper price.
However, that should mean that the longer the time span of a copyright, the lower the price per copy, as the payback time has been increased. I haven't seen any evidence of that. In fact, as the copyright term increases, an alternate effect of increasing prices because of less competition (from the public domain) seems to be more dominant.
That's not even taking into account the injustice of changing the term of copyright for existing works after the fact; if the only reason for copyright is to encourage new works, there's absolutely no reason to extend the length for works that have already been created. Those works have already received their payback, with prices set based on a shorter term. It is unfair to steal those works from us.
Perhaps a patronage system would work better. Someone with means pays the artist, author, programmer a salary, giving that person the freedom to create without the pressure of producing something that is financially lucrative. Release the results to the public. Results: the patron gets Karma Points, and a popular creator gets more patrons and/or can get a higher salary.
Wouldn't it make more sense to simply give permission to anyone in the world to decrypt it, using any DVD player (authorized by the MPAA or not)? The only reason that a DVD player is legal at all is because of the "permission of the copyright owner"; if you and the MPAA are both copyright owners, there's no reason your permission is any less valid than the MPAA's.
Why not build an Open Firmware clone? There are free Forth systems readily available, the extensions necessary for OF are not excessive. Doing a byte-code compiler using the OF specs would be interesting, but not particularly difficult (starting with a normal Forth environment).
OF even supports multiple partition schemes and filesystems. Linux has support for using OF as a bootstrap interface. I've always thought that an interesting approach to booting BIOS-based machines would be to use a boot loader that implemented Open Firmware, then use that to boot an OF-aware Linux. Of course, you'd need to keep old-BIOS support if you wanted to boot MS operating systems (and support the new boot ROM "standard" that Intel is introducing for IA-64).
If you're talking about OpenBoot, that turned into Open Firmware, which is alive and well. OF is an excellent solution, and makes a lot more sense than a LinuxBIOS.
Oh My God, you think maybe Apple could implement something like this in MacOS too?
Linux has support for the Apple partition map scheme, which is quite flexible. MacOS and LinuxPPC can both mount FAT partitions on disks with Apple partition maps, or can mount disks with DOS-style partition maps.
There are a few dumb things about Apple partitions, especially with some dumb drivers for drives with block sizes other than 512 bytes (e.g. CD-ROM or some optical drives), but overall it works very well.
I think an ideal first language to implement on your 4000-transistor wire-wrapped 4K machine would be something like Forth; an almost trivial core, a very simple set of primitives, and the rest is effectively self-compiling and very space-conservative. You could hand-assemble your first Forth interpreter, then write an assembler in Forth. I know some people will say LISP would be a better choice, but I just don't like it as much as Forth.
Here's my list of inventions/subject areas that I think would be both possible to implement and most useful with very little base technology behind it (metal working is the only real requirement; you'd have to introduce that as well if it wasn't sufficient). Someone should write a "Future Knowledge for Dummies (1100 A.D. edition)"
Note that some of these might seem a little ambitious, but they are all based on very simple principles (once you know them, that is). For instance, making a transistor wouldn't be possible early on, but with an electric oven and some chemistry developed, you could probably produce crude doped semiconductors, then improve upon the process as you move along.
Germ theory of disease
Printing press
Gunpowder
Steel
Basic principles of chemistry/physics
Steam engine
Electricity (voltaic cells, generators, alternators, transformers, motors, light)
Refrigeration
Hydraulics
Glass (or would that already be known?)
Optics
Photography
Internal combustion engine
Automobile/motorcycle
Flight (wings and lighter-than-air)
Telegraph
Simple electronics (vacuum tubes, diode, transistor, LED)
Radio
Television
Computer
This clause seems to prevent RMS from "selling out".
This sounds a lot like some of the really hacked code people have come up with trying to cram as much functionality into a small amount of memory (e.g. using an instruction as the offset for an overlapping branch instruction, then making sure that the correct code is at that location; self-modifying code; funky arbitrary-looking calculations that produce addresses that just happen to do the right think). This is also the type of code that genetic algorithms tend to come up with.
If there is some evolutionary pressure to keep the genome short (such as increased chance of lethal mutation), I'd expect a really kludged and hacked solution such as what appears to be the actual case.
Although Linus could give his permission for anything he wrote or was assigned to him, all other pieces of the Kernel remain under their own copyright. To the extent that some of those pieces are straight GPL, Linus has no authority to change the terms.
However, I don't see his "NOTE!" as being anything other than an interpretation of the effects of the standard GPL. To the extent that other contributors to the Kernel knew about his interpretation and/or agree with it, it is presumably valid. Where other code was borrowed from other GPLed programs, such an interpretation would have little weight (although I happen to think he is correct).
Yes, that's one possible scenario. You haven't shown how it is the only possible scenario, nor even why it even the most likely scenario.
"everyone" didn't think that ISO/OSI was the Next Big Thing, certainly not at the time that TCP/IP was introduced and became popular.
To an extent, I DO agree with Microsoft. I'd much rather have government sponsored or funded research that produces software be required to be in the public domain than to allow it to be kept proprietary by a private company. If that means that it also can't be GPL, I'd still be for it.
Now, I'd still prefer that it be required to be GPLed, but I can see that some companies and people might not agree with that, and I certainly support their right (even if they are Microsoft) to try to convince the legislators that government policy should be to require it to be public domain. However, it is just as legitimate a point of view to try to convince them that it should be GPLed, or some other "open source" license.
What would be inconsistent would be to allow publically funded software to be made proprietary AND to disallow the GPL.
I'm not sure how you can say that TCP/IP wouldn't have taken off if it had first been made available under the GPL. It's quite possible that we'd not only still be using TCP/IP, but that a whole lot more software would now be available under the GPL instead of being proprietary with fractured standards.
It seems to me (NAL) that it is extremely unlikely that the GPL, or any term of it, would be invalidated. What is more likely to happen is that some portion of it would be interpreted in a way that changes how most people have used it; for example, what "distribution", "source code", "incorporate", "independent and separate works", "normally distributed with the major components" mean, and how dynamic linking is affected by the GPL and LGPL; what about plug-ins, what about source code distributions which use both GPL and restricted libraries (e.g. the RIPEM controversy), etc.
Light reflecting off snow or granite or anything else doesn't depend on the atmosphere. It scatters all by itself. That's why you can easily get a sunburn under your chin on the snow. Sunlight doesn't bounce "straight out" from most surfaces. It scatters all over the place. You need a very smooth surface, like still water or a mirror, for light to behave the way you seem to think it does.
Except that Darwin won't run on older machines, certainly not on Nubus machines. When I got a new Mac (clone), my old desktop machine, a 6100/60, became a mail gateway machine running MkLinux. Part of the decision was actually technical, because the code we were using had originally been on a Sun 3 and had some byte-order dependencies. Other portions of code had been developed on x86 Linux.
That machine is still running, although we upgraded from DR1 to a more recent version of MkLinux when we needed to upgrade sendmail. There is a small Mac partition that is used only to boot into MkLinux (do they have boot code that works without MacOS yet? I know they made it so it could boot from a Mac file system, which makes installing it a lot easier than the original kludge of porting mke2fs and tar into a Mac console-window application).
Other than a very rare glitch with the virtual paging code, the system has been rock solid, with the only downtimes being a power failure longer than our UPS could handle, a hard drive failure, and reloading the OS to upgrade the kernel and sendmail.
I've also tried some of the earlier versions of LinuxPPC; other than the bootstrap process (where I had fun digging into OpenFirmware; I really wish Intel had adopted that instead of inventing Yet Another Bootstrap Protocol for Itanium), it was almost indisinguishable from an Intel box running Linux.I've ordered the newest LinuxPPC distribution, which should ship tomorrow or so; first thing I'll do is update it to 2.4.0.
Look at the slide shows. They show one slide with the spectrum as seen by a dichromat. There are no gaps in it, just yellow at one end and blue at the other
The problem with Red/Green colorblindness is that you have a large area of the spectrum with only one receptor able to sense it. Once it is beyond the range of the blue receptors, you can't tell what shade it is. The slide is marked as yellow on one end, but could just as well have been marked red or green. Depending on which receptor is missing, a constant-level spectrum would look brighter around one particular frequency, but otherwise look the same hue.
A tetrachromat would see the extra primary color between yellow and red or between yellow and green. The other colors would be better defined. So what most of us see as an indistinct shade of greenish yellow, the tetrachromat would see as a distinct shade
A single-frequency light source would look the same to either. What the tetrachromat would be able to do is distinguish between combinations that look the same to a trichromat. In your example, a mixture of red and green (with a bit more of green than red) would look different from something that was actually greenish-yellow (i.e. with a frequency between red and green) that looks to a trichromat exactly the same.
In particular, color prints and color monitors showing real-world images, being tuned for our RGB eyes, would look different for someone with RYGB eyes. What looks very natural to us would look unnatural to them, since the world is made up of many more frequencies than just RGB. The same thing would happen with someone who had a color shift in their sensors, e.g. RYB, but I don't know how common that might be. From the article, it sounds as if offspring of a tetrachromat might be something like that. I've always wanted to see if having a monitor with 5 colors (Red/Yellow/Green/Cyan/Blue) would look more natural to at least some people (I imagine that a continuous-spectrum monitor would be rather difficult to make).
Much of the prior art in both computer-enabled business methods and computer science will be in the existance of computer programs that were written before the priority date. However, the prior art requested in the BountyQuest postings is exclusively looking for "published documents".
I realize that the easiest way to invalidate an existing patent is to come up with published information that exactly matches the patent. Does this mean that programs that are extremely close, and were in public use, but whose source code was never made public, should be excluded? As an example:
I wrote a program in 1980 that accepted orders for "Think Ahead - Impeach Reagan Now" buttons (hence I can be certain of the date). I still have the source of this program available, and the program was used by "the general public" to order buttons. It was essentially one-click: you ordered the buttons, your identify was stored, and I would deliver the buttons to you and accept payment (because I either knew the person or could look them up in an on-line directory to find out where they lived, and could contact them through the system). It was a reverse auction (actually, more like a dutch auction, as I understand the term) because you specified the price you'd pay and how many buttons you'd want at that price. Later on, I (the vendor) accepted or rejected the price (as it happened, I chose to set the price the same for everyone, at the profit-maximizing level for me; I don't see anything in the priceline patent that requires that the decision as to whether to accept a price or not has to be any particular algorithm, nor that the offered price can't be lowered by the vendor). No, I didn't accept credit cards, and no, I didn't offer services to other "vendors".
Assuming I can prove that the program existed, was used and was publically available in 1980, would that qualify as prior art for these two patents, or be useful in showing that either or both patents are "obvious" (as used in patent law)?
I see no way that BountyQuest is enabling the consolidation of this kind of information. I'd be perfectly willing to submit such information even without the carrot of a bounty.
As noted, the existence of DiVX did weaken the case substantially. I thought the judge made a significant error with respect to DiVX, though. He said several times that there was significance to the fact that decrypted movies became available on the Internet shortly after DeCSS was created. However, it seems to me that both temporally and causatively a much more likely explanation was due to DiVX becoming widespread, not DeCSS. Not brought up in the trial record was the fact that "transcoding" will lead to a lower quality than compression from the original digital transfer. There is a fair bit of value to having the DVD in its original compression rather than compressed down to 650MB, and much more value having the original film with no compression. That's why many DVD enthusiasts feel strongly about having an anamorphic transfer rather than a letterboxed widescreen transfer. You've lost a lot of resolution already. Getting a 650MB version of a movie is no substitute for having the DVD - as future compression technologies evolve, you won't be able to take advantage of them as you've already had something like a generational loss. Since that was the whole justification for an abomination like the DMCA in the first place, perhaps Congress should reconsider the whole thing.