KnowledgeKinetics (www.knowledgekinetics.net) handles document (content) management as well as a whole slew of other things. Quite a nifty set of tools.
Binary data in XML is stored uu-encoded... ( I've been told that's the standard way of doing things with that... me, I prefer putting text data in --easier to debug, so what do I know?)
The funniest thing is that no matter how many times someone complains about the atrocious grammar or spelling... nothing is done. Remember the "put Hemos through English class" sigs? Ah well, it makes for entertaining commentaries.
They are so Windows-centric that they forget about the importance of portability. Like MFC, and half the other shit they make which ends up stangling a developer to a particular platform. There's a reason Java will remain popular, and that's Sun's devotion to portability and stability. ?
Does it bother you that variables within C# are all COM objects? Does your operating system support COM? Do you have a program that allows COM objects to exist on your machine? Then it can't run C#. There goes portability...
Stability. Hmmm. Complete control over the spec would imply stability. Not open-ness, but at least stability. For all the cruft in Microsoft's closed API's, there's very little change --> nice and stable, a bit like Java (I know, heresy:)
I'd like to see an anarchic DNS system. You choose who you want to be your DNS server and go with whatever garbage they serve up. And the sam goes for name propogation. And so on, for the DNS servers themselves. Everyone can register any name they want on their own DNS server. Mmmm... people will have to start memorizing more numbers...:)
So it's crazy. But *I* like the idea. That way there can be more seamus.org's than just the one (why, oh why was I so slow?:) Of course, eventually people would get pissy about having "convenient" IP addresses to type in... can you imagine network card and router sales/manufacturing in this future? Neat...
For the Web developer, the tools to build, test and deploy engaging Web sites are hopelessly inadequate. Many focus more on building attractive rather than useful Web sites.
Doesn't Microsoft's current website work this way... (i.e. hopelessly inadequate). Could they have a worse method of accessing their "Knowledge Base" aka "how to work around our bugs"?).
What's the minimum for cos(x)? One variable, massive search space... (well, cheating really, but you get the point...:) Finding minimums, is... hard... And numerical optimizers are REALLY bad at doing non-linear problems, having worked with CPLEX and a few others of its ilk. CPLEX, in specific, is great for linear problem solving, I rather liked using it... but it did squat when it came time to work on the non-linear problems (nuclear waste stream disposal -- very cool stuff, check out M.A.W.S.)
I wish I could find the link to an article I read about a year ago... basically saying that the people who originally used the SUV type vehicles for real work (ranchers, etc.) can now no longer afford to buy them... psychotic, eh?
While the majority of Open Source (or Free) projects are for Linux, there are plenty out there for Windows (maybe there should be more). You can look at the source code to a LOT of stuff on Unix (not just Linux). So, while it's an excellent point (and a definite bonus), this isn't something that is tied intrinsically to Linux (I'd like to see more large commercial closed-source products using Linux, as an affirmation of it's popularity).
This bothers me. Bump-mapping involves perturbing the geometry of a surface. What is being done in pretty much all of the examples is a clever lighting hack. It incorporates about half of the bump-mapping algorithm, but doesn't bother to perturb... This is annoying, as the actual process to shade the pixels is the same process that could be used to alter their location. Shame on the video-card manufacturers for not properly implementing this. Shame on the developers and manufacturers for propagating the belief that "EMBM" or other styles of shading is actually bump-mapping! Shame on us for being clueless!
That being said, it's a nice fast hack to get something LIKE bump-mapping. Downside : extreme color washout (over-burning or over-dodging for the photo-shop crowd,eh?) Funky lighting considerations need to be taken into account. Oh sure, doing it properly would take a few extra clock-cycles, god forbid anyone would make CUSTOM hardware to do this... oh, wait... the equations almost exactly match what the texturing units do, with a few extra steps... oh wait... there are loads of texturing units all in parallel in hardware now, oh maybe we could get them to run in parallel to do real bump-mapping... gah!
Adaption as speed killer? Hm... no, on-line training of a neural network can be quite fast. In what is implemented in the router, there would be no need at all for adaption.
This is not a traditional feed-forward network trained through back-propagation of error. This is a competitive neural network along the lines of a Kohonen, Grossberg, or Hamming network. What you've been reasoning about is a feed-forward network, which is a completely different beast altogether.
My suggestion: Actually read a book on the subject matter. An excellent one is "Neural Network Design" by Hagan, Demuth, and Beale. It's quite handy, and does an excellent job explaining the how's and the why's...
What you point out, A good programmer is marked by his or her ability to think about a problem and understand a good (not necessarily the best, since there are too many metrics involved to optimize for all of them) way to solve it. , Case-based reasoning (admittedly an infant field, more or less) attempts to solve this problem. Especially CBR's like Meta-AQUA that reason ABOUT reasoning, in an attempt to understand WHY certain problems exist and HOW they were solved.
By remembering how a problem was solved, a previously successfull methodology can be applied to a new problem. Problems (of course) arise depending on how you approach the application of previous solutions. A simple transformation may not work (transforming a sort routine from C to Lisp is not very viable, the domains are too different), so instead a derivational approach can be used. This derivation follows the reasoning that was used behind writing, say, a sort routing in one particular language, and not the specific actions.
This reasoning about reasoning gets you a bit ahead of the neural networks... Not that they don't do a great job, and there is a vaste amount of research just dying to be done on/with them, it's just that they don't currently scale up too well. We don't know a good architecture that begins with a neuron, can create outputs, but can also re-configure itself to deal with new problems. Since a traditional feed-forward network is trained to minimize a particular error function, it doesn't work at all when you introduce a problem that has a completely different error function (Let alone different amounts of input).
Cognitive Psychology has given us some interesting models of how humans might think, and they're relatively easy to mimic with CBRs (I'm sure I've got researchers laughing on the floor with that one:) They give us a good means of generically solving problems. They are, I think, a step in the right direction, especially in that they show a capablitiy for reasoning on multiple domains, a weakness of neural nets (An instance of a neural net is limited to solving only one specifc problem).
Complain about neural networks the next time you use some metal alloy, eh? More than a few have been "discovered" through neural network applications. And let's not think about the untidy world of nuclear wasted disposal. Or anything that does non-linear shortest path finding... Plenty of real-world uses in robotics for AI, among other things... Perhaps you should read some technical articles on Neural Network applications. I'd look at the various IEEE-CS publications. And, of course, that's just Neural Networks. There's all sorts of fun and applied stuff done with Expert Systems, Case-Based Reasoning, Genetic Algorithms&Programming, and more recently, there seems to be a glut of agent-based "stuff".
Er, and you believed pundits talking about events far in the future? I'd be pretty pissed I didn't fly to work and eat little food-pellets if I thought that way:)
And After all that... it's unfortunate that either words are taken out of context, or fools get publicized, but that's the way it is in a lot of fields...
Wasn't it Dijsktra who said, "Anyone attempting to generate random numbers on a digital device is living in a state of sin."? All sorts of fun stuff from my simulation class just comes poring back... At least in terms of the randommness of numbers and their generation. Damn you all for posting to this story!:)
The CD-R drive that it would fit into would probably suck down a lot power as it burned that CD. And let's face it, it would be a pretty damn big camera with a CD stuck in it... And then there's the burning time, you'd probably need a HUGE chunk of buffer RAM since you couldn't burn the data as fast as you can record it...
Course, it would be nice to get more video out of it than just 20 minutes... Personally I like the TRV-900 they make: very pretty.
You know, I don't recall... But the pigeon scenes were hilarious... Er, not intentionally funny, just howlingly funny if you've seen a John Woo film before.
How is it possible that "Hard Target" with VanDamme is a much better movie than this one? Perhaps they should have got someone to write a script. It's like they just pieced together some scenes, and worked out what to say on the fly... very clever... erg...
Shanghai Noon, on the other hand, was absolutely great. Entertaining action, and farcical comedy. Delicous summer movie (no plot, just comedy and action, eh?).
Did anyone else notice that the theme song, just like the movie, was awful in comparison to the first? Handy hint: re-use/re-license work done by Bono and The Edge, they're better musicians than whoever they got for this MI2, eh?
Throwing more chips at processing tasks is nice. However, the article nicely pointed out some severe problems Apple has with bandwidth, and linking together all of their components on a nice fast bus. That's what makes SGI's machines so beautiful, and Sun's so remarkably, well, sun-like:) I think PCs have pretty poor bus designs, but they've got better throughput than current Macs. Bonus points to whoever can implement something like Wildfire or SGI's NUMA hardware.
I am but a lowly grad-student in CS (A.I. being my main interest), and I did nothing but cringe through his (Bill Joy's) paper. I think your average man off the street, or say, 100 monkeys, could have come up with it. The article made me quite painfully aware of his complete lack of knowledge about AI. This current article looks to be mildly educated rambling. Totally uninteresting to a researcher, and bandies around enough terms and references to confuse your average layperson. Try checking out the Principia Cybernetica for similar wackiness. They're entertaining, but the what-if scenarios they posit are so far beyond anything we can do today, or probably in the next 50 or 100 years, that they're meaningless. They're this decades version of 50's pulp fiction.
Personally, this article at least raises some interesting questions and/or directions AI research can take (well in the future for the most part). Unlike Bill Joy's drivel.
Ah, all done venting now:)
And in response to your post, I agree completely:)
It's nice to see they fixed the case design from their original Intel stuff. We've got a 320 (I believe) here, and the damn sliding front access door/case swoopiness made it a pain to put anything in the floppy/cdrom/zip drives. Of course, they're probably the same people who didn't adequately cool the SGI 02's (can you say "toaster"?). That being said, we love our Octanes and Onyxs.
Slashdot Mirror
KnowledgeKinetics (www.knowledgekinetics.net) handles document (content) management as well as a whole slew of other things. Quite a nifty set of tools.
Oddly enough, you can go to Ball Aerospace and ask for just that... satellite buses
Binary data in XML is stored uu-encoded... ( I've been told that's the standard way of doing things with that... me, I prefer putting text data in --easier to debug, so what do I know?)
The funniest thing is that no matter how many times someone complains about the atrocious grammar or spelling... nothing is done. Remember the "put Hemos through English class" sigs? Ah well, it makes for entertaining commentaries.
It was pretty easy, but I had run him through the printer first, so he was on an 8x11 piece of paper...
They are so Windows-centric that they forget about the importance of portability. Like MFC, and half the other shit they make which ends up stangling a developer to a particular platform. There's a reason Java will remain popular, and that's Sun's devotion to portability and stability. ?
Does it bother you that variables within C# are all COM objects? Does your operating system support COM? Do you have a program that allows COM objects to exist on your machine? Then it can't run C#. There goes portability...
Stability. Hmmm. Complete control over the spec would imply stability. Not open-ness, but at least stability. For all the cruft in Microsoft's closed API's, there's very little change --> nice and stable, a bit like Java (I know, heresy :)
Absolutely! It would make life interesting... er, at least for about a week, then it would get old and annoying. But what a week!
So it's crazy. But *I* like the idea. That way there can be more seamus.org's than just the one (why, oh why was I so slow? :) Of course, eventually people would get pissy about having "convenient" IP addresses to type in... can you imagine network card and router sales/manufacturing in this future? Neat...
Doesn't Microsoft's current website work this way... (i.e. hopelessly inadequate). Could they have a worse method of accessing their "Knowledge Base" aka "how to work around our bugs"?).
What's the minimum for cos(x)? One variable, massive search space... (well, cheating really, but you get the point... :) Finding minimums, is... hard... And numerical optimizers are REALLY bad at doing non-linear problems, having worked with CPLEX and a few others of its ilk. CPLEX, in specific, is great for linear problem solving, I rather liked using it... but it did squat when it came time to work on the non-linear problems (nuclear waste stream disposal -- very cool stuff, check out M.A.W.S.)
I wish I could find the link to an article I read about a year ago... basically saying that the people who originally used the SUV type vehicles for real work (ranchers, etc.) can now no longer afford to buy them... psychotic, eh?
While the majority of Open Source (or Free) projects are for Linux, there are plenty out there for Windows (maybe there should be more). You can look at the source code to a LOT of stuff on Unix (not just Linux). So, while it's an excellent point (and a definite bonus), this isn't something that is tied intrinsically to Linux (I'd like to see more large commercial closed-source products using Linux, as an affirmation of it's popularity).
That being said, it's a nice fast hack to get something LIKE bump-mapping. Downside : extreme color washout (over-burning or over-dodging for the photo-shop crowd,eh?) Funky lighting considerations need to be taken into account. Oh sure, doing it properly would take a few extra clock-cycles, god forbid anyone would make CUSTOM hardware to do this... oh, wait... the equations almost exactly match what the texturing units do, with a few extra steps... oh wait... there are loads of texturing units all in parallel in hardware now, oh maybe we could get them to run in parallel to do real bump-mapping... gah!
This is not a traditional feed-forward network trained through back-propagation of error. This is a competitive neural network along the lines of a Kohonen, Grossberg, or Hamming network. What you've been reasoning about is a feed-forward network, which is a completely different beast altogether.
My suggestion: Actually read a book on the subject matter. An excellent one is "Neural Network Design" by Hagan, Demuth, and Beale. It's quite handy, and does an excellent job explaining the how's and the why's...
By remembering how a problem was solved, a previously successfull methodology can be applied to a new problem. Problems (of course) arise depending on how you approach the application of previous solutions. A simple transformation may not work (transforming a sort routine from C to Lisp is not very viable, the domains are too different), so instead a derivational approach can be used. This derivation follows the reasoning that was used behind writing, say, a sort routing in one particular language, and not the specific actions.
This reasoning about reasoning gets you a bit ahead of the neural networks... Not that they don't do a great job, and there is a vaste amount of research just dying to be done on/with them, it's just that they don't currently scale up too well. We don't know a good architecture that begins with a neuron, can create outputs, but can also re-configure itself to deal with new problems. Since a traditional feed-forward network is trained to minimize a particular error function, it doesn't work at all when you introduce a problem that has a completely different error function (Let alone different amounts of input).
Cognitive Psychology has given us some interesting models of how humans might think, and they're relatively easy to mimic with CBRs (I'm sure I've got researchers laughing on the floor with that one :) They give us a good means of generically solving problems. They are, I think, a step in the right direction, especially in that they show a capablitiy for reasoning on multiple domains, a weakness of neural nets (An instance of a neural net is limited to solving only one specifc problem).
Ah well, just some crack-addled rambling, eh?
Er, and you believed pundits talking about events far in the future? I'd be pretty pissed I didn't fly to work and eat little food-pellets if I thought that way :)
And After all that... it's unfortunate that either words are taken out of context, or fools get publicized, but that's the way it is in a lot of fields...
High Resolution of the digital format? Cute...
I'm sick of looking at avi's and mpegs with blocky flecks of color from compression, when are movie theatres going to fix this? Oh, wait... :)
But Ebay also sells cars, paintings, etc. that cost far more than that... Odd.
Wasn't it Dijsktra who said, "Anyone attempting to generate random numbers on a digital device is living in a state of sin."? All sorts of fun stuff from my simulation class just comes poring back... At least in terms of the randommness of numbers and their generation. Damn you all for posting to this story! :)
Course, it would be nice to get more video out of it than just 20 minutes... Personally I like the TRV-900 they make: very pretty.
How is it possible that "Hard Target" with VanDamme is a much better movie than this one? Perhaps they should have got someone to write a script. It's like they just pieced together some scenes, and worked out what to say on the fly... very clever... erg...
Shanghai Noon, on the other hand, was absolutely great. Entertaining action, and farcical comedy. Delicous summer movie (no plot, just comedy and action, eh?).
Did anyone else notice that the theme song, just like the movie, was awful in comparison to the first? Handy hint: re-use/re-license work done by Bono and The Edge, they're better musicians than whoever they got for this MI2, eh?
Throwing more chips at processing tasks is nice. However, the article nicely pointed out some severe problems Apple has with bandwidth, and linking together all of their components on a nice fast bus. That's what makes SGI's machines so beautiful, and Sun's so remarkably, well, sun-like :) I think PCs have pretty poor bus designs, but they've got better throughput than current Macs. Bonus points to whoever can implement something like Wildfire or SGI's NUMA hardware.
Personally, this article at least raises some interesting questions and/or directions AI research can take (well in the future for the most part). Unlike Bill Joy's drivel.
Ah, all done venting now :)
And in response to your post, I agree completely :)
It's nice to see they fixed the case design from their original Intel stuff. We've got a 320 (I believe) here, and the damn sliding front access door/case swoopiness made it a pain to put anything in the floppy/cdrom/zip drives. Of course, they're probably the same people who didn't adequately cool the SGI 02's (can you say "toaster"?). That being said, we love our Octanes and Onyxs.