$15/gal gas is a bit absurd, but, in a way, he's got the right idea. The inability to use large trucks and other gas guzzlers for the distribution of food would cause industry to choose an alternative which is more economical. It's not just the population at large that would have to adapt, it's industry as a whole. You'd probably see a lot more mass transit spring up in areas, because suddenly it would be a requirement for a LARGE percentage of the population. There would, of course, be food distributors who would choose a cheaper method of transportation, like rail, that would be able to offer food cheaper than their competition (who would simply raise prices to pay for gas) and people would buy their food instead of the expensive shit. Likewise, the increased dependence on something like rail (which can simply tap into the energy grid for power) would probably result in the public wanting high speed rail for the transport of goods... in short, if gas prices went to $15/gal in a slow, but, steady rate, these changes would happen... and people would only starve if the government refused to provide incentives to hasten the changes that would be required as a result of expensive gas.
It's all a question of political will. We live in a society where we can build gigantic structures that touch the sky and tunnel through the earth. Changing our infrastructure from road to something else is barely a challenge worth of our combined intellects.... but only if we actually give a shit enough to try in a coordinated and not-half-assed manner...
Maybe I'm just not jaded enough, but I didn't interpret his statement to be directed at the 17 year old... his statement seemed to be directed at the ever popular "they". As in, this is old news, and if a kid can do it (albeit a very smart kid) then why the fuck aren't "they" doing it in larger scales, and why are "they" still burning coal and oil ffs? ... unless I missed the whoosh...
I know that graphics-guru's used to have to write a lot of inline assembly and other optimizers in order to get certain parts of their code to run a little bit faster. So, the majority of people won't need to delve into that kind of depth... but there are some who push the envelope a bit and who, traditionally, needed to get their hands dirty on the low level stuff. No idea if this is still the case... maybe someone a little more involved in pushing the envelope could answer... Mr. Carmack, are you out there?
yes, and if you run an I-don't-have-to-outrun-the-bear-just-you experiment on a species for a couple generations, they will eventually evolve some mechanism for going faster and faster. Which is what "happened" here... except, partway into the beginning of the experiment, they realized that they didn't have to do any real outrunning, they just needed to climb a tree. So, naturally, the ones that were better at climbing trees survived and had more chilluns... thems chilluns had the same features as their better-surviving parents (shorter legs more suited for climbing) and so they survived "better" than their neighbours whose parents didn't have such short legs (but managed to reproduce anyway...) Then these lizards went on to have chilluns, but their neighbours, who weren't quite as good at climbing, were yesterdays lunch for the predator...
mutations give the genetic diversity (i.e. differing leg lenghts) and natural selection ensures that a certain group survives... that group has kids, who have kids, etc... and eventually, the differing genetic "chunks" that helped the species thrive get propagated from parent to child, and eventually becomes dominant.
I know this thread is a couple days old... but I feel bad for the guy posting. No one actually answered his question.... take the parent, for example. He gives a lot of good reasons why the OP should take a B.Sc. in either comp sci or math... totally missing the fact that the OP already has one and is talking about getting an M.Sc. in Game Programming. (no offense adisakp... 95% of the posters in this "story" missed that...)
It's probably too late to answer effectively (ie. no one will read this) but I'd say that the parent is right in his reasoning, but needs to rewrite his post to suggest a masters in CS and not a bac...
My suggestion would be to do the M.Sc. in comp sci, but take courses that mimic/are those given in the gaming M.Sc. and gear your thesis around something that the game companies would find useful, and that the non-gaming companies would not dismiss out of hand. Perhaps an optimized algorithm for a transport mechanism that's faster than TCP but more reliable than UDP... or an optimized algorithm for some graphics raytracing/rasterizing/whatever techniques.
In that way, you'd still be taking what you are interested in, but you wouldn't risk having some dolt think that you're not the real deal just because your degree is in game programming. With the comp sci M.Sc. geared towards game programming, you'll be able to show the interviewers in the gaming companies that you are "t3h sh1t" and when you eventually go somewhere else, they'll just see the M.Sc. Comp. Sci. on your VC and they won't immediately assume that you're just sh1t... good luck!
We are at a total loss to understand how this policy has developed, who is behind it and why there is such haste in enacting it into law -- with little if any public debate.
If it were me, I'd rush laws like this in, with as little roof for public debate as possible. I'd also go out of my way to make the most insanely draconian DRM measures manditory in all consumer electronics and force consumers to return/submit any and all non-drm'ed devices... why?
Because until copyright law (and especially, copyright protecting DRM) becomes a massive pain in the ass to consumers, they're not going to care enough to stop it...
Exactly! If only blizzard would make this kind of thing a part of the game, or at least legal, I'd probably start playing WoW again. It's just silly that, as fun as WoW is, it's got such boring repetitive grinding in it. It's really quite annoying.
he really only talks about the gameplay code... code that throws a lot of exceptions... ai code is generally a lot smoother. I mean, stuff like neural nets and genetic algorithms would be absolutely fine on these things... but anything coded like an expert system would potentially suffer from performance issues. He says that that kind of code will run 1/3 to 1/10th it's "normal" speed on an out-of-order processor. But this code is normally vying for processor time in between context switches... and who really cares about ai anyway right? as long as the game looks pretty... well... now the same code can run 1/3 slower, but 100% of the time on a cell.... I seriously think that the ai in games could get a helluva lot better as a result of cells.
funny... I write my ai code in C... game developers usually use whatever they're coding their games in... I still don't see how poor handling of branch prediction* (or no handling of it at all) is going to have any effect on whether or not a coder can code in an ai using a language that completely and utterly abstracts that low level crap away? There's a difference between not being able to easily code something, and not having your code run in quite the optimized way that it used to. The former simply isn't true from anything I've seen (hell... even ASSEMBLER abstracts that extremely low level shit away...) and the latter is a matter that's up for debate; having your code run a bit slower than normal is somewhat bad, but having it be able to run on a dedicated processor and hence much faster than normal more than makes up for it.
So I'm still calling bullshit on this one so far, but only because no one has given any details... if you know them, please show 'em...
* - it should be noted that branch prediction is just an OPTIMIZATION technique at the hardware level... code in whatever language you want, it'll still run fine, the processor just won't have already prefetched the instruction and associated memory blocks... if this is all the FUD boils down to, then it's massively busted since my (and anyone elses) ai code will still run perfectly fine on it....
I'm replying, so I obviously didn't mod it as such... but I'd have to say because it's an assinine comment. I'm just getting immersed in my M.Sc. in CS and so far it's like going from night to day.... I would strongly argue that in the CS realm you don't start really getting into CS until you start working on your thesis. I got strong marks in undergrad, took a boatload of theoretical courses as well as a large number of coding projects, and none of them come close to what I'm working on now in terms of sheer theoretic complexity... Saying that theoretical CS work beyond the B.Sc. level is superfluous is just uninformed. I mean, cripes, this wasn't done by undergrads, and I'd say it's some cool shit... so yeah... the GP looks like a cleverly disguised regurgitation of a number of other trollish posts about the subject.
Even with the inefficiencies of having an in-order processor doing ai calculations, I still think that the fact that you can have a processor completely dedicated is what will make the difference. I'm not entirely sure what game makers typically use for their opponent ai, but on the cutting edge of ai research in general, all of the algorithms I've looked at can benefit from having multiple processors (especially any type of algorithm which uses a population of candidates who all have to try out the task before they determine which is best, like Genetic Algorithms/Programming). There's also a lot of brute force calculations (like positional location, line of sight determination for ai algorithm inputs, etc...) which are usually tackled in a "quick and dirty" manner because of horsepower constraints... these constraints would all but disappear with some extra cells onto which you can offload these calculations... I think I've got my prof convinced to buy a couple after the Christmas rush and, like I said before, I'm pretty psyched to get them (and not just because it'll give me something to do in the lab when I'm feeling brainfscked...)
CMU Dragons? Dude, can I talk to you or one of your guys about design? I've seen the videos of you guys in Germany this year and they're insane. There's no Canadian presence in the small league atm (to my knowledge) and I'm trying to get funding to get a team in there... I'm hoping to present a PIPE variant in Hotlanta in July, but other than some simulation-league stuff that I'm hoping to put together for souzhou the year after, my university doesn't have anything:-(
your comment was interesting... pray tell, how is it that a system custom-built for parallel processing is *bad* for ai? I'm an ai researcher and I'm dying to get my hands on one of these things cuz they'd be perfect for MAS/MRS research... hell, you could have 6 of those processors running the ai code for the 6 bots you're playing against... so please, tell me, how is it bad?
I don't see anyone divided, what I see is lots of people that fear for Novell : they're basically toast once GPLv3 is out.
And what a lovely piece of FUD that will make when Novell (who by that time likely will have MS IP in their OS) has to either rollback their OS a couple years (which will likely kill it) or simply folds up their OS shop... I can see the MS sponsored FUD now: "Linux and FOSS in general is such a broken way of doing things that even access to MS money can't make it profitable and successful... just look at Novell"
Is there some way anyone could see what patents they've allegedly got that are being infringed?
erm... that's complicated:
a - probably not, since there probably aren't any,
b - if, in fact, there are some, I don't think they've gotten around to figuring out which ones they are yet.
Another poster hit the nail on the head, this is a masterstroke of FUD which only cost microsoft $500mil. The masterstroke is that it is having better FUD-like effects than any previous MS-FUD that I can remember, but it's targeted at a movement which encompasses the company they paid the $500mil to, so it's that much harder to recognize it as FUD.... since they're trying, basically, to throw the battle in order to win the war... Kinda like a delaying battle designed to suck your opponent into a KZ... sure you're getting attritted a little bit, but you've caused them to break ranks and the bulk of them are coming into range...
(and technically, without an export license, closed source software must be reviewed first..., even today)
Yes, but thankfully if you get caught (assuming what you say is true) you won't get crucified for being an arms dealer... which is what they wanted to do to him, for allowing it to get out. And actually, it wasn't illegal for him to make it available on usenet... it became illegal the second a non-USian downloaded a copy.
I heard somewhere that one of the biggest reasons why this attempt to screw Phil over failed is because of MIT. Apparently they immediately began sharing copies of it and distributing it as widely as they could, sort of to say: "if you're going to call him an arms dealer, you'll have to call us arms dealers as well..." it was, apparently, at that point that someone in the administration of the time realized that the statute didn't quite make sense... still a LONG and STUPID battle for which Phil deserves massive kudos.
sure it does: Set up the experiment, look at numbers 31, 57 and 68 for the message, and at a later date, when you want to send the message back to yourself, you encode it in 31,57 and 68.... course doing so will change your actions, so you'll have to remember to send them to yourself in your new future.... course since you're now sending your message in a different "future" than originally, if you make a typo you could be fucked... aren't paradoxes fun?:)
Yes, but he PUBLISHED it on teh INTARWEB (well... usenet...but anyway) which the gubment then claimed was EXPORTATION (Read the first two paras in the History section)... not only is the GP 100% correct, but so are you... go figure, huh
The post, and the reaction to it remind me of a story my ex-wife (a helicopter mechanic in the military) once told me about a trouble-ticket on the aircraft:
Pilot's Problem: Collective requires more than the rated 5lbs of pressure to lift Tech Solution: Pilot needs to put the purse down...
uh... "suck it up" is mostly a military term refering to sucking in the gut, puffing out the chest, and soldiering on. Princess refers to the "hoighty-toighty" princess-type that can't get comfortable on 15 mattresses because there's a pea in between a pair of them. So not gay baiting at all... but fuck you very much anyway... (not posted anon cuz I've at least got the parts to stand behind the mud I sling...)
Slashdot Mirror
$15/gal gas is a bit absurd, but, in a way, he's got the right idea. The inability to use large trucks and other gas guzzlers for the distribution of food would cause industry to choose an alternative which is more economical. It's not just the population at large that would have to adapt, it's industry as a whole. You'd probably see a lot more mass transit spring up in areas, because suddenly it would be a requirement for a LARGE percentage of the population. There would, of course, be food distributors who would choose a cheaper method of transportation, like rail, that would be able to offer food cheaper than their competition (who would simply raise prices to pay for gas) and people would buy their food instead of the expensive shit. Likewise, the increased dependence on something like rail (which can simply tap into the energy grid for power) would probably result in the public wanting high speed rail for the transport of goods... in short, if gas prices went to $15/gal in a slow, but, steady rate, these changes would happen... and people would only starve if the government refused to provide incentives to hasten the changes that would be required as a result of expensive gas.
It's all a question of political will. We live in a society where we can build gigantic structures that touch the sky and tunnel through the earth. Changing our infrastructure from road to something else is barely a challenge worth of our combined intellects.... but only if we actually give a shit enough to try in a coordinated and not-half-assed manner...
Maybe I'm just not jaded enough, but I didn't interpret his statement to be directed at the 17 year old... his statement seemed to be directed at the ever popular "they". As in, this is old news, and if a kid can do it (albeit a very smart kid) then why the fuck aren't "they" doing it in larger scales, and why are "they" still burning coal and oil ffs?
... unless I missed the whoosh...
I know that graphics-guru's used to have to write a lot of inline assembly and other optimizers in order to get certain parts of their code to run a little bit faster. So, the majority of people won't need to delve into that kind of depth... but there are some who push the envelope a bit and who, traditionally, needed to get their hands dirty on the low level stuff. No idea if this is still the case... maybe someone a little more involved in pushing the envelope could answer... Mr. Carmack, are you out there?
yes, and if you run an I-don't-have-to-outrun-the-bear-just-you experiment on a species for a couple generations, they will eventually evolve some mechanism for going faster and faster. Which is what "happened" here... except, partway into the beginning of the experiment, they realized that they didn't have to do any real outrunning, they just needed to climb a tree. So, naturally, the ones that were better at climbing trees survived and had more chilluns... thems chilluns had the same features as their better-surviving parents (shorter legs more suited for climbing) and so they survived "better" than their neighbours whose parents didn't have such short legs (but managed to reproduce anyway...) Then these lizards went on to have chilluns, but their neighbours, who weren't quite as good at climbing, were yesterdays lunch for the predator...
mutations give the genetic diversity (i.e. differing leg lenghts) and natural selection ensures that a certain group survives... that group has kids, who have kids, etc... and eventually, the differing genetic "chunks" that helped the species thrive get propagated from parent to child, and eventually becomes dominant.
I know this thread is a couple days old... but I feel bad for the guy posting. No one actually answered his question.... take the parent, for example. He gives a lot of good reasons why the OP should take a B.Sc. in either comp sci or math... totally missing the fact that the OP already has one and is talking about getting an M.Sc. in Game Programming. (no offense adisakp... 95% of the posters in this "story" missed that...)
It's probably too late to answer effectively (ie. no one will read this) but I'd say that the parent is right in his reasoning, but needs to rewrite his post to suggest a masters in CS and not a bac...
My suggestion would be to do the M.Sc. in comp sci, but take courses that mimic/are those given in the gaming M.Sc. and gear your thesis around something that the game companies would find useful, and that the non-gaming companies would not dismiss out of hand. Perhaps an optimized algorithm for a transport mechanism that's faster than TCP but more reliable than UDP... or an optimized algorithm for some graphics raytracing/rasterizing/whatever techniques.
In that way, you'd still be taking what you are interested in, but you wouldn't risk having some dolt think that you're not the real deal just because your degree is in game programming. With the comp sci M.Sc. geared towards game programming, you'll be able to show the interviewers in the gaming companies that you are "t3h sh1t" and when you eventually go somewhere else, they'll just see the M.Sc. Comp. Sci. on your VC and they won't immediately assume that you're just sh1t... good luck!
We are at a total loss to understand how this policy has developed, who is behind it and why there is such haste in enacting it into law -- with little if any public debate.
If it were me, I'd rush laws like this in, with as little roof for public debate as possible. I'd also go out of my way to make the most insanely draconian DRM measures manditory in all consumer electronics and force consumers to return/submit any and all non-drm'ed devices... why?
Because until copyright law (and especially, copyright protecting DRM) becomes a massive pain in the ass to consumers, they're not going to care enough to stop it...
Exactly! If only blizzard would make this kind of thing a part of the game, or at least legal, I'd probably start playing WoW again. It's just silly that, as fun as WoW is, it's got such boring repetitive grinding in it. It's really quite annoying.
he really only talks about the gameplay code... code that throws a lot of exceptions... ai code is generally a lot smoother. I mean, stuff like neural nets and genetic algorithms would be absolutely fine on these things... but anything coded like an expert system would potentially suffer from performance issues. He says that that kind of code will run 1/3 to 1/10th it's "normal" speed on an out-of-order processor. But this code is normally vying for processor time in between context switches... and who really cares about ai anyway right? as long as the game looks pretty... well... now the same code can run 1/3 slower, but 100% of the time on a cell.... I seriously think that the ai in games could get a helluva lot better as a result of cells.
funny... I write my ai code in C... game developers usually use whatever they're coding their games in... I still don't see how poor handling of branch prediction* (or no handling of it at all) is going to have any effect on whether or not a coder can code in an ai using a language that completely and utterly abstracts that low level crap away? There's a difference between not being able to easily code something, and not having your code run in quite the optimized way that it used to. The former simply isn't true from anything I've seen (hell... even ASSEMBLER abstracts that extremely low level shit away...) and the latter is a matter that's up for debate; having your code run a bit slower than normal is somewhat bad, but having it be able to run on a dedicated processor and hence much faster than normal more than makes up for it.
So I'm still calling bullshit on this one so far, but only because no one has given any details... if you know them, please show 'em...
* - it should be noted that branch prediction is just an OPTIMIZATION technique at the hardware level... code in whatever language you want, it'll still run fine, the processor just won't have already prefetched the instruction and associated memory blocks... if this is all the FUD boils down to, then it's massively busted since my (and anyone elses) ai code will still run perfectly fine on it....
I'm replying, so I obviously didn't mod it as such... but I'd have to say because it's an assinine comment. I'm just getting immersed in my M.Sc. in CS and so far it's like going from night to day.... I would strongly argue that in the CS realm you don't start really getting into CS until you start working on your thesis. I got strong marks in undergrad, took a boatload of theoretical courses as well as a large number of coding projects, and none of them come close to what I'm working on now in terms of sheer theoretic complexity... Saying that theoretical CS work beyond the B.Sc. level is superfluous is just uninformed. I mean, cripes, this wasn't done by undergrads, and I'd say it's some cool shit... so yeah... the GP looks like a cleverly disguised regurgitation of a number of other trollish posts about the subject.
Even with the inefficiencies of having an in-order processor doing ai calculations, I still think that the fact that you can have a processor completely dedicated is what will make the difference. I'm not entirely sure what game makers typically use for their opponent ai, but on the cutting edge of ai research in general, all of the algorithms I've looked at can benefit from having multiple processors (especially any type of algorithm which uses a population of candidates who all have to try out the task before they determine which is best, like Genetic Algorithms/Programming). There's also a lot of brute force calculations (like positional location, line of sight determination for ai algorithm inputs, etc...) which are usually tackled in a "quick and dirty" manner because of horsepower constraints... these constraints would all but disappear with some extra cells onto which you can offload these calculations... I think I've got my prof convinced to buy a couple after the Christmas rush and, like I said before, I'm pretty psyched to get them (and not just because it'll give me something to do in the lab when I'm feeling brainfscked...)
CMU Dragons? Dude, can I talk to you or one of your guys about design? I've seen the videos of you guys in Germany this year and they're insane. There's no Canadian presence in the small league atm (to my knowledge) and I'm trying to get funding to get a team in there... I'm hoping to present a PIPE variant in Hotlanta in July, but other than some simulation-league stuff that I'm hoping to put together for souzhou the year after, my university doesn't have anything :-(
your comment was interesting... pray tell, how is it that a system custom-built for parallel processing is *bad* for ai? I'm an ai researcher and I'm dying to get my hands on one of these things cuz they'd be perfect for MAS/MRS research... hell, you could have 6 of those processors running the ai code for the 6 bots you're playing against... so please, tell me, how is it bad?
oh, I'm sorry, but thanks for coming out.
.... for the love of god, tell me where such a holy-land can be found??
I don't see anyone divided, what I see is lots of people that fear for Novell : they're basically toast once GPLv3 is out.
And what a lovely piece of FUD that will make when Novell (who by that time likely will have MS IP in their OS) has to either rollback their OS a couple years (which will likely kill it) or simply folds up their OS shop... I can see the MS sponsored FUD now: "Linux and FOSS in general is such a broken way of doing things that even access to MS money can't make it profitable and successful... just look at Novell"
Is there some way anyone could see what patents they've allegedly got that are being infringed?
erm... that's complicated:
a - probably not, since there probably aren't any,
b - if, in fact, there are some, I don't think they've gotten around to figuring out which ones they are yet.
Another poster hit the nail on the head, this is a masterstroke of FUD which only cost microsoft $500mil. The masterstroke is that it is having better FUD-like effects than any previous MS-FUD that I can remember, but it's targeted at a movement which encompasses the company they paid the $500mil to, so it's that much harder to recognize it as FUD.... since they're trying, basically, to throw the battle in order to win the war... Kinda like a delaying battle designed to suck your opponent into a KZ... sure you're getting attritted a little bit, but you've caused them to break ranks and the bulk of them are coming into range...
and owns ~80 of the market.
A WHOLE 80 SERVERS!? ... shit man... that's like 95% of the market ;-)
/ducks
Oh SNAP!
there ya go ;-)
(and technically, without an export license, closed source software must be reviewed first..., even today)
Yes, but thankfully if you get caught (assuming what you say is true) you won't get crucified for being an arms dealer... which is what they wanted to do to him, for allowing it to get out. And actually, it wasn't illegal for him to make it available on usenet... it became illegal the second a non-USian downloaded a copy.
I heard somewhere that one of the biggest reasons why this attempt to screw Phil over failed is because of MIT. Apparently they immediately began sharing copies of it and distributing it as widely as they could, sort of to say: "if you're going to call him an arms dealer, you'll have to call us arms dealers as well..." it was, apparently, at that point that someone in the administration of the time realized that the statute didn't quite make sense... still a LONG and STUPID battle for which Phil deserves massive kudos.
sure it does: Set up the experiment, look at numbers 31, 57 and 68 for the message, and at a later date, when you want to send the message back to yourself, you encode it in 31,57 and 68.... course doing so will change your actions, so you'll have to remember to send them to yourself in your new future.... course since you're now sending your message in a different "future" than originally, if you make a typo you could be fucked... aren't paradoxes fun? :)
Yes, but he PUBLISHED it on teh INTARWEB (well... usenet...but anyway) which the gubment then claimed was EXPORTATION (Read the first two paras in the History section)... not only is the GP 100% correct, but so are you... go figure, huh
/me hands Tom a KitKat
chillax...
Hallo Minkey?
The post, and the reaction to it remind me of a story my ex-wife (a helicopter mechanic in the military) once told me about a trouble-ticket on the aircraft:
Pilot's Problem: Collective requires more than the rated 5lbs of pressure to lift
Tech Solution: Pilot needs to put the purse down...
uh... "suck it up" is mostly a military term refering to sucking in the gut, puffing out the chest, and soldiering on. Princess refers to the "hoighty-toighty" princess-type that can't get comfortable on 15 mattresses because there's a pea in between a pair of them. So not gay baiting at all... but fuck you very much anyway... (not posted anon cuz I've at least got the parts to stand behind the mud I sling...)
okay...
Q: Why don't sharks attack lawyers?
A: Professional courtesy.
... ohhhhh... no more fucking shark jokes... meh... suck it up princess....