It's given that porting games to linux isn't economically viable for windows gaming shops. I hope Loki recovers, and of course the Carmack is going to keep porting Quake because he has plenty of money already, but for the most part, here and now in 2001, Linux gaming is in trouble.
I've never tried to use WINE (I game in windoze for now), but from what I read in that article yesterday they're making some really good progress. If they manage to get WINE running as well as or better than the real windows API, that could do a lot for the community all by itself. "Linux runs windows apps better than windows" sounds almost as good as "Linux has a lot of apps."
If WINE keeps progressing, it could be the shot in the arm that saves Linux gaming. It may not do anything for the Linux gaming industry, but everyone knew that trying to make money in it was risky.
It seems to me that this is cooler than binary for the same reasons CISC is better than RISC. Unless manufacturing changes dramaticly, it will probibally never catch on for the same reasons RISC did.
As far as "closer to e", I'm not an expert, but I would guess that in computing applications, wouldnt a truncation of e be more useful than a rounding most of the time?
Not to mention that if Intel did come out with a better product, I would EXPECT him to switch sides.
Tom was an Intel freak back in the old days. His opinion was basically "If you want performance, buy a Pentium MMX. AMD's chips are nice, but not nice enough." When AMD got the Athlons out, he started recommending them. If Intel comes out with something that beats the Athlon, then I would hope he'd recommend Intel again.
Tom continues to to promote AMD because he thinks that they're better than the P4 (and I agree). Now a days the reasons are a little more complex though.
In the previous two generations of PC hardware, things were a lot simpler. Back when all CPU's were socket 7, Intel chipsets were vastly, vastly superior to the competition, and there was really no reason to use anything else, ever. Intel's CPU's didn't have a specific relibiliy advantage, but they were always a year ahead of AMD in the speed department. Then came Socket 1 vs Super 7. Intel's P2's wern't that far ahead of the K6-2, but the BX chipset was amazing, and it was worth getting a celery or P2 to use it.
Nowadays, there is no BX. IMO every chipset sucks now (I'm still using a BX myself, currently with a celery 850). Intel chipsets cripple themselves with rambus, and the competition is only in the last six months or so really reaching Intel Quality. Again IMO, an ideal situation would be an Athlon in an i815, but of course that's not possible.
If the nForce turns out to be all that and a bag of chips, we'll finally get some homogeny back. But for now we have a bunch of CPU's with no clear leader until you look at price, and a slew of chipsets that frankly make me want to cry.
Re:The Lexus and the Olive Tree
on
Globalization
·
· Score: 1
Disclaimer: I've only read about two thirds of Friedmans book, for a class. I have read the end.
I think that Friedman's book is more wishful thinking then a picture of the globalist world. It seems to me that in reality, the strait-jacket helps the multinationals, and if a nation's standard of living improves it's a by-product (and maybe a temporary one at that).
I don't want to spend a lot of time on this, because I suspect Katz is going to cover it in part 2. But in a lot of places, Friedman's strait jacket isn't approprate, but the IMF and the WTO make countries reform thier economies to fit in it anyway, because it's what's best for the banker's wallets.
For example, nations without public debt problems but lots of private bankrupcy could be handle thier problems internally with a little IMF help. But this keeps all the profit from economic improvement in the hands of locals. Individual, wealthy locals yes, but that's not who the IMF is looking out for. The individual IMF members seem to be more interested in making huge amounts of money for the americian and multinational corperations that they're about to buy controling interests in.
Not that I'm a fan of the bill or anything, but if this is the only legislation that goes through as a result of 911, then civil liberties got lucky. It could have been a lot worse.
Making it easy to use is the distru's job, and I don't think that the hackers should have to deal with it.
I'd just as soon have the guys writing kernel and tool stuff just make a great OS than worry about how the user will interact with it. The high level X app writer is responsible for making stuff easy to use within that application, of course...
But the Distru makes the SYSTEM easy. They put everything in a happy installer, and give the user easy to understand, point and grunt instructions. Probibally more importantly, they also provide easy to use updates for the system to keep it secure and happy. And I don't mean RPM's here, I mean stuff to make RPM's even easier to use.
IMO, the best case scenario for a Linux for Users distru will involve a (manual or outomatic) weekly update check. The updater proggie downloads a file from somewhere, untars it into RPM's or whatever, figgures out what's approprate based on the current setup, and does it's thing with miminum user input. Of course, there's capibility for out of schedule emergency stuff, and a update re-run when the user changes the installation settings (like installing additional stuff from the CD).
Of course, people who do all these rolls should take the approprate mindset for the task at hand.
Obviously, there's no new ideas here, but it's a long road to full implementation. I believe the community would be doing the Wrong Thing if we started putting stopgap and kludged up things in Linux for ease of use/ease of sysadmin. They may be helpful in the short run, but these things always manage to screw a few people's enviorments up, and they usually become a burden down the road.
NASA's biggest problem (in my uninformed opinion) is that it doesn't know when to quit. Programs like ISS and shuttle replacement should have been either dropped or re-designed a long time ago, and it seemed that even then everyone I talked to knew it. I don't know if the top boys at NASA get glossed over by contractor pitches, or just roll over to congressional pressure to fast, or if they're just silly, but one of the jobs of an executive is to come down and say "no" when the time comes, before things get out of control.
As much as I hate to say it, I also think we should get out of manned space flight for a while. Right now, almost all our big science gains and almost all our spacelift is or could have been done with unmanned missions. The shuttle still needs replaced, but do we even need a shuttle? I would cancel the shuttle program, and not even start another replacement for 5-10 years.
Then, when the time comes, NASA's primary goal should be to eliminate the differences between orbital craft, sub-orbital craft, and aircraft. Star Trek aside, space shouldn't be a big deal exotic situation, but a normal part of the process of travel, just like flying is part of travel now.
To sum up: Dont be afraid to put the hammer down, and get rid of "middle" projects. NASA should have two mission concepts: Unmanned probes and spacelift that give us the best bang for our space-bucks, and long term avionic concept projects that work in baby steps to get really, really good space exploitation down the road.
Well, it isn't the killer performace piece I was hoping for, but still not a bad chipset. I'd like to get one just to get rid of my God damn SBLive card, if nothing else!
It's isn't really meaningful until we're looking at the end product from the mobo manufactures. Still, I wouldn't mind picking up one of the 220-D chipsets, a nice cheap thunderbird, and some of that fancy Dance Dance RAM. I'd just drop my GF1DDR in there until the GF3's got nice and cheap.
If they intercept a message, and it's plain text, they should be able to use it. If it's encrypted, but they decrypt using brute force methods, or if for some reason your private key is public (on a hacker board or somthing) and they get it, then they should be able to use it.
In other words, if you don't stop the govt from reading your email (or whatever), then why shouldn't they? What's bad is when the govt makes it illegal for you to try and stop them when they have a court order.
Not really. A code green "infected" box will stop transmitting, or will only send a few code green infection attempts. So while a server sending them does the same amount of traffic, the total number of servers goes down.
Run Copper? If you're gonna do that, you might as well just use eithernet. Get some active hubs or switches (do they make passive switches? Prolly not) and go house to house. Actually, that sounds kinda fun...
To me, it read like he was just explaining the concept of the slide rule for readers whe don't know how they work.
Developing these kinds of tools don't take intellegence, they're the product of ingenuity. None of those gunners cared about how the damn things worked from a mathmatical standpoint, all they cared about was that this baby did all the work in one step.
I think that MS (and anyone else) should have gotten some early warning before the vuneribility should have been posted, so they'd have a change to have a patch and all that, notify who they can, ect. I know that it wouldn't really solve the problem in the case of something as widespread and arcane as IIS, but still, just dumping a no-fix bug to the world isn't good unless the provider just refuese to do anything about it.
Having said that, I still believe that hiding this kind of information is not a good thing. Once a fix and a distru system is established, bug reports should remain available to the public.
In this case, MS could have had a fix on windowsupdate, and a generic setup for the SMS world ready to go, emailed SMS admins, and all that, at the same time as public disclosure.
Not too early. ATI dropped this at just the right time. That's because these supposedly early, unoptomized, buggy drivers perform at the same benchmark speeds that the final, less buggy but still un-optimized drivers will.
Anyone who expects ATI to come out with better drivers later is just setting himself up for disappointment again.
I didn't have a lot of time that day, so instead of trying to simulate a human, it simulated the responses of a taco. I called it "Dr. Taco," and it responded to everything you typed with a line feed. VARY HIGH QUALITY!!!
This is almost a year away, and I would imagine that they can get pretty close to 100% yields on these things by now even if operating at full capacity. And it's not like it costs anything to warehouse a few thousand of the things (a decent file cabnet goes for a couple hundred bucks), so these embedded guys should't have any troubles.
Of course, the NYT isn't going to mention that he rotted for two weeks without bail. The FBI and it's corprate backers know they may not win the legal battle, so they gotta try to scare the hell out of the tech crowd too...
"Code Blue" is also what hosptials say over the intercom when someone dies and then need a trama team / defibulator.
"Adult code blue, room 412. Adult code blue, room 412."
Perhaps it's time to phase out the idea of a "top level domain ruleset" in the first place. Think about it: it wouldn't take much effort to set the second tier name servers to querry based on the first letter of the last part of a domain name. If need be, we can have the "n" server ("officially" a primary server) to actually do a lookup on the second letter if there's too many names, and so on...
We could do this, make exceptions for the traditional TLDs but otherwise sell the first 2 names rather than just one. Then, if people pick a second "name word" that has never been used before, it becomes a nes TLD equivlent just by the defauslt of the system.
Thank you for pointing that out. I guess Caduceus1's comment is an invalid example now. Wait a minute, his example isn't invalidated... Oh, I see the problem: your comment was pointless.
Slashdot Mirror
I've never tried to use WINE (I game in windoze for now), but from what I read in that article yesterday they're making some really good progress. If they manage to get WINE running as well as or better than the real windows API, that could do a lot for the community all by itself. "Linux runs windows apps better than windows" sounds almost as good as "Linux has a lot of apps."
If WINE keeps progressing, it could be the shot in the arm that saves Linux gaming. It may not do anything for the Linux gaming industry, but everyone knew that trying to make money in it was risky.
It seems to me that this is cooler than binary for the same reasons CISC is better than RISC. Unless manufacturing changes dramaticly, it will probibally never catch on for the same reasons RISC did.
As far as "closer to e", I'm not an expert, but I would guess that in computing applications, wouldnt a truncation of e be more useful than a rounding most of the time?
Not to mention that if Intel did come out with a better product, I would EXPECT him to switch sides.
Tom was an Intel freak back in the old days. His opinion was basically "If you want performance, buy a Pentium MMX. AMD's chips are nice, but not nice enough." When AMD got the Athlons out, he started recommending them. If Intel comes out with something that beats the Athlon, then I would hope he'd recommend Intel again.
Tom continues to to promote AMD because he thinks that they're better than the P4 (and I agree). Now a days the reasons are a little more complex though.
In the previous two generations of PC hardware, things were a lot simpler. Back when all CPU's were socket 7, Intel chipsets were vastly, vastly superior to the competition, and there was really no reason to use anything else, ever. Intel's CPU's didn't have a specific relibiliy advantage, but they were always a year ahead of AMD in the speed department. Then came Socket 1 vs Super 7. Intel's P2's wern't that far ahead of the K6-2, but the BX chipset was amazing, and it was worth getting a celery or P2 to use it.
Nowadays, there is no BX. IMO every chipset sucks now (I'm still using a BX myself, currently with a celery 850). Intel chipsets cripple themselves with rambus, and the competition is only in the last six months or so really reaching Intel Quality. Again IMO, an ideal situation would be an Athlon in an i815, but of course that's not possible.
If the nForce turns out to be all that and a bag of chips, we'll finally get some homogeny back. But for now we have a bunch of CPU's with no clear leader until you look at price, and a slew of chipsets that frankly make me want to cry.
Disclaimer: I've only read about two thirds of Friedmans book, for a class. I have read the end.
I think that Friedman's book is more wishful thinking then a picture of the globalist world. It seems to me that in reality, the strait-jacket helps the multinationals, and if a nation's standard of living improves it's a by-product (and maybe a temporary one at that).
I don't want to spend a lot of time on this, because I suspect Katz is going to cover it in part 2. But in a lot of places, Friedman's strait jacket isn't approprate, but the IMF and the WTO make countries reform thier economies to fit in it anyway, because it's what's best for the banker's wallets.
For example, nations without public debt problems but lots of private bankrupcy could be handle thier problems internally with a little IMF help. But this keeps all the profit from economic improvement in the hands of locals. Individual, wealthy locals yes, but that's not who the IMF is looking out for. The individual IMF members seem to be more interested in making huge amounts of money for the americian and multinational corperations that they're about to buy controling interests in.
Not that I'm a fan of the bill or anything, but if this is the only legislation that goes through as a result of 911, then civil liberties got lucky. It could have been a lot worse.
I'd just as soon have the guys writing kernel and tool stuff just make a great OS than worry about how the user will interact with it. The high level X app writer is responsible for making stuff easy to use within that application, of course...
But the Distru makes the SYSTEM easy. They put everything in a happy installer, and give the user easy to understand, point and grunt instructions. Probibally more importantly, they also provide easy to use updates for the system to keep it secure and happy. And I don't mean RPM's here, I mean stuff to make RPM's even easier to use.
IMO, the best case scenario for a Linux for Users distru will involve a (manual or outomatic) weekly update check. The updater proggie downloads a file from somewhere, untars it into RPM's or whatever, figgures out what's approprate based on the current setup, and does it's thing with miminum user input. Of course, there's capibility for out of schedule emergency stuff, and a update re-run when the user changes the installation settings (like installing additional stuff from the CD).
Of course, people who do all these rolls should take the approprate mindset for the task at hand.
Obviously, there's no new ideas here, but it's a long road to full implementation. I believe the community would be doing the Wrong Thing if we started putting stopgap and kludged up things in Linux for ease of use/ease of sysadmin. They may be helpful in the short run, but these things always manage to screw a few people's enviorments up, and they usually become a burden down the road.
As much as I hate to say it, I also think we should get out of manned space flight for a while. Right now, almost all our big science gains and almost all our spacelift is or could have been done with unmanned missions. The shuttle still needs replaced, but do we even need a shuttle? I would cancel the shuttle program, and not even start another replacement for 5-10 years.
Then, when the time comes, NASA's primary goal should be to eliminate the differences between orbital craft, sub-orbital craft, and aircraft. Star Trek aside, space shouldn't be a big deal exotic situation, but a normal part of the process of travel, just like flying is part of travel now.
To sum up: Dont be afraid to put the hammer down, and get rid of "middle" projects. NASA should have two mission concepts: Unmanned probes and spacelift that give us the best bang for our space-bucks, and long term avionic concept projects that work in baby steps to get really, really good space exploitation down the road.
OK, to me this looks like the work of those photoshoping freaks at somethingawful.com.
It's isn't really meaningful until we're looking at the end product from the mobo manufactures. Still, I wouldn't mind picking up one of the 220-D chipsets, a nice cheap thunderbird, and some of that fancy Dance Dance RAM. I'd just drop my GF1DDR in there until the GF3's got nice and cheap.
If they intercept a message, and it's plain text, they should be able to use it. If it's encrypted, but they decrypt using brute force methods, or if for some reason your private key is public (on a hacker board or somthing) and they get it, then they should be able to use it.
In other words, if you don't stop the govt from reading your email (or whatever), then why shouldn't they? What's bad is when the govt makes it illegal for you to try and stop them when they have a court order.
Not really. A code green "infected" box will stop transmitting, or will only send a few code green infection attempts. So while a server sending them does the same amount of traffic, the total number of servers goes down.
Run Copper? If you're gonna do that, you might as well just use eithernet. Get some active hubs or switches (do they make passive switches? Prolly not) and go house to house. Actually, that sounds kinda fun...
Developing these kinds of tools don't take intellegence, they're the product of ingenuity. None of those gunners cared about how the damn things worked from a mathmatical standpoint, all they cared about was that this baby did all the work in one step.
I think that MS (and anyone else) should have gotten some early warning before the vuneribility should have been posted, so they'd have a change to have a patch and all that, notify who they can, ect. I know that it wouldn't really solve the problem in the case of something as widespread and arcane as IIS, but still, just dumping a no-fix bug to the world isn't good unless the provider just refuese to do anything about it. Having said that, I still believe that hiding this kind of information is not a good thing. Once a fix and a distru system is established, bug reports should remain available to the public. In this case, MS could have had a fix on windowsupdate, and a generic setup for the SMS world ready to go, emailed SMS admins, and all that, at the same time as public disclosure.
Not too early. ATI dropped this at just the right time. That's because these supposedly early, unoptomized, buggy drivers perform at the same benchmark speeds that the final, less buggy but still un-optimized drivers will. Anyone who expects ATI to come out with better drivers later is just setting himself up for disappointment again.
I didn't have a lot of time that day, so instead of trying to simulate a human, it simulated the responses of a taco. I called it "Dr. Taco," and it responded to everything you typed with a line feed. VARY HIGH QUALITY!!!
This is almost a year away, and I would imagine that they can get pretty close to 100% yields on these things by now even if operating at full capacity. And it's not like it costs anything to warehouse a few thousand of the things (a decent file cabnet goes for a couple hundred bucks), so these embedded guys should't have any troubles.
Of course, the NYT isn't going to mention that he rotted for two weeks without bail. The FBI and it's corprate backers know they may not win the legal battle, so they gotta try to scare the hell out of the tech crowd too...
"Code Blue" is also what hosptials say over the intercom when someone dies and then need a trama team / defibulator. "Adult code blue, room 412. Adult code blue, room 412."
Everything here is unsigned, could you maybe use size_t instead? I'd check it out myself but I'm like, lazy and stuff.
With an L, an A, an M, to the E!!!
And have The Doctor be played by Christfor Walken!!! YES!!!
Sounds very nice, but it doesn't mean a thing if it doesn't push bits as fast as the old guys. Or if they can't get the thing to market.
We could do this, make exceptions for the traditional TLDs but otherwise sell the first 2 names rather than just one. Then, if people pick a second "name word" that has never been used before, it becomes a nes TLD equivlent just by the defauslt of the system.
Thank you for pointing that out. I guess Caduceus1's comment is an invalid example now. Wait a minute, his example isn't invalidated... Oh, I see the problem: your comment was pointless.