games are where mindshare is in the eyes of Joe Sixpack buying a PC for his kids.
Uhm, nope. If Joe Sixpack wants his kids to play games, he'll get them a PlayStation or DreamCast just like all of their friends have.
The reason that there are so many games under Windows is that it is, relatively speaking, easy to create them.
Nope again. Games existed for DOS/Windows long before DirectX and related goodies in huge numbers. Games are written for Windows because of the huge installed user base. Developers put up with DOS's terrible memory model (I was about to say memory management but reconsidered) and built drivers for every sound and graphics card from scratch for every game or line of games. They put up with this because people used DOS and were going to buy games for DOS, and didn't care much about how easy it was to develop for it.
The availability of "productivity" apps, like an office suite, is a much more important factor in swaying users to Linux. Conquering the home market is hard, and not the first priority. Conquering the small- and mid-size workstation market in the office is a much more likely target.
"It can't do anything my PC can do except e-mail and Web browsing" isn't the problem. The problem is that it won't do the new things your PC will do one day, that you never even envisaged when you bought your PC. Let me explain.
A lot of people in this discussion have re-stated the most common argument against Internet appliances: That they can't do many of the things you can do with a PC. Others have re-stated the common counter-argument: That most people don't want to do these things.
However, I think there is a much more important problem with these machines: That they have a limited function.
The thing about a PC, or any computer, is that you might design it to do one thing (e.g. word processing, games playing), and people might buy it for this thing, but it is very easy for it to do something completely different in the future. The Internet is a good example. Before the Net explosion, people bought PCs to do their word processing, spreadsheets, maybe play a game or two. When the Internet came along, they used those same PCs to access it; that's the real power of a non-crippled computer: It can do things that you never planned for.
This thing might be great for e-mail and Web browsing, and people who only need to do these things will use it. But when they want to do something new and different (god knows what!) a year down the line, it will be useless. While those who own a PC will, at the most, install a piece of software or buy a small peripheral.
This may seem like a small problem for the individual user, but it is a HUGE problem for the market as a whole. If the PC is replaced by specialised appliances, deploying new technologies will not just involve releasing some software and maybe a Web site. It will involve getting an entirely new appliance to all of your prospective users.
To illustrate, imagine this scenario: Roll back history for about 2 years, and imagine that everyone on the Internet had an Internet Appliance instead of a PC. And someone comes up with MP3s. Could the technology have spread as it has in our world? No, because unless this person managed to convince appliance manufacturers to include it in the next version, and waited for that version to actually get out and occupy a significant portion of the market, nobody could use it.
Internet appliances raise the barrier of entry for new technologies. This is their main disadvantage, and it's a scary prospect.
The answer to your question is "maybe", but I sincerely doubt it.
Crusoe, as the article mentions, uses VLIW. VLIW normally depends on the compiler ordering instructions in the most effective manner, and taking care of issuing them to the most appropriate pipeline. This takes a lot of complexity out of your chip. You get the advantages of super-scalar design without all of that multi-issue hardware that bloats up your processor.
The downside is that your VLIWs are hardcoded for a specific processor design. The Crusoe has four pipelines, each of them specialised for certain functions. If, in a future version of the Crusoe, they decide to change the number and/or type of these pipelines, the optimal structure of these VLIWs will be completely different.
Now, as I said, normally making the VLIWs (or "molecules" as TransMeta calls them) depends on the compiler. This means that every time you change the internals of the processor implementation, you need to completely re-design your compiler as well - not really possible. But in Crusoe, it's the Code Morphing software's job. And since this comes part & parcel with the CPU itself, you'll always get super-scalar optimization that works with your CPU.
Of course, this brings us to another question. If Crusoe can emulate x86, can't it emulate other instruction sets? Of course it can. But I'm not talking about Sparc, Alpha or MIPS here.
Imagine this scenario:
As someone else mentioned here, let's say you put, say, 8 Crusoes on a motherboard and have the Code Morphing software handle part of the MP. And let's assume that these processors initially act as CPUs in SMP mode. And now you start Quake, and you want some polygon-pushing power. What if the Code Morphing software picks this up, and automatically switches the ISAs of a couple of your processors to GPU mode. Bingo, you've got 3D acceleration WITHOUT buying an expansion card, WITHOUT going through the system bus, and WITHOUT investing into a big bunch of computation power that sits idle 99.9% of the time.
Need a DSP to do sound processing? An MPEG decoder chip to play DVDs? Just switch one or more of your processors into that mode. Need massive computation power for mathematical calculations? Switch them all back to x86 mode.
All this can be done by the Code Morphing software. Granted, maybe you'd like to have different kinds of processors (i.e. some of them might have more FP pipelines) to make this more efficient, but it's all transparent to the software.
I think this could be the killer feature that brings Crusoe out of the mobile market and into the desktop/workstation/server market.
A lot of people are commenting on how possible sanctions on MS (esp. those concerning limiting its ability to cut deals with licensees and hiding APIs) would set a precedent and generally create a very difficult environment to work in for other companies.
But it's important to note that we're not talking about other companies. Other companies don't have a wide range monopoly (i.e. a monopoly on desktop OSs that feeds monopolies on office software, server OSs, back-end tools, development tools, browsers or anything else).
Let's face it, the only way for Microsoft to stop stifling innovation in the computer industry is for its monopoly to be completely deposed in a few sectors. They have to loose, say, the browser wars and the OS wars. Once that happens, they'll have to play fair with others because it will be impossible to offer a Microsoft-only solution as once can do now.
Today, you can set up an entire company network using only Microsoft software. Win98 + WinNT + Office + BackOffice + IE + VisStudio covers practically anything you need, with only niche markets available for others (and since they ARE niche markets, they have to work with the MS non-niche products that even niche clients will use in addition to niche products - not everyone needs CAD, but even those who need CAD need word processing).
You can't make them loose these monopolies through a direct court action, but it's not necessary. We all know that Microsoft will definitely loose out to competing products in many sectors if only competing products are allowed to compete. If this means that the courts need to give them an extra push by handicapping Microsoft for a certain amount of time, that's fine. It doesn't mean everyone will now have to publish their APIs or not have premier partnership programs, just that only Microsoft, the abusive monopoly, won't be able to.
There are several factors that determine which browser wins the Wars. Specifically:
Distribution channels. The DOJ trial FoF made a great case for the fact that most users will not switch from the browser that comes installed with their PC or given to them by their ISP. The Slashdot crowd can find this hard to relate to, as we're used to installing our own software, but an increasing number of Web users don't, won't or can't. Hopefully, the final ruling on the case will limit MS's ability to dominate distribution channels, and even if it doesn't, the attention that has been drawn to these practices will probably give OEMs and IAPs the guts to challenge Microsoft and install Navigator instead of IE for their users.
Componentization and Embedability. Software developers want an HTML (and friends) layout component to embed into their applications. Intuit uses IE in Quickbooks. SoftQuad use IE in HoTMetaL and XMetaL. Others do similar things. Mozilla is uniquely poised to dominate this area. Gecko is an incredibly lightweight, powerful component that is much better than IE's counterpart in a million different ways. The wonderful thing about this is that a software developer's decision to use Gecko over IE's HTML control is based entirely on technology considerations - they don't care about user interface eye candy or marketing, they only want a component that will work well with their software. If Mozilla/Gecko is released soon, I expect it to dominate this market.
Standards Compliance. This battle has to be won, and it has to be won now. Microsoft has no reason to support Web standards, and never will. It's plainly against their best interests. Again, looking at the FoF we see that the Web is considered a development platform, and Microsoft is Microsoft because it controls platforms. But anyone playing the Web game that has their head screwed on right would rather die than let that happen.
However, in addition to all of the above there exist the very important factors that govern the success of any software product: features, stability, system footprint and general technical excellence. Mozilla has the upper hand here (except for stability, for the time being, but we'll have to be patient).
Who will prevail? It all depends on when we'll actually get a release of Mozilla. As JWZ said when he quit Netscape, Mozilla failed in many ways because it was grossly mishandled on the part of Netscape. Mozilla got a terrible codebase and Netscape expected thousands of hackers to dive in and fix the mess. It is interesting to find that the work got done a lot faster on the parts of Mozilla that were written from scratch (e.g. NGLayout) than anything else: Netscape's code was such complete crap it was difficult to hack, and most people didn't even bother. AFAIK, Mozilla as it exists today was almost completely written from scratch. That's how bad the code was.
What we need to do now is hold our breath and hope for a release, SOON. Netscape is planning to have the beta go live Real Soon Now (I remember they asked me for a "Developer Testimonial" back in September to include in the launch... and that was supposed to be rushed in...), which basically means that we'll have a beta some time in January, with a release some time in February or March. Amen. It's late, but it might not be too late.
What will happen then? If all goes well, ISP start bundling Mozilla with their sign-on packs, because the average ISP employee is going to drool over Mozilla. I used to work in an ISP, and I can tell you that User Friendly is not far off the mark. AOL might be bound into using IE in its software for another year, and probably have plans to continue to do so longer than that, but if the rest of the world rocks the boat one way, AOL will do the same. AOL picked IE over NS because IE could be embedded into AOL's software. Now, Netscape can do the same, and with more ease, and with smaller download times. AOL will not switch over fast, even though it owns Netscape, but I believe that unless Mozilla fails completely, it will switch soon.
Also, Mozilla will completely dominate the browser market for anyone not using Windows or MacOS within a few months. That might be a small (tiny?) percentage of the market, but it's market segment that carries a lot of weight - a lot of us are developers and have a direct influence on content creation and corporate acceptance.
What Netscape has to do is aggressively market Gecko as an alternative to MS's HTML layout component, and FAST. If they can get companies like Intuit to use Gecko, we will get more and more content developed for NS (i.e. standards-based) instead of IE.
But I think the major driving force behind browser acceptance will be content providers. As someone pointed out, Joe Webmaster might not care about standards and just preview his webpage on the latest IE, but most professional Web developers (i.e. those that work for search engines, portals and other high-traffic sites) will, and most of them are sick and tired of porting code from NS to IE and back. They'll start doing things with Mozilla that IE can't do, and then the market dynamics will pressure ISVs, IAPs, OEMs and individual users to use NS over IE.
It all hangs in the balance, but it's very important that Mozilla is released soon. We're in the final stretch - I hope it's sonner rather than later.
Let's assume that Transmeta (or the MIT group someone else mentioned around here) is indeed working on something like this.
A processor that has a core that uses an ISA of one-cycle or few-cycle instructions and all sorts of goody good technology to make it run fast. This will do vanilla, FP and SIMD amongst other things.
Around that core a translator that takes code in various ISAs and translates it and then profiles it so that it makes sense to the core.
In other words, one chip to rule them all, one chip to find them, one chip to bring them all and in the darkness bind them:-)
Benefits? This chip can be re-programmed on the fly to do anything: "Normal" CPU tasks (memory access, bit blits, integer ops), FP-intensive maths, 3D rendering, Multimedia encoding/decoding. However, having a CPU that can be re-programmed to do these things is pretty useless. Reprogramming it will probably take time measured in the thousands of cycles and the scheduling would be a nightmare. But...
Stick two or four of these babies onto a motherboard and configure them depending on system useage... keep one in "traditional" CPU mode, to do the basic stuff, i.e. run the OS code most of the time. Then, when you run a 3D rendering app or game, you can switch one of these to 3D rendering mode and it will take over the load. Fire up a DVD movie, and another switches to MPEG decoding. And the cherry on the cake.... Fire up a non-native app, and another CPU switches over to the different instruction set. And you get all of this without feeding anything through a bus.
We're talking about a new approach to computer architecture. Instead of having several high-performance chips of different types inside a computer, with one at the center and the rest residing at the other end of a bus, we get an array of configureable chips that take the role of any specialised processor as the need arises. Which means that if you stick enough of these babies onto a board, you have a system that is good at *everything*. Good at maths, good at 2D graphics, good at 3D graphics, good at multimedia, able to run ANY legacy code, good at anything else they come up with later...
So let's hear it from someone more knowledgeable: is this feasible?
the base station can handle POTS and ISDN (haven't you noticed it's dead? must be pretty smelly by now)
ISDN is dead in the US and only the US. I know DSL is a lot cheaper and better over there, but here in Greece you can get an ISDN line from the (only) telco for about $20 a month (the time charges are the same as for POTS) and the ISP accounts cost a trivial extra amount compared to a 56k modem dialup. AFAIK it's also taken off in Germany and other places in mainland Europe. I also spend half the year in the UK where prices from telcos (actually I've only checked out BT) are crazy, but over here ISDN is the best solution for a digital dialup connection. I'm getting an ISDN line Real Soon Now and I'm really looking for devices that support it. I also have a DECT wireless phone that would work with the FreePad base station (doing what, I don't know, but hey, it works).
This statement is bizarre. These things cannot exist but in a democracy. If there is no way to hold leaders accountable, then you can not stop them from silencing speech or making people "disappear".
I'll stick by my point here. What you're saying is that what is good about our system is that leaders are accountable and in direct contact with an electorate that has the internationally ratified human rights and not that leaders are elected in a nation-wide poll. You can definitely have elected leaders without human rights; look at ancient Athens. A great number of intellectuals were ostracized (lit. deported for life) or executed in Athens when they spoke against the government or were thought to be amassing too much individual power. Solon himself was condemned to forced suicide because he was believed that his teachings contradicted accepted religious practises.
What about today? Yes, you can. The media, for one thing, is an imperfect tool for informing the public about its government's actions. As to wether the average American knows what goes on in government, ask Oliver Stone for his opinion. Who influences government decisions? Lobbyists do, and the ones with the money usually do most, because especially in America electoral campaigns don't come cheap. As for human rights, America was segregated until only a few decades ago, and had elected government long before abolishing slavery.
Can you have human rights without democracy? Though I cannot think of any historical examples, I do not see why you can not. Maybe you cannot ensure them perpetually, but Nazi Germany shows that the same holds for democracy. In the United Kingdom today, there is no constitution; a law could be voted tomorrow that abolishes parliament, and the Queen is still the ruler of the state and commander of the military, and has the legal right to do whatever she wants with either. Just because you put something down on paper does not mean that you ensure its perpetuity or its practise. There are many ways to bring down a constitution, and the most dangerous do not involve any bloodshed at all.
Also note that accountability itself is useless unless you have empowerment. Giving the people the right to criticize is pointless unless:
(a) The people have a clear idea of how government works (how many people in America would you say have that?)
(b) The people have the right to enforce a specific course of action (in America they can only pick between two parties; if both refuse to address an issue - and there are many reasons why they may wish to do this - the public can do nothing about it)
(c) The people get a perfectly clear picture of what the government is doing in all areas (which is sometimes counter-productive in itself, such as in some issues of national security, and anyway is a pipe dream; 99% of the people only get 30-second clips on TV)
(d) The people are involved (how many people actually vote in the States? How many vote in elections for regional government especially?)
These things are important, yes. But we do not have them today, not in practice.
With a significantly reduced reign and less power. And they can be removed from power without bloodshed. And anyone can become a politician.
Yes, but can anyone become a succesful politician? And can someone become a politician that can be elected to a post that makes decisions about his life in general? Can a Serb be elected to a post that will influence wether his home town is bombed?
The fact that you prefix "supposed" in front of "ethnic cleansing" says about your bias. I assume you get news from the international wire services in Greece.
The fact that you take these wire services at face value shows that you are biased in their favour. Personally, I do not disagree that huge atrocities have been comitted by the Milosevic administration. But equally questionable actions have been taken by the KLA, who are supported and equipped by the US government, and are NATO's primary source of intelligence about anything that goes on in Kosovo. What I strongly disagree with is NATO's place as an international peace-keeping organization that can act punitively without the ratification of the UN Security Council (as was not the case in the Gulf War) and its unwilligness to even aknowledge the confirmed, prolonged and institutionalized ethnic cleansing that is executed by its own member states (e.g. Turkey).
The point I'm trying to make is this: Elected government is a nice idea, but it's not a pannacea (sp? - I can write it in Greek if you want:-)). A lot of the arguments you made were valid, but mostly they criticize or praise this or that administration. My objection is to the fact that elected government is the source of all that is good in our civilization. No. I believe that human rights, education, a good (=slim, efficient and limited to where it's really needed) welfare system and a free market economy are much more important. These are what can really abolish elites; by allowing anyone to succeed in any way he sees best, and guaranteeing that he doesn't trample on others on his way up. None of these are things that we have implemented successfully in our civilization; and none of these are directly tied to elected government. If Lucas wants to show that an elected government can fail in all of the things it's supposed to be good at, I say let him; his accusations are not unfounded.
What I found most disturbing in the whole affair, and the only point I will address, is that Brin's objection is Lucas presenting a story that supports the opinion that democracy might be inherently bad.
This being an opinion I currently hold, I would like to offer a couple of comments. Brin seems to be confusing several of the underlying concepts of American society and over-generalizing about how your political system works.
Democracy is not a cut-and-paste solution. Democratic governments vary greatly, and the principle behind democracy is elected government. A lot of other things that exist in American democracy and various other types of government in the Western world, such as human rights, international law, freedom of speech and expression etc. have little to do with democracy in itself.
I am not an American. I am Greek, and I live in Athens, the place where democracy was born. I would like to point out that the ancient Athenian democracy came with little of the sprinklings and egalitarian human-rights laws that come with most modern forms of government. And it didn't work well. It didn't work well at all, and it could be argued that democracy was largely to blame for Athens' fall and decline after the Pelloponesian War.
I've also read the Iliad and the Odyssey as well as several plays by Evripides (in the original, no less!) and would like to wonder about how they differ with the kind of stories Mr. Brin advocates.
For one thing, Mr. Brin agrees that a story must have heroes. And, believe it or not, this is the way history works. I'd like to see someone come up and disagree with me when I say that a handful of people have made an incredible impact on the history of human civilization while most people did indeed play the role of un-named spear-carriers. Especially in the case of war, the decision to go to war is usually taken by a small group of leaders to whom many have sworn allegiance by one means or another. It was the American government's decision to start the war in Kosovo over supposed ethnic cleansing just as it was Agamemnon's decision to start the war in Troy over the abduction of Helen.
The only moral objection that Mr. Brin brings up is how the elite is selected. The only difference he really sees between the Homerian demi-gods and everyday-Joes-cum-heroes is that one is selected by fate / hereditary rights / genetics while the other is selected by a handful of people around him. For Mr. Brin, the deciding difference between Agamemnon Atreides and Bill Clinton is that Agamemnon was king because he was the son of Atreus, while Mr. Clinton is president because of the small percentage of the American public that actually voted, more than half chose him over the other candidate.
The fact that he's at least partially responsible for bombing the houses of people in Serbia, half way around the world from the Oval office, is not the problem. Supposedly, our beloved democracy would have struck him down in his place if his action was morally reprehensible.
Please get me straight. Although I do take interest in these matters, I do not wish to debate the moral right of this or that leader to wage war against another nation. That is a matter of a different discussion. What I am debating is Mr. Brin's point that all the good in society comes from the fact that our leaders are elected.
Let me tell you something you might have forgotten. The Nazi government was elected, through due democratic process. The public loved them. For the average German in the 30s, a government that promised freedom from the economic hell imposed by the winners of WWI and getting rid of the Jewish commercial elite that they believed was the source of all their problems (there, another elite creeps into place). HITLER AND THE NAZI PARTY HAD A DIRECT MANDATE FROM THE GERMAN PEOPLE, AND YET THEY ARE STILL THE CAUSE FOR SOME OF THE WORST ATROCITIES EVER COMMITTED.
Mr. Brin supports that power corrupts, and that despots invariably become egotistical and power-hungry, incapable of acting in the best interests of their people. Just because a (relatively small) bunch of people picked an elected president over his opponent does not mean that the above cannot apply to him as well!
Democracy is, IMHO, just another way to pick an elite. But the way the world is governed at the time being means we NEED an elite. We just switched from passing the crown down from father to son, and are now passing the mandate from politician to politician.
The average citizen might have an opinion, and even a valid and just way of thinking about how the country should be run. But he might not. Putting the decision of one political party over another in the hands of the public does NOT mean you get good government.
This is Slashdot, a site for geeks, people who see ourselves as an intellectual elite. Saying that if more people pick a Democrat over a Republican (or the other way around) means we get good government is like saying that having people pick Windows over Linux (which they do, mostly, and the reasons, though bad, apply equally to democracy: lack of choice, lack of support, bad media coverage, lack of education and awareness, lack of interest) means they get good computing.
If George Lucas wants to thrash democracy, let him. You might disagree with him (and me), but it's a valid point and not unethical or morally reprehensible. Mr. Brin's point about GL thrashing a culture that has been good to him is invalid. GL is thrashing ELECTED GOVERNMENT, not the institutions that allow freedom of expression and opportunities for financial success. The two do not go together hand-in-hand. You can have one without the other. And in ALL forms of government to date, be it monarchy, aristocracy, communism, tyrrany, democracy or anything else, (perhaps not socialism? But the examples are few and far apart, and hard to judge) an ELITE has governed the masses, who have had little say in government except when they rebelled or went to the polls to exchange one elite for another. The question here is how an elite is selected. Elected government might be the best we have so far , but it is not close to being a good and effective system. You can judge individual elites (e.g. Nazis were bad, JFK was good, Agamemnon was bad, Pericles was good etc.) but you will find that the way they were brought into power has little to do with their effectiveness. Mr. Brin's objection to this I find appalling.
Slashdot Mirror
Uhm, nope. If Joe Sixpack wants his kids to play games, he'll get them a PlayStation or DreamCast just like all of their friends have.
Nope again. Games existed for DOS/Windows long before DirectX and related goodies in huge numbers. Games are written for Windows because of the huge installed user base. Developers put up with DOS's terrible memory model (I was about to say memory management but reconsidered) and built drivers for every sound and graphics card from scratch for every game or line of games. They put up with this because people used DOS and were going to buy games for DOS, and didn't care much about how easy it was to develop for it.
The availability of "productivity" apps, like an office suite, is a much more important factor in swaying users to Linux. Conquering the home market is hard, and not the first priority. Conquering the small- and mid-size workstation market in the office is a much more likely target.
"It can't do anything my PC can do except e-mail and Web browsing" isn't the problem. The problem is that it won't do the new things your PC will do one day, that you never even envisaged when you bought your PC. Let me explain.
A lot of people in this discussion have re-stated the most common argument against Internet appliances: That they can't do many of the things you can do with a PC. Others have re-stated the common counter-argument: That most people don't want to do these things.
However, I think there is a much more important problem with these machines: That they have a limited function.
The thing about a PC, or any computer, is that you might design it to do one thing (e.g. word processing, games playing), and people might buy it for this thing, but it is very easy for it to do something completely different in the future. The Internet is a good example. Before the Net explosion, people bought PCs to do their word processing, spreadsheets, maybe play a game or two. When the Internet came along, they used those same PCs to access it; that's the real power of a non-crippled computer: It can do things that you never planned for.
This thing might be great for e-mail and Web browsing, and people who only need to do these things will use it. But when they want to do something new and different (god knows what!) a year down the line, it will be useless. While those who own a PC will, at the most, install a piece of software or buy a small peripheral.
This may seem like a small problem for the individual user, but it is a HUGE problem for the market as a whole. If the PC is replaced by specialised appliances, deploying new technologies will not just involve releasing some software and maybe a Web site. It will involve getting an entirely new appliance to all of your prospective users.
To illustrate, imagine this scenario: Roll back history for about 2 years, and imagine that everyone on the Internet had an Internet Appliance instead of a PC. And someone comes up with MP3s. Could the technology have spread as it has in our world? No, because unless this person managed to convince appliance manufacturers to include it in the next version, and waited for that version to actually get out and occupy a significant portion of the market, nobody could use it.
Internet appliances raise the barrier of entry for new technologies. This is their main disadvantage, and it's a scary prospect.
The answer to your question is "maybe", but I sincerely doubt it.
Crusoe, as the article mentions, uses VLIW. VLIW normally depends on the compiler ordering instructions in the most effective manner, and taking care of issuing them to the most appropriate pipeline. This takes a lot of complexity out of your chip. You get the advantages of super-scalar design without all of that multi-issue hardware that bloats up your processor.
The downside is that your VLIWs are hardcoded for a specific processor design. The Crusoe has four pipelines, each of them specialised for certain functions. If, in a future version of the Crusoe, they decide to change the number and/or type of these pipelines, the optimal structure of these VLIWs will be completely different.
Now, as I said, normally making the VLIWs (or "molecules" as TransMeta calls them) depends on the compiler. This means that every time you change the internals of the processor implementation, you need to completely re-design your compiler as well - not really possible. But in Crusoe, it's the Code Morphing software's job. And since this comes part & parcel with the CPU itself, you'll always get super-scalar optimization that works with your CPU.
Of course, this brings us to another question. If Crusoe can emulate x86, can't it emulate other instruction sets? Of course it can. But I'm not talking about Sparc, Alpha or MIPS here.
Imagine this scenario:
As someone else mentioned here, let's say you put, say, 8 Crusoes on a motherboard and have the Code Morphing software handle part of the MP. And let's assume that these processors initially act as CPUs in SMP mode. And now you start Quake, and you want some polygon-pushing power. What if the Code Morphing software picks this up, and automatically switches the ISAs of a couple of your processors to GPU mode. Bingo, you've got 3D acceleration WITHOUT buying an expansion card, WITHOUT going through the system bus, and WITHOUT investing into a big bunch of computation power that sits idle 99.9% of the time.
Need a DSP to do sound processing? An MPEG decoder chip to play DVDs? Just switch one or more of your processors into that mode. Need massive computation power for mathematical calculations? Switch them all back to x86 mode.
All this can be done by the Code Morphing software. Granted, maybe you'd like to have different kinds of processors (i.e. some of them might have more FP pipelines) to make this more efficient, but it's all transparent to the software.
I think this could be the killer feature that brings Crusoe out of the mobile market and into the desktop/workstation/server market.
A lot of people are commenting on how possible sanctions on MS (esp. those concerning limiting its ability to cut deals with licensees and hiding APIs) would set a precedent and generally create a very difficult environment to work in for other companies.
But it's important to note that we're not talking about other companies. Other companies don't have a wide range monopoly (i.e. a monopoly on desktop OSs that feeds monopolies on office software, server OSs, back-end tools, development tools, browsers or anything else).
Let's face it, the only way for Microsoft to stop stifling innovation in the computer industry is for its monopoly to be completely deposed in a few sectors. They have to loose, say, the browser wars and the OS wars. Once that happens, they'll have to play fair with others because it will be impossible to offer a Microsoft-only solution as once can do now.
Today, you can set up an entire company network using only Microsoft software. Win98 + WinNT + Office + BackOffice + IE + VisStudio covers practically anything you need, with only niche markets available for others (and since they ARE niche markets, they have to work with the MS non-niche products that even niche clients will use in addition to niche products - not everyone needs CAD, but even those who need CAD need word processing).
You can't make them loose these monopolies through a direct court action, but it's not necessary. We all know that Microsoft will definitely loose out to competing products in many sectors if only competing products are allowed to compete. If this means that the courts need to give them an extra push by handicapping Microsoft for a certain amount of time, that's fine. It doesn't mean everyone will now have to publish their APIs or not have premier partnership programs, just that only Microsoft, the abusive monopoly, won't be able to.
There are several factors that determine which browser wins the Wars. Specifically:
However, in addition to all of the above there exist the very important factors that govern the success of any software product: features, stability, system footprint and general technical excellence. Mozilla has the upper hand here (except for stability, for the time being, but we'll have to be patient).
Who will prevail? It all depends on when we'll actually get a release of Mozilla. As JWZ said when he quit Netscape, Mozilla failed in many ways because it was grossly mishandled on the part of Netscape. Mozilla got a terrible codebase and Netscape expected thousands of hackers to dive in and fix the mess. It is interesting to find that the work got done a lot faster on the parts of Mozilla that were written from scratch (e.g. NGLayout) than anything else: Netscape's code was such complete crap it was difficult to hack, and most people didn't even bother. AFAIK, Mozilla as it exists today was almost completely written from scratch. That's how bad the code was.
What we need to do now is hold our breath and hope for a release, SOON. Netscape is planning to have the beta go live Real Soon Now (I remember they asked me for a "Developer Testimonial" back in September to include in the launch... and that was supposed to be rushed in...), which basically means that we'll have a beta some time in January, with a release some time in February or March. Amen. It's late, but it might not be too late.
What will happen then? If all goes well, ISP start bundling Mozilla with their sign-on packs, because the average ISP employee is going to drool over Mozilla. I used to work in an ISP, and I can tell you that User Friendly is not far off the mark. AOL might be bound into using IE in its software for another year, and probably have plans to continue to do so longer than that, but if the rest of the world rocks the boat one way, AOL will do the same. AOL picked IE over NS because IE could be embedded into AOL's software. Now, Netscape can do the same, and with more ease, and with smaller download times. AOL will not switch over fast, even though it owns Netscape, but I believe that unless Mozilla fails completely, it will switch soon.
Also, Mozilla will completely dominate the browser market for anyone not using Windows or MacOS within a few months. That might be a small (tiny?) percentage of the market, but it's market segment that carries a lot of weight - a lot of us are developers and have a direct influence on content creation and corporate acceptance.
What Netscape has to do is aggressively market Gecko as an alternative to MS's HTML layout component, and FAST. If they can get companies like Intuit to use Gecko, we will get more and more content developed for NS (i.e. standards-based) instead of IE.
But I think the major driving force behind browser acceptance will be content providers. As someone pointed out, Joe Webmaster might not care about standards and just preview his webpage on the latest IE, but most professional Web developers (i.e. those that work for search engines, portals and other high-traffic sites) will, and most of them are sick and tired of porting code from NS to IE and back. They'll start doing things with Mozilla that IE can't do, and then the market dynamics will pressure ISVs, IAPs, OEMs and individual users to use NS over IE.
It all hangs in the balance, but it's very important that Mozilla is released soon. We're in the final stretch - I hope it's sonner rather than later.
So let me get this straight.
Let's assume that Transmeta (or the MIT group someone else mentioned around here) is indeed working on something like this.
A processor that has a core that uses an ISA of one-cycle or few-cycle instructions and all sorts of goody good technology to make it run fast. This will do vanilla, FP and SIMD amongst other things.
Around that core a translator that takes code in various ISAs and translates it and then profiles it so that it makes sense to the core.
In other words, one chip to rule them all, one chip to find them, one chip to bring them all and in the darkness bind them :-)
Benefits? This chip can be re-programmed on the fly to do anything: "Normal" CPU tasks (memory access, bit blits, integer ops), FP-intensive maths, 3D rendering, Multimedia encoding/decoding. However, having a CPU that can be re-programmed to do these things is pretty useless. Reprogramming it will probably take time measured in the thousands of cycles and the scheduling would be a nightmare. But...
Stick two or four of these babies onto a motherboard and configure them depending on system useage... keep one in "traditional" CPU mode, to do the basic stuff, i.e. run the OS code most of the time. Then, when you run a 3D rendering app or game, you can switch one of these to 3D rendering mode and it will take over the load. Fire up a DVD movie, and another switches to MPEG decoding. And the cherry on the cake.... Fire up a non-native app, and another CPU switches over to the different instruction set. And you get all of this without feeding anything through a bus.
We're talking about a new approach to computer architecture. Instead of having several high-performance chips of different types inside a computer, with one at the center and the rest residing at the other end of a bus, we get an array of configureable chips that take the role of any specialised processor as the need arises. Which means that if you stick enough of these babies onto a board, you have a system that is good at *everything*. Good at maths, good at 2D graphics, good at 3D graphics, good at multimedia, able to run ANY legacy code, good at anything else they come up with later...
So let's hear it from someone more knowledgeable: is this feasible?
ISDN is dead in the US and only the US. I know DSL is a lot cheaper and better over there, but here in Greece you can get an ISDN line from the (only) telco for about $20 a month (the time charges are the same as for POTS) and the ISP accounts cost a trivial extra amount compared to a 56k modem dialup. AFAIK it's also taken off in Germany and other places in mainland Europe. I also spend half the year in the UK where prices from telcos (actually I've only checked out BT) are crazy, but over here ISDN is the best solution for a digital dialup connection. I'm getting an ISDN line Real Soon Now and I'm really looking for devices that support it. I also have a DECT wireless phone that would work with the FreePad base station (doing what, I don't know, but hey, it works).
I'll stick by my point here. What you're saying is that what is good about our system is that leaders are accountable and in direct contact with an electorate that has the internationally ratified human rights and not that leaders are elected in a nation-wide poll. You can definitely have elected leaders without human rights; look at ancient Athens. A great number of intellectuals were ostracized (lit. deported for life) or executed in Athens when they spoke against the government or were thought to be amassing too much individual power. Solon himself was condemned to forced suicide because he was believed that his teachings contradicted accepted religious practises.
What about today? Yes, you can. The media, for one thing, is an imperfect tool for informing the public about its government's actions. As to wether the average American knows what goes on in government, ask Oliver Stone for his opinion. Who influences government decisions? Lobbyists do, and the ones with the money usually do most, because especially in America electoral campaigns don't come cheap. As for human rights, America was segregated until only a few decades ago, and had elected government long before abolishing slavery.
Can you have human rights without democracy? Though I cannot think of any historical examples, I do not see why you can not. Maybe you cannot ensure them perpetually, but Nazi Germany shows that the same holds for democracy. In the United Kingdom today, there is no constitution; a law could be voted tomorrow that abolishes parliament, and the Queen is still the ruler of the state and commander of the military, and has the legal right to do whatever she wants with either. Just because you put something down on paper does not mean that you ensure its perpetuity or its practise. There are many ways to bring down a constitution, and the most dangerous do not involve any bloodshed at all.
Also note that accountability itself is useless unless you have empowerment. Giving the people the right to criticize is pointless unless:
(a) The people have a clear idea of how government works (how many people in America would you say have that?)
(b) The people have the right to enforce a specific course of action (in America they can only pick between two parties; if both refuse to address an issue - and there are many reasons why they may wish to do this - the public can do nothing about it)
(c) The people get a perfectly clear picture of what the government is doing in all areas (which is sometimes counter-productive in itself, such as in some issues of national security, and anyway is a pipe dream; 99% of the people only get 30-second clips on TV)
(d) The people are involved (how many people actually vote in the States? How many vote in elections for regional government especially?)
These things are important, yes. But we do not have them today, not in practice.
Yes, but can anyone become a succesful politician? And can someone become a politician that can be elected to a post that makes decisions about his life in general? Can a Serb be elected to a post that will influence wether his home town is bombed?
The fact that you take these wire services at face value shows that you are biased in their favour. Personally, I do not disagree that huge atrocities have been comitted by the Milosevic administration. But equally questionable actions have been taken by the KLA, who are supported and equipped by the US government, and are NATO's primary source of intelligence about anything that goes on in Kosovo. What I strongly disagree with is NATO's place as an international peace-keeping organization that can act punitively without the ratification of the UN Security Council (as was not the case in the Gulf War) and its unwilligness to even aknowledge the confirmed, prolonged and institutionalized ethnic cleansing that is executed by its own member states (e.g. Turkey).
The point I'm trying to make is this: Elected government is a nice idea, but it's not a pannacea (sp? - I can write it in Greek if you want
What I found most disturbing in the whole affair, and the only point I will address, is that Brin's objection is Lucas presenting a story that supports the opinion that democracy might be inherently bad.
This being an opinion I currently hold, I would like to offer a couple of comments. Brin seems to be confusing several of the underlying concepts of American society and over-generalizing about how your political system works.
Democracy is not a cut-and-paste solution. Democratic governments vary greatly, and the principle behind democracy is elected government. A lot of other things that exist in American democracy and various other types of government in the Western world, such as human rights, international law, freedom of speech and expression etc. have little to do with democracy in itself.
I am not an American. I am Greek, and I live in Athens, the place where democracy was born. I would like to point out that the ancient Athenian democracy came with little of the sprinklings and egalitarian human-rights laws that come with most modern forms of government. And it didn't work well. It didn't work well at all, and it could be argued that democracy was largely to blame for Athens' fall and decline after the Pelloponesian War.
I've also read the Iliad and the Odyssey as well as several plays by Evripides (in the original, no less!) and would like to wonder about how they differ with the kind of stories Mr. Brin advocates.
For one thing, Mr. Brin agrees that a story must have heroes. And, believe it or not, this is the way history works. I'd like to see someone come up and disagree with me when I say that a handful of people have made an incredible impact on the history of human civilization while most people did indeed play the role of un-named spear-carriers. Especially in the case of war, the decision to go to war is usually taken by a small group of leaders to whom many have sworn allegiance by one means or another. It was the American government's decision to start the war in Kosovo over supposed ethnic cleansing just as it was Agamemnon's decision to start the war in Troy over the abduction of Helen.
The only moral objection that Mr. Brin brings up is how the elite is selected. The only difference he really sees between the Homerian demi-gods and everyday-Joes-cum-heroes is that one is selected by fate / hereditary rights / genetics while the other is selected by a handful of people around him. For Mr. Brin, the deciding difference between Agamemnon Atreides and Bill Clinton is that Agamemnon was king because he was the son of Atreus, while Mr. Clinton is president because of the small percentage of the American public that actually voted, more than half chose him over the other candidate.
The fact that he's at least partially responsible for bombing the houses of people in Serbia, half way around the world from the Oval office, is not the problem. Supposedly, our beloved democracy would have struck him down in his place if his action was morally reprehensible.
Please get me straight. Although I do take interest in these matters, I do not wish to debate the moral right of this or that leader to wage war against another nation. That is a matter of a different discussion. What I am debating is Mr. Brin's point that all the good in society comes from the fact that our leaders are elected.
Let me tell you something you might have forgotten. The Nazi government was elected, through due democratic process. The public loved them. For the average German in the 30s, a government that promised freedom from the economic hell imposed by the winners of WWI and getting rid of the Jewish commercial elite that they believed was the source of all their problems (there, another elite creeps into place). HITLER AND THE NAZI PARTY HAD A DIRECT MANDATE FROM THE GERMAN PEOPLE, AND YET THEY ARE STILL THE CAUSE FOR SOME OF THE WORST ATROCITIES EVER COMMITTED .
Mr. Brin supports that power corrupts, and that despots invariably become egotistical and power-hungry, incapable of acting in the best interests of their people. Just because a (relatively small) bunch of people picked an elected president over his opponent does not mean that the above cannot apply to him as well!
Democracy is, IMHO, just another way to pick an elite. But the way the world is governed at the time being means we NEED an elite. We just switched from passing the crown down from father to son, and are now passing the mandate from politician to politician.
The average citizen might have an opinion, and even a valid and just way of thinking about how the country should be run. But he might not. Putting the decision of one political party over another in the hands of the public does NOT mean you get good government.
This is Slashdot, a site for geeks, people who see ourselves as an intellectual elite. Saying that if more people pick a Democrat over a Republican (or the other way around) means we get good government is like saying that having people pick Windows over Linux (which they do, mostly, and the reasons, though bad, apply equally to democracy: lack of choice, lack of support, bad media coverage, lack of education and awareness, lack of interest) means they get good computing.
If George Lucas wants to thrash democracy, let him. You might disagree with him (and me), but it's a valid point and not unethical or morally reprehensible. Mr. Brin's point about GL thrashing a culture that has been good to him is invalid. GL is thrashing ELECTED GOVERNMENT, not the institutions that allow freedom of expression and opportunities for financial success. The two do not go together hand-in-hand. You can have one without the other. And in ALL forms of government to date, be it monarchy, aristocracy, communism, tyrrany, democracy or anything else, (perhaps not socialism? But the examples are few and far apart, and hard to judge) an ELITE has governed the masses, who have had little say in government except when they rebelled or went to the polls to exchange one elite for another. The question here is how an elite is selected. Elected government might be the best we have so far , but it is not close to being a good and effective system. You can judge individual elites (e.g. Nazis were bad, JFK was good, Agamemnon was bad, Pericles was good etc.) but you will find that the way they were brought into power has little to do with their effectiveness. Mr. Brin's objection to this I find appalling.