AmigaOS was largely written in C, with the original command line interface for it written in BCPL (a port of Tripos, an OS for minicomputers designed by Cambridge University.)
I'd also like a reduction in the amount of stuff that can be copyrighted (clearly creative non-functional works, yes, but with a reduction in the term lengths available for types of work the more functional, and therefore necessary to build upon and understand, they become. Yes, I'm arguing for a dramatic decrease in the terms available for computer software.)
And I'd like an additional "cost" imposed on copyright holders, namely that to possess a copyright, you must publish and continue to publish. Lack of reasonable availability should be a fair-use defense for non-profit copyright infringement.
How is this different to when a property owner declares a police officer should go and arrest someone for stealing from them, or a friend of a murder victim should go and arrest someone for killing their friend? (No flames please about the degree of crime involved, I'm not interested, I'm just picking two examples of has-a-victim crimes at random.)
All parties involved, in all these cases, are paying taxes. The suggestion here is that an additional tax, over and above income and sales taxes, should be levied on copyright owners. There may be logical arguments for this, but I don't think it's reasonable to suggest it corrects a mismatch that doesn't exist elsewhere.
Try this thought-experiment. Follow a dollar as it goes from your pocket into a theater. Who gets it? Does nobody, in the end, receive it in the form of a salary? With the possible exception of monies spent on production costs outside of your country, there's little chance that it will not end up as straightforward income for someone somewhere. It might, at worst, end up being a shareholder's dividend, but even those are taxed.
It might be reinvested into a new film, but that film will cost money to make, largely in the form of salaries. Or it might be more direct, it might be that it'll be part of the salary for the very ticket seller you handed the dollar to in the first place.
One can make cases for where additional taxes should or shouldn't be levied. But it's false to assume that because corporations can get out of income taxes (usually by the perfectly legitimate policy of reinvesting surplusses of revenue back into the business to make it grow), the government doesn't end up getting huge amounts of money from those businesses. While employees pay income tax, and while those same employees constitute a substantial part of the expenses of any real company, money is being funnelled the IRS's way.
He said "Income Tax", not corporate taxes. The actors, director, producers, etc, were all subject to income taxes from the wages they earned from Spiderman II, and no matter how you slice it, 20-50% of the revenues from SII will ultimately end up with the government because it'll end up going to people, one way or another.
If you follow the developers links, you'll eventually get to a PDF documenting the porting considerations. Unfortunately, part of the Rosetta subsystem's limitations is that it doesn't run Classic Apps. Which means pretty much all 68k apps.
My guess is that your best hope is that Ardi, makers of Executor, will port the latter to the Intel OS X.
Why not? All three of the 2006 videogame consoles are going to be PPC-based (if you consider the Cell to be PPC-derived). More and more personal computers and even some embedded devices are using the PPC processor architecture. The only reasons for Intel not to start pressing PPC chips are 1) the "Not Developed Here" syndrome, and 2) to not take any fabrication resources away from the x86 cash cow. I'd imagine there could be a business case that overrides both of those concerns.
Absolutely none of these devices needs to be PowerPCs. It happens that right this second IBM is doing sterling work creating custom CPU "solutions" for various third parties. If Intel has a better CPU in 2008, Microsoft will use it for Xbox 3. Likewise Sony and Nintendo.
The question isn't "Why not" but "Why". What is it about PowerPCs that would help Intel? Can they not produce custom CPU solutions with an ix86, i860, Itanium, XScale, etc, architecture, but some-how can with an architecture they have absolutely no control over and no experience of? Would PowerPCs give them access to a new market?
And who better than Intel's chip designers to implement such improvements? More so the "faster" than the "less power", but still.
IBM's chip designers? They are actually familiar with the design. If IBM's can't, why should Intel's be able to?
So, if I understand this correctly, you're saying Intel can save PowerPC but they can't save their own lines of processors?
Heh.
If I had to bet money, right now I'd say ix86. Itanium is nice but it doesn't scale and it's Intel only. ix86 has 32 bit "low" and "medium" ends, in the various incarnations of the Pentium M, plus the 64 bit "high" end Xeon. Apple wouldn't be/isn't the only customer, so Apple gets the benefits that massive mass-production generates. The only potential hinderance is that enterprising hackers may produce tools to shoe-horn OS X onto non-Apple machines, but it's questionable how damaging that'll be in practice. It may even work in Apple's favour.
Intel's XScale is because they bought the rights from DEC's faltering CPU business. It worked well with Intel at the time because the ARM didn't compete with any successful CPU line.
The PowerPC competes in three areas - the embedded space (where XScale competes), the desktop space (badly) (where the Pentium reigns), and the workstation market (where Intel is trying to get the Itanium to take off.) By producing PowerPCs, they confer additional credibility onto their biggest competitor, and help it attack Intel's other products.
It would be a strategic mistake for Intel to start producing PowerPCs.
I think we're beginning to look at an answer that makes sense, even if it'd end up being a death knell for the processor as a whitebox CPU (IMHO)
It'd be nice to see the Itanium get some steam. Intel are obviously a talented bunch, but they're having to pour their most visible efforts into a CPU range that's a great big kludge, by necessity.
Problems with this theory: 1. Intel probably wouldn't want to produce PowerPCs. 2. Apple has little or nothing to gain by Intel producing 970 clones except, possibly, fractional improvements in price.
Apple's problem at the moment is two-fold: The 970 is capped at well under the 3GHz originally promised, and it's still a power hog. It can't be used in laptops. Intel building clones isn't going to help, the 970 will need some design work to get faster and/or use less power.
Meanwhile, not only are we supposedly asking Intel to undermine the credibility of its own competitors to the PowerPC range, but also to make a CPU for one manufacturer that's come to it saying "We went to Freescale and they suck, so we went to IBM and they suck too, and now we're coming to you". For Intel to bite, it has to have real value for them, and being a second-tier chip cloner is not where Intel is at. They need Apple to confer credibility on what Intel does.
This still doesn't make any sense. The problem with the 7400 that prompted Apple to go to IBM was that Motorola couldn't make enough of them, not that they weren't fast enough. While this is a problem with the 970, it's not the biggest problem by a long shot. On top of that, IBM had the right to make PowerPCs. IBM was a member of the AIM alliance. There were technology sharing agreements between Motorola and IBM that made this possible. Intel isn't.
Intel is supposedly being brought in because development of the 970 has stagnated. It's not getting any faster and IBM has been forced to start offering Apple multiple core designs to get around the inability of the technology to increase speeds any other way. At the same time, it's an expensive chip. Supposedly it's cheaper than the 74xx/75xx series, but, presumably because of the relatively small market, IBM isn't selling the chip at anything close to a commodity price, and is being expected to improve them over time. By comparison, the console chips IBM's making are going to be unchanged for three years and will likely be sold in the hundreds of millions each.
So I still don't buy it.
If you take a look at my journal, I restate your comment about the problems with a move to ix86. I think it's a little more likely that Intel's trying to use Apple to kickstart the Itanium. However, I also think Apple picking Itanium would be the death-knell for it as a White Box chip, as Microsoft would never agree to support a white box market based on the same CPU that Apple uses.
I have enormous difficultly believing what'll be announced is that Intel will produce PowerPC chips.
You're talking about Intel somehow coming up with a PowerPC compatable chip in the space of a year, one that'll be significantly better than those IBM and FreeScale are producing now. That's a design that'll be built, tested, and put into mass-production, in the space of one year.
Even if we assume some of the work's been done already, it's still a highly implausable time table. Two years instead of one? The public turnaround of the 970 was almost that long, and that's with IBM having much of the design work already done.
And what is this going to buy anyway? Intel starts a PowerPC division for ONE CUSTOMER? One thing's for sure, Intel's not going to be undercutting IBM on price, and if the aim is to put these chips in the Mac mini, that's going to be a big problem.
Where's the evidence, incidentally, that Apple can just add third parties to AIM if IBM screws up? Did Motorola agree to this?
I just don't buy it. IF, and it's a huge if, Apple will announce something on Monday about switching the Mac line to Intel produced chips, it'll be based upon an existing Intel design. It might be a version of Itanium with a close, easy to translate into, approximation to Altivec. And that'll be only because Intel wants to get Itanium out there, and will probably come at the cost of ensuring it'll never end up in white box PCs, so representing a desperate last throw by Intel. It might be an ARM, though Intel just doesn't make them at that speed. It might be a Pentium M, though Intel had better announce a 64 bit version if they do.
It's not impossible an Intel PowerPC is on the horizon, but it seems far from likely.
The PowerPC was a good fit to replace the 68k range. It was significantly faster, and could be used with the same endianisms, probably the only CPU-specific thing that programmers end up tripping over when coding in Pascal, C and direct derivatives (C++, Objective C, etc)
I don't think the same can be said for the current ix86 range, and I'm doubtful the Itanium either, though I'd love someone to point me at a document that says I'm wrong there. The ix86 isn't significantly faster than what it's allegedly replacing, so its ability to emulate the PowerPC isn't comparable to the PowerPC/68k thing. And the endianisms are reversed for the 68k order (still used by OS X apps) and Intel chips.
So it'll not be as trivial and painless as the original switch over. In practice, users will have to put up with their older apps running much more slowly and requiring a lot more memory, and developers will find a lot of things tripping them up when trying to cross-compile for both architectures. It'll be hairy.
He already manages the funding of various projects, so it'd simply merge two potentially diverse funding sources. We don't really need or want that. Different mindsets are good.
Well, it's been a day, and nobody really answered the point...
Let me thank you for taking the time to respond to me, I didn't realise you were one of the developers. However, I have to say the evidence so far is far from convincing that the IP spoofing support exists "as a side effect". It's hard to see what could have been developed, related to anonymisation technologies, that would have resulted in such support.
Right at the moment there's a battle between content producers and P2P users, one that, frankly, the P2P users are on the wrong side of. There's absolutely no legitimate reason to distribute someone else's copyrighted content on the Internet without their consent. Every single time someone does something that makes it easy to infringe copyrights, not accidentally, but as deliberate policy - be it Shawn Fanning's inspiration of IRC MP3 trading of largely illegitimate files to build Napster, to the attempts to create distributed, anonymous, file trading systems like yours - it harms the credibility of the tech community and encourages those who want more draconian laws to seek them. At a time we need more accountability, to ensure people do not abuse the technologies, you're doing what you can to prevent that.
What you're doing is not a good idea. It's an unnecessary shot in a war that shouldn't be fought this way. Those who oppose copyrights on principle do not have to violate them to undermine the system, it's important to render copyrights unnecessary than harder to enforce. To do the latter is destructive, something wonderful - a system that encourages the creation of new content - is destroyed and nothing fills that vaccuum. Regardless of that though, the end result of the latter is draconian law writing, not a copyrightless utopia.
With respect, I'm pretty sure that if you'd found a way to put the latest Hollywood blockbuster or your entire music collection up for easy, workable, download back in 1992, they'd have sued you then too.
What we're seeing is the addition of content, so nothing's no longer free, it's just there's a lot of non-free stuff that's been added to the free stuff. Actually, there's also a hell of a lot of free stuff that's been added then too.
Certain groups are reacting very negatively to the notion that people should put things they developed at often enormous financial risk, on the explicit understanding that copyright allows them to have users of their content pay for the use of it, up for download for free. The commercial interests you speak of frequently would be perfectly happy to have no role whatsoever in the Internet, it's just people keep posting their stuff onto it.
There's absolutely no reason to do IP address spoofing as part of a file transfer protocol from what I can see. Load balancing certainly doesn't require IP address spoofing...
Worse still, this system seems to be ripe to create DDoS attacks, not prevent them. Just tell a bunch of servers that the site you want to attack is "C".
Nonetheless, they're implemented, correct? These are not optional in the sense that someone who wants to use them has to actually write the code to make them work, and they weren't created as a natural side effect of the system. Someone, somewhere, said "We need to make it possible to anonymously transfer files, in such a way that people cannot get caught when deliberately transfering files in a way that breaks the law." And given where the legal pressure is on P2P at the moment, it's safe to know what type of breaking the law we're talking about.
It does. If it was up to me, I'd have the system break down every file into managable chunks, and download each chunk, CRCd, etc, separately. Packets of packets. You can even make the system optimize itself by having it do this on the fly, reducing the chunk size when errors occur.
If this protocol is supposed to be based upon BitTorrent, then given that's how much of BT works already, I suspect that's what they've done.
How does spoofing the sender address in the IP packets help with decentralizing the content search? Remember, it was that part of the system I was talking about, that I referred to as having the primary reason you quoted.
I think the "too smart for their own good" part of the plan is to assume that even though A knows B's address, it's not important, because A can't prove B is sending anything. All A knows is that to get a file sent, B has to be contacted. B could then pass the request on to D, for all A knows.
FWIW, I agree with you in general and I suspect that, whatever else is the case, "They can just continue doing what they are currently doing on BT and other P2P networks, being A, and get the people who are distributing the (software|music|movies)'s IP addresses.". I think a geek who turns up in court and says to the judge "*snuk* you can't prove it was me, all I was doing was collecting requests, I might have been passing them on to someone else" will get a quick slap and a reminder both that the phrase is "reasonable doubt", not "even 'Aliens stole my homework' is a valid defense", and that it's perfectly possible to be a party to a crime, it doesn't become legal because several people did different parts.
Slashdot Mirror
For what it's worth.
I'd also like a reduction in the amount of stuff that can be copyrighted (clearly creative non-functional works, yes, but with a reduction in the term lengths available for types of work the more functional, and therefore necessary to build upon and understand, they become. Yes, I'm arguing for a dramatic decrease in the terms available for computer software.)
And I'd like an additional "cost" imposed on copyright holders, namely that to possess a copyright, you must publish and continue to publish. Lack of reasonable availability should be a fair-use defense for non-profit copyright infringement.
All parties involved, in all these cases, are paying taxes. The suggestion here is that an additional tax, over and above income and sales taxes, should be levied on copyright owners. There may be logical arguments for this, but I don't think it's reasonable to suggest it corrects a mismatch that doesn't exist elsewhere.
Try this thought-experiment. Follow a dollar as it goes from your pocket into a theater. Who gets it? Does nobody, in the end, receive it in the form of a salary? With the possible exception of monies spent on production costs outside of your country, there's little chance that it will not end up as straightforward income for someone somewhere. It might, at worst, end up being a shareholder's dividend, but even those are taxed. It might be reinvested into a new film, but that film will cost money to make, largely in the form of salaries. Or it might be more direct, it might be that it'll be part of the salary for the very ticket seller you handed the dollar to in the first place.
One can make cases for where additional taxes should or shouldn't be levied. But it's false to assume that because corporations can get out of income taxes (usually by the perfectly legitimate policy of reinvesting surplusses of revenue back into the business to make it grow), the government doesn't end up getting huge amounts of money from those businesses. While employees pay income tax, and while those same employees constitute a substantial part of the expenses of any real company, money is being funnelled the IRS's way.
He said "Income Tax", not corporate taxes. The actors, director, producers, etc, were all subject to income taxes from the wages they earned from Spiderman II, and no matter how you slice it, 20-50% of the revenues from SII will ultimately end up with the government because it'll end up going to people, one way or another.
My guess is that your best hope is that Ardi, makers of Executor, will port the latter to the Intel OS X.
The question isn't "Why not" but "Why". What is it about PowerPCs that would help Intel? Can they not produce custom CPU solutions with an ix86, i860, Itanium, XScale, etc, architecture, but some-how can with an architecture they have absolutely no control over and no experience of? Would PowerPCs give them access to a new market?
IBM's chip designers? They are actually familiar with the design. If IBM's can't, why should Intel's be able to?Heh.
If I had to bet money, right now I'd say ix86. Itanium is nice but it doesn't scale and it's Intel only. ix86 has 32 bit "low" and "medium" ends, in the various incarnations of the Pentium M, plus the 64 bit "high" end Xeon. Apple wouldn't be/isn't the only customer, so Apple gets the benefits that massive mass-production generates. The only potential hinderance is that enterprising hackers may produce tools to shoe-horn OS X onto non-Apple machines, but it's questionable how damaging that'll be in practice. It may even work in Apple's favour.
I just read something in the Wall Street Journal suggesting he was going to phase out the 200 TTL chips, and switch to Intel...
Intel's XScale is because they bought the rights from DEC's faltering CPU business. It worked well with Intel at the time because the ARM didn't compete with any successful CPU line.
The PowerPC competes in three areas - the embedded space (where XScale competes), the desktop space (badly) (where the Pentium reigns), and the workstation market (where Intel is trying to get the Itanium to take off.) By producing PowerPCs, they confer additional credibility onto their biggest competitor, and help it attack Intel's other products.
It would be a strategic mistake for Intel to start producing PowerPCs.
I think we're beginning to look at an answer that makes sense, even if it'd end up being a death knell for the processor as a whitebox CPU (IMHO)
It'd be nice to see the Itanium get some steam. Intel are obviously a talented bunch, but they're having to pour their most visible efforts into a CPU range that's a great big kludge, by necessity.
Apple's problem at the moment is two-fold: The 970 is capped at well under the 3GHz originally promised, and it's still a power hog. It can't be used in laptops. Intel building clones isn't going to help, the 970 will need some design work to get faster and/or use less power.
Meanwhile, not only are we supposedly asking Intel to undermine the credibility of its own competitors to the PowerPC range, but also to make a CPU for one manufacturer that's come to it saying "We went to Freescale and they suck, so we went to IBM and they suck too, and now we're coming to you". For Intel to bite, it has to have real value for them, and being a second-tier chip cloner is not where Intel is at. They need Apple to confer credibility on what Intel does.
Microsoft is getting a custom PowerPC from IBM, they're not putting (so far as I can see) off-the-shelf 970s in them.
Intel is supposedly being brought in because development of the 970 has stagnated. It's not getting any faster and IBM has been forced to start offering Apple multiple core designs to get around the inability of the technology to increase speeds any other way. At the same time, it's an expensive chip. Supposedly it's cheaper than the 74xx/75xx series, but, presumably because of the relatively small market, IBM isn't selling the chip at anything close to a commodity price, and is being expected to improve them over time. By comparison, the console chips IBM's making are going to be unchanged for three years and will likely be sold in the hundreds of millions each.
So I still don't buy it.
If you take a look at my journal, I restate your comment about the problems with a move to ix86. I think it's a little more likely that Intel's trying to use Apple to kickstart the Itanium. However, I also think Apple picking Itanium would be the death-knell for it as a White Box chip, as Microsoft would never agree to support a white box market based on the same CPU that Apple uses.
You're talking about Intel somehow coming up with a PowerPC compatable chip in the space of a year, one that'll be significantly better than those IBM and FreeScale are producing now. That's a design that'll be built, tested, and put into mass-production, in the space of one year.
Even if we assume some of the work's been done already, it's still a highly implausable time table. Two years instead of one? The public turnaround of the 970 was almost that long, and that's with IBM having much of the design work already done.
And what is this going to buy anyway? Intel starts a PowerPC division for ONE CUSTOMER? One thing's for sure, Intel's not going to be undercutting IBM on price, and if the aim is to put these chips in the Mac mini, that's going to be a big problem.
Where's the evidence, incidentally, that Apple can just add third parties to AIM if IBM screws up? Did Motorola agree to this?
I just don't buy it. IF, and it's a huge if, Apple will announce something on Monday about switching the Mac line to Intel produced chips, it'll be based upon an existing Intel design. It might be a version of Itanium with a close, easy to translate into, approximation to Altivec. And that'll be only because Intel wants to get Itanium out there, and will probably come at the cost of ensuring it'll never end up in white box PCs, so representing a desperate last throw by Intel. It might be an ARM, though Intel just doesn't make them at that speed. It might be a Pentium M, though Intel had better announce a 64 bit version if they do.
It's not impossible an Intel PowerPC is on the horizon, but it seems far from likely.
I don't think the same can be said for the current ix86 range, and I'm doubtful the Itanium either, though I'd love someone to point me at a document that says I'm wrong there. The ix86 isn't significantly faster than what it's allegedly replacing, so its ability to emulate the PowerPC isn't comparable to the PowerPC/68k thing. And the endianisms are reversed for the 68k order (still used by OS X apps) and Intel chips.
So it'll not be as trivial and painless as the original switch over. In practice, users will have to put up with their older apps running much more slowly and requiring a lot more memory, and developers will find a lot of things tripping them up when trying to cross-compile for both architectures. It'll be hairy.
He already manages the funding of various projects, so it'd simply merge two potentially diverse funding sources. We don't really need or want that. Different mindsets are good.
Well, that's disappointing. I have to hope, I guess, that this isn't something that becomes popular.
Let me thank you for taking the time to respond to me, I didn't realise you were one of the developers. However, I have to say the evidence so far is far from convincing that the IP spoofing support exists "as a side effect". It's hard to see what could have been developed, related to anonymisation technologies, that would have resulted in such support.
Right at the moment there's a battle between content producers and P2P users, one that, frankly, the P2P users are on the wrong side of. There's absolutely no legitimate reason to distribute someone else's copyrighted content on the Internet without their consent. Every single time someone does something that makes it easy to infringe copyrights, not accidentally, but as deliberate policy - be it Shawn Fanning's inspiration of IRC MP3 trading of largely illegitimate files to build Napster, to the attempts to create distributed, anonymous, file trading systems like yours - it harms the credibility of the tech community and encourages those who want more draconian laws to seek them. At a time we need more accountability, to ensure people do not abuse the technologies, you're doing what you can to prevent that.
What you're doing is not a good idea. It's an unnecessary shot in a war that shouldn't be fought this way. Those who oppose copyrights on principle do not have to violate them to undermine the system, it's important to render copyrights unnecessary than harder to enforce. To do the latter is destructive, something wonderful - a system that encourages the creation of new content - is destroyed and nothing fills that vaccuum. Regardless of that though, the end result of the latter is draconian law writing, not a copyrightless utopia.
Please remove this feature from your code.
What we're seeing is the addition of content, so nothing's no longer free, it's just there's a lot of non-free stuff that's been added to the free stuff. Actually, there's also a hell of a lot of free stuff that's been added then too.
Certain groups are reacting very negatively to the notion that people should put things they developed at often enormous financial risk, on the explicit understanding that copyright allows them to have users of their content pay for the use of it, up for download for free. The commercial interests you speak of frequently would be perfectly happy to have no role whatsoever in the Internet, it's just people keep posting their stuff onto it.
Worse still, this system seems to be ripe to create DDoS attacks, not prevent them. Just tell a bunch of servers that the site you want to attack is "C".
Nonetheless, they're implemented, correct? These are not optional in the sense that someone who wants to use them has to actually write the code to make them work, and they weren't created as a natural side effect of the system. Someone, somewhere, said "We need to make it possible to anonymously transfer files, in such a way that people cannot get caught when deliberately transfering files in a way that breaks the law." And given where the legal pressure is on P2P at the moment, it's safe to know what type of breaking the law we're talking about.
If this protocol is supposed to be based upon BitTorrent, then given that's how much of BT works already, I suspect that's what they've done.
How does spoofing the sender address in the IP packets help with decentralizing the content search? Remember, it was that part of the system I was talking about, that I referred to as having the primary reason you quoted.
FWIW, I agree with you in general and I suspect that, whatever else is the case, "They can just continue doing what they are currently doing on BT and other P2P networks, being A, and get the people who are distributing the (software|music|movies)'s IP addresses.". I think a geek who turns up in court and says to the judge "*snuk* you can't prove it was me, all I was doing was collecting requests, I might have been passing them on to someone else" will get a quick slap and a reminder both that the phrase is "reasonable doubt", not "even 'Aliens stole my homework' is a valid defense", and that it's perfectly possible to be a party to a crime, it doesn't become legal because several people did different parts.