Well in theory if you can solve any NP complete problem you can solve them all - you just need to figure out how you can map the new problem onto the solvable one.
Of course there are practical difficulties using these types of analog computer as you scale up, but I don't know how much serious effort has been taken to look for analog approaches that would scale.
NP complete problems are only hard if you insist on using a digital computer to solve them.
Want to find the shortest set of roads to build to interconnect a bunch of cities? You need a couple of pieces of glass, some round plastic rod and some soapy water...
Cut the rod into 1" lengths, so that you have one per city, then glue them as separators between the two pieces of glass so that the positions of the rods represent the locations of the cities. Dip the completed "computer" into soapy water, and let surace tension and energy minimization do it's job - the soap bubbles that form between the rods will have edges that join where you should build your roads.
Well, my subject about says it all. Just because most scientists can and do use cars and pencils, we don't refer to them as racing drivers or pencil-operators.
"All scientists are programmers" would have been a truer headline, as would "All programmers are not computer scientists".
"If I were 21 years old," he said at a company conference in New Orleans, "I probably wouldn't go into computing. The computing industry is about to become boring. I'd go into genetic engineering."
This rings true to me. Much of Comp Sci (my chosen profession, though I suck at it) seems to have a lack of discovery and / or innovation these days, with the exception of nanocomputing. Much of the rest of it is innovation, not invention / discovery. How many Turings do we have in Comp Sci now?
Computing / computer science is a skill rather than an industry. While I'm a programmer, and have worked for a computer company (Acorn), I've also worked for a medical company and a couple of communications companies. I'm sure I could get a programming job at a medical company doing genetic research if I set my mind to it.
Secondly, how many people working in genetics are making fundamental discoveries, and how many are just grunts doing their job? For that matter, how many people's jobs in *any* field allow them to do blue skies research of the type that may lead to fundamental discovery?
I've long ago realized I had to separate my intellectual interests from my job. While I've been lucky to have extremely interesting work assignments, it's at home that I become the "mad scientist".:-)
If you want Jabber or some other open source IM protocol to be able to interoperate with AOL clients, that requires that you use AOL servers to talk to them.
Assuming AOL is willing to consider this as a business case, how much are YOU as a user willing to pay for this use of AOL's servers?
OSCAR is AOL's low level protocol, which they do not support for 3rd part clients. TOC is a different high level protocol that sits on top of OSCAR and provides a slightly limited subset of the functionality. AOL *DO* support use of TOC - they even (used?) to provide source of a Java-based TOC client, as well as the Tk/Tcl Tik-TOC client.
There's no reason that Brook's subsumption architecture has to work at the level of raw sensory input - it can just as well work on internal models. If fact, this is along the lines of what Minsky is proposing in his "Society of Mind" - that we're based on a set on low level interacting intelligences.
Sure there are still plenty of Intel boxes at retail, but they're all low end = 1 GHz. AMD has all the high end. What a turnaround from a couple of years ago!!!!
You've got that backwards - AMD owns the desktop retail sector. Intel owns the server market and laptops.
AMD's 760MP SMP chipset coupled with the incredible 1.5GHz AMD Palomino just shown at CeBIT are going to be used in servers from Compaq and apparently IBM also.
AMD's Morgan (portable Palomino) comes out later this year.
With Rambus rapidly going out of business and DDR P4 chipsets proliferating, Intel is in a world of hurt. Already their dismal selling (~1% of total PC sales) P4 is bested by the slower clocked AMD Athlon, and with P4 hobbled with DDR (it was designed for the huge memory bandwidth of RDRAM), the comparisons of P4 vs Athlon and Palomino just get more and more lopsided. Intel has dropped the ball.
Re:It goes against reason, check your bible !
on
New Human Ancestor?
·
· Score: 1
If we really are descended from monkeys, how come we don't all enjoy swinging from trees, eating bananas and mindless copulation with the closest memeber of either sex?
I'm confused by the term "memeber". Perhaps you think that copulation and banana-eating are memes rather than activities people actually engage in?
I don't see the OpenDivX effort as really being good for open source causes, since it's not GPL'd.. you couldn't legally take their enhancements and incorporate them into a GPL'd project (of course it'll happen anyway, but...).
Second, as a Linux user all I want is to be able to play whatever digital media is available, and to be able to compress to those common standards also. Even if there's a great open source alternative to MPEG-4, I still want to be able to play MPEG-4 if (as it seems) it's going to be widely adopted. Given the MPEG-1/2 situation, I don't see any reason to believe that MPEG-4 licencing fees will be pursued from end users, and if I ever got to the point of wanting to commercially author MPEG-4 movies I wouldn't begrudge them the licence fees.
'R' rated movies or 'explicit lyrics' CDs never stopped anyone from seeing them, anymore than an 'extreme violence' sticker on a game would.
In fact we all know that these things can gave the opposite effect than that intended. What game ratings would do, which seems like a *good* thing, would be to allow parents to have some control on what they give their kids as presents, or even to have a clue what the stuff lying around in their room is.
http://www.divx-digest.com/software/divxcodec.ht ml
One source for the original "3.11 alpha" DivX;-) CODEC, that is just a binary hack of Microsoft's (almost) MPEG-4 CODEC to allow it to be used with AVIs. This is the CODEC that practically all the DivX's on the net are compressed with.
http://www.projectmayo.com/index.php
Home of OpenDivX and run by divxnetworks.com. OpenDivX is based on the MoMuSys MPEG-4 source code, and unlike DivX claims to be MPEG-4 compliant, albeit using AVI rather than MPEG-4 file format. This encoder is S-L-O-W and also incompatible with DivX;-), despite the "OpenDivX" name.
You can play, and encode(!) OpenDivX movies with the awesome mplayer media player for linux:
http://thot.banki.hu/esp-team/MPlayer.html
http://www.3ivx.com/
A commercial company producing a free MPEG-4 decoder, and about to announce an encoder at CeBIT. Quality and decoder speed are good. An xanim plug in is available, as is Windows etc support. Uses Quicktime as a file format
http://rachmaninoff.ti.uni-mannheim.de/sampeg/
A work-in-progress GPL'd MPEG-4 encoder from Dirk Farin, the guy who wrote the impressive SAMPEG-2 MPEG-2 encoder. Actually uses MPEG-4 file format. Sounds very promising.
http://sparky.sourceforge.net/
Not quite MPEG-4, but GPL'd and competetive in terms of compression. From the guy who wrote the avifile win32 CODECs on Linux library. Currently slow, but an impressice start for a real open source CODEC.
http://www.opencodex.com/news.html
The guys who ran the original $50K DivX for Quicktime port competition. Web site keeps claiming good things, but nothing is getting released... It turns out the CODEC is H.263 based abyway, not true MPEG-4 (although MPEG-4 and H.263 are quite closely related - they use the same quantizer).
The story that/. rejected was centered around the stunning (to me, at least!;-) press release from Sorenson at MacWorld about a month ago that they now have a VERY impressive MPEG-4 CODEC (incl. 2-pass VBR)...I'm amazed and happy that Sorenson would so aggressively push an open standard when their cash cow is a proprietary competitor.
It sucks that/. thought their lame story quoting the divxnetworks.com weasels was better than this stuff!;-)
I don't know. I assume they're trying to build on the DivX "brand name" and usurp adoption of standard MPEG-4... As you say, with people like Phillips and Sorenson already technically way ahead, and OpenDivX incompatible with DivX anyway, it seems a long shot!
Not exactly all theirs. It's the MoMuSys MPEG-4 implementation with a slightly hacked (S-L-O-W) encoder and a fast decoder. I don't know whey they chose AVI rather than Quicktime or MPEG-4 file format - I guess they don't care about A/V sync. Oh, well.
ProjectMayo is a scam. It's not a real open source project, but rather an open source freeloader poroject run by the commercial enterprise divxnetworks.com, with $100M of backing.
http://www.divxnetworks.com/aboutus.html
The "open source" licence is not GPL - it requires you to do stuff like adding some kind of "made with divxnetwork" header to your movies.
OpenDivX is based on the MoMuSys source, claims to be MPEG-4 compliant (aside from using an AVI vs MPEG-4 transport), and is incompatible with DivX.
I just got a/. article rejected today where I explained all this plus gave links to all the free MPEG-4 implementations and the Sorenson MPEG-4 press release.
Slashdot Mirror
Well in theory if you can solve any NP complete problem you can solve them all - you just need to figure out how you can map the new problem onto the solvable one.
Of course there are practical difficulties using these types of analog computer as you scale up, but I don't know how much serious effort has been taken to look for analog approaches that would scale.
Well, yeah, but it's not free. You've still got to figure out how to map your specific problem onto some other one that you know how to solve.
NP complete problems are only hard if you insist on using a digital computer to solve them.
Want to find the shortest set of roads to build to interconnect a bunch of cities? You need a couple of pieces of glass, some round plastic rod and some soapy water...
Cut the rod into 1" lengths, so that you have one per city, then glue them as separators between the two pieces of glass so that the positions of the rods represent the locations of the cities. Dip the completed "computer" into soapy water, and let surace tension and energy minimization do it's job - the soap bubbles that form between the rods will have edges that join where you should build your roads.
No - strong AI is a philosophical problem (for some). There's no reason to believe it's NP complete.
Well, my subject about says it all. Just because most scientists can and do use cars and pencils, we don't refer to them as racing drivers or pencil-operators.
"All scientists are programmers" would have been a truer headline, as would "All programmers are not computer scientists".
"If I were 21 years old," he said at a company conference in New Orleans, "I probably wouldn't go into computing. The computing industry is about to become boring. I'd go into genetic engineering."
:-)
This rings true to me. Much of Comp Sci (my chosen profession, though I suck at it) seems to have a lack of discovery and / or innovation these days, with the exception of nanocomputing. Much of the rest of it is innovation, not invention / discovery. How many Turings do we have in Comp Sci now?
Computing / computer science is a skill rather than an industry. While I'm a programmer, and have worked for a computer company (Acorn), I've also worked for a medical company and a couple of communications companies. I'm sure I could get a programming job at a medical company doing genetic research if I set my mind to it.
Secondly, how many people working in genetics are making fundamental discoveries, and how many are just grunts doing their job? For that matter, how many people's jobs in *any* field allow them to do blue skies research of the type that may lead to fundamental discovery?
I've long ago realized I had to separate my intellectual interests from my job. While I've been lucky to have extremely interesting work assignments, it's at home that I become the "mad scientist".
If you want Jabber or some other open source IM protocol to be able to interoperate with AOL clients, that requires that you use AOL servers to talk to them.
Assuming AOL is willing to consider this as a business case, how much are YOU as a user willing to pay for this use of AOL's servers?
If your answer is zero, then shut the fuck up.
OSCAR is AOL's low level protocol, which they do not support for 3rd part clients. TOC is a different high level protocol that sits on top of OSCAR and provides a slightly limited subset of the functionality. AOL *DO* support use of TOC - they even (used?) to provide source of a Java-based TOC client, as well as the Tk/Tcl Tik-TOC client.
There's no reason that Brook's subsumption architecture has to work at the level of raw sensory input - it can just as well work on internal models. If fact, this is along the lines of what Minsky is proposing in his "Society of Mind" - that we're based on a set on low level interacting intelligences.
Sure there are still plenty of Intel boxes at retail, but they're all low end = 1 GHz. AMD has all the high end. What a turnaround from a couple of years ago!!!!
I wonder if that'll change as soon as Compaq AMD-based servers start eating into their high end profits later this year...
You've got that backwards - AMD owns the desktop retail sector. Intel owns the server market and laptops.
AMD's 760MP SMP chipset coupled with the incredible 1.5GHz AMD Palomino just shown at CeBIT are going to be used in servers from Compaq and apparently IBM also.
AMD's Morgan (portable Palomino) comes out later this year.
With Rambus rapidly going out of business and DDR P4 chipsets proliferating, Intel is in a world of hurt. Already their dismal selling (~1% of total PC sales) P4 is bested by the slower clocked AMD Athlon, and with P4 hobbled with DDR (it was designed for the huge memory bandwidth of RDRAM), the comparisons of P4 vs Athlon and Palomino just get more and more lopsided. Intel has dropped the ball.
If we really are descended from monkeys, how come we don't all enjoy swinging from trees, eating bananas and mindless copulation with the closest memeber of either sex?
I'm confused by the term "memeber". Perhaps you think that copulation and banana-eating are memes rather than activities people actually engage in?
Kind of a later reply here, but...
I don't see the OpenDivX effort as really being good for open source causes, since it's not GPL'd.. you couldn't legally take their enhancements and incorporate them into a GPL'd project (of course it'll happen anyway, but...).
Second, as a Linux user all I want is to be able to play whatever digital media is available, and to be able to compress to those common standards also. Even if there's a great open source alternative to MPEG-4, I still want to be able to play MPEG-4 if (as it seems) it's going to be widely adopted. Given the MPEG-1/2 situation, I don't see any reason to believe that MPEG-4 licencing fees will be pursued from end users, and if I ever got to the point of wanting to commercially author MPEG-4 movies I wouldn't begrudge them the licence fees.
I've just modified my preferences to ignore your articles.
:-)
Congratulations - you're the only thing on my ignore list.
Your articles are vapid trolls, and you manage to say less in 1000 words than most people can say in 100.
Oh, but don't feel bad - I'm willing to lose Karma just to say this:
Bye bye!
Huh? Think for a second before you post!
Can you watch Sorenson Video 3 encoded movies on Linux? -- No
Would you be able to watch Sorenson MPEG-4 encoded movies on Linux? -- Yes
Why? -- Because MPEG-4 is an open standard.
Get it???????
'R' rated movies or 'explicit lyrics' CDs never stopped anyone from seeing them, anymore than an 'extreme violence' sticker on a game would.
In fact we all know that these things can gave the opposite effect than that intended. What game ratings would do, which seems like a *good* thing, would be to allow parents to have some control on what they give their kids as presents, or even to have a clue what the stuff lying around in their room is.
OK, but I'm lazy, so no HTML links!
t ml
;-) CODEC, that is just a binary hack of Microsoft's (almost) MPEG-4 CODEC to allow it to be used with AVIs. This is the CODEC that practically all the DivX's on the net are compressed with.
;-), despite the "OpenDivX" name.
/. rejected was centered around the stunning (to me, at least! ;-) press release from Sorenson at MacWorld about a month ago that they now have a VERY impressive MPEG-4 CODEC (incl. 2-pass VBR)...I'm amazed and happy that Sorenson would so aggressively push an open standard when their cash cow is a proprietary competitor.
/. thought their lame story quoting the divxnetworks.com weasels was better than this stuff! ;-)
http://www.divx-digest.com/software/divxcodec.h
One source for the original "3.11 alpha" DivX
http://www.projectmayo.com/index.php
Home of OpenDivX and run by divxnetworks.com. OpenDivX is based on the MoMuSys MPEG-4 source code, and unlike DivX claims to be MPEG-4 compliant, albeit using AVI rather than MPEG-4 file format. This encoder is S-L-O-W and also incompatible with DivX
You can play, and encode(!) OpenDivX movies with the awesome mplayer media player for linux:
http://thot.banki.hu/esp-team/MPlayer.html
http://www.3ivx.com/
A commercial company producing a free MPEG-4 decoder, and about to announce an encoder at CeBIT. Quality and decoder speed are good. An xanim plug in is available, as is Windows etc support. Uses Quicktime as a file format
http://rachmaninoff.ti.uni-mannheim.de/sampeg/
A work-in-progress GPL'd MPEG-4 encoder from Dirk Farin, the guy who wrote the impressive SAMPEG-2 MPEG-2 encoder. Actually uses MPEG-4 file format. Sounds very promising.
http://sparky.sourceforge.net/
Not quite MPEG-4, but GPL'd and competetive in terms of compression. From the guy who wrote the avifile win32 CODECs on Linux library. Currently slow, but an impressice start for a real open source CODEC.
http://www.opencodex.com/news.html
The guys who ran the original $50K DivX for Quicktime port competition. Web site keeps claiming good things, but nothing is getting released... It turns out the CODEC is H.263 based abyway, not true MPEG-4 (although MPEG-4 and H.263 are quite closely related - they use the same quantizer).
The story that
It sucks that
I don't know. I assume they're trying to build on the DivX "brand name" and usurp adoption of standard MPEG-4... As you say, with people like Phillips and Sorenson already technically way ahead, and OpenDivX incompatible with DivX anyway, it seems a long shot!
Not exactly all theirs. It's the MoMuSys MPEG-4 implementation with a slightly hacked (S-L-O-W) encoder and a fast decoder. I don't know whey they chose AVI rather than Quicktime or MPEG-4 file format - I guess they don't care about A/V sync. Oh, well.
ProjectMayo is a scam. It's not a real open source project, but rather an open source freeloader poroject run by the commercial enterprise divxnetworks.com, with $100M of backing.
http://www.divxnetworks.com/aboutus.html
The "open source" licence is not GPL - it requires you to do stuff like adding some kind of "made with divxnetwork" header to your movies.
divxnetworks.com is the company behind "Project Mayo" and "OpenDivX".
/. article rejected today where I explained all this plus gave links to all the free MPEG-4 implementations and the Sorenson MPEG-4 press release.
Check out the flames forum at:
Project Mayo (aka divxnetworks.com).
OpenDivX is [u]incompatible[/u] with DivX.
DivX is the hacked Microsoft CODEC.
OpenDivX is based on the MoMuSys source, claims to be MPEG-4 compliant (aside from using an AVI vs MPEG-4 transport), and is incompatible with DivX.
I just got a
*sigh*
Best of luck with your upgrade. I hope it doesn't fail horribly!
;-)
Yeah, that would suck!
I'm not sure if you got modded up for the ICQ protocol link, or for ESR's sex tips! ;-)
Exactly so!
...) of theory, that arn't necessarily related.
Math is a bunch of islands (number theory, topology,
In fact, usually when someone such as Andrew Weil DOES build a bridge between a couple of these islands then we herald that as a rare triumph!
Also, as others have stated, it's a bit late to be worried about Godel's incompleteness theorem... kind of old news!