I dunno. That site consists of a bunch of people with a REALLY good sense of humor, or complete and blithering lunatics. Articles about how you should traumatize cattle so the meat tastes good!? I don't want to take it seriously, but I got the impression that several of the posters there do. I'm a bit unsettled by that.
But, given that their main poll had 137 votes, I guess most people have already voted with their [metaphorical] feet.
Following your lead, I have deduced that Othello must be a game-theoretic treatsie, while the Comedy of Errors is clearly a C compiler.
I suspect that All's Well that Ends Well must be an early formulation of the halting problem, but I'm still waiting for the mechanical proof solver to come back with results. It may be that the last word of the title is a transcription error.
The one thing that sets VOD appart from the normal uses of MP3 or MPEG is that the VOD can require a cray to compress, as long as the decompression is easy. Just like fractal compression, which is theoretically possible, just REALLY time consuming to compress.
One way might be to send executable code: stored procedures that manipulate image regions. Think of it as motion prediction on steriods. Now, I don't have a clue as to how the compresssor would figure out these code snippets. Exhaustive search? Mind you, it doesn't necessarily have to be turing complete, but perhaps a very advanced command langauge.
In fact, VOD compression doesn't have to be 100% automatic. You can very well justify having an operator select regions where the compressor should try harder to optimize, and what kind of optimization is likely to be needed. Say, you mark _this_ as background for the next X frames, _this_ is foreground, _that_ is a repetetive element, so store that in our image cache -- see _there_ it is again.
These sort of human driven superoptimisations might be able to achieve a very high level of compression with acceptable results.
A somewhat constructive question, but along the lines of the above:
How will this 25X be programmed? It seems to be like the CM-1. You have all these tiny processors comunicating quickly, but with almost no local instructions or data.
I tend to think that the compactness of Forth and the ~ 1 KiloInstruction availible to each processor will be enough to store a useful program in each core, while still leaving room for housekeeping chores. I'll be willing to accept that IO won't be a complete killer.
But! I'm dammed if I'm going to sit and hand optimize the communication between 25 cores. What if I decide that my application really needs 25X25 cores? I mean, Occam wasn't that fun, nor was Lisp*. Ask Transputer or Danny Hillis.
So for the question:
Do you have any programming tools where I can express my algorithm in a communication neutral way, and then have it tuned for the architecture at hand? Or is it not that hard to make this architecture fly?
where to start? Java has not actually RUN on a virtual machine for many years. Compiled to it, yes, but JITted to native code in pretty much all cases.
So you can see the JVM as a convenient intermediate stack format.
Translating to registers is notso hard, if you have a few registers handy. You can view the register file as an array implementing a circular list (the stack). When you catch up to the tail, start flushing to memory. Using this technique, caller saves is perfectly implementable (ie, the caller guarantees that the callee will not overwrite the caller's data. It is up to the callee to not overwrite its own).
I believe the compiler long ago surpassed the programmer as the best suited to optimize whole programs. Flow analysis, specialisation, profiling feedback... all of these just are too much work for the programmer to optimise.
When humans optimize, some can do a better job than most compilers for short sequences of code, but they spend too much time optimizing the wrong thing, at the expense of bugs, development time, and maintenance.
You could never fit a show inside the message size limits most email has. However, if I have Broadband and you have broadband, then I could send you a somewhat secure URL that points to my net-enabled PVR where you could stream it from.
I'd imagine that the URL is all you are emailing.
Of course, this feature alone is going to keep the lawyers busy busy. And for once, I kinda see their point.
The thing that sets this device appart from VCRs is that it can record the entire commercial and not incurr a penalty for fastforwarding through it. So the PVR has the luxury of only needing to decide whether that was a commercial, not whether this is one.
The way I would implement it is to record everything and use a user-tunable heuristic to mark blocks as likely commercials which are then skipped during playback. If you get it wrong, the user can view the block w/o skipping it. For example, commercials tend to be a bit louder than average programming. You know that there will be a big change (got that from another post) in picture before and after the block, AND you know that it will be a multiple of 30 seconds.
The first and last of these criteria, in conjunction with post-facto marking rather than the pre-commercial guessing makes the PVR much better suited to the task of identifying commercials than VCRs.
Only digital product placement is likely to be able to foil these sorts of heuristics, esp if the user is able to write their own rules and assign levels of certainty to them.
the DMCA says very little about breaking encryption. Rather it speaks about circumventing copyright protection devices. I don't see what copyright the studios would be breaking if they reverse engineer the replay heuristics.
On the contrary; All studios need to do is to create a licence that their programming is licenced for viewing as a whole. Any attempt to select parts of the programming could be argued to be an attempt to circumvent licencing, and hence by extension copyright.
Not a watertight case in either direction, but it seems that the DMCA would once again play into the big studios hand if it has any bearing at all.
ya know... this is turning into a real pet peeve of mine:
100,000 times thinner, reduced by 200%, 3 times less power...
I see what they mean, most of the time (although the 200% is kinda vague) but if you mean 1/100,000th as thick, that is not really the same as 100,000 times thinner. I tend to view thin and thick as complements under addition, not reciprocals.
mutter mutter. Pedants are never bored, but tend to be boring.
Say I release versions 1,2,3, and 4 of my Program under the GPL. These versions, by being "released under the GPL", are exactly what you say : GPL 4Life.
I am under no obligation of releasing version 5 under the GPL. I can also re-release versions 1 - 4 under someother licence. For example, I can sell them as closed source to M$. However, these actions in no way invalidate the licence on the existing released code. And only I can take these actions.
You on the other hand, are free and encouraged to build on my GPLed code, but if you do so, it MUST be released under the GPL 4Life as well. You have no rights to resell/relicence your _derivative_ work. ( I guess I could sell/transfer these rights to you tho, once again, without invalidating the licence on already released code).
great idea; but it fails on the glacially slow refresh times for LCDs. Many people don't realise that when an LCD pixel turns on or off, the molecules in the pixel physically reconfigure themselves. It takes time for them to react to the... erm guessing here... elecrical field applied to them.
However, TI had a display-on-a-chip (using mirrors mounted on micro-actuators, iirc) that used exactly this color cycling technique. As always, this was discussed on/. a while back.
But search is down, so I can't point you to the discussion. Make do with the original site instead:
http://www.dlp.com/dlp/default.asp
click on the "see how it works" for a flash animation. The "color" tab describes the process. With a physically spinning color disk. How oldskool!
The problem is that even if the DCMA gets overturned in the US as being unconstutional or just plain stupid, now you have to overturn it in all of these other countries that have been forced to pass it.
Likely, they will be unwilling to look like complete marionettes, and thus resist repealing the law immediately, arguing that they actually passed it because they thought it was a good idea, not because the US forced them to.
was it 8MB or 8 Mword? I seem to recall crays using some non-standard wordsize.
While I'm at it, here's another:
How fast were those 160 MFlops; I suspect that sustained throughput would play a big part in it. Is that about as fast -- in real world speed, not peak tight loop speed -- as today's desktops, or have we finally caught up to that?
Re:Misc Icons - Opera
on
Netscape 6.1
·
· Score: 2
Multiple Document Interface.
I HATE that. I want the speed and simplicity of Opera, but I just can't use those MDI programs.
Even MS-Word, for which I think microsoft invented/developed the whole MDI thing, has given up on that.
I mean, let the window manager take care of windows. Mind you, I love emacs's buffers, but that's different...
Re:there's actually a few useful features
on
Netscape 6.1
·
· Score: 2
dictionary?!
when did you last use a dictionary for anything other than an authored document, as opposed to an email or news posting?
Speaking of news, how is it for that? Is NS 6.1 better than the crap 4.7? I swear that thing has an O(N^2) algorithm for loading group overviews. Anthing over 1000 msgs, and you're out to lunch before its done. *mp3 has 60000 msgs these days..
M$ almost never gets it right, the first, second, or even third time (it is widely acknowledged even among their most ardent fans, that Windows -- their flagship -- was only usable in its FOURTH incarnation: WfWG).
But they always get it right in the end.
Word is almost half as good as emacs (well, it's not perfect, while emacs only lacks multithreading and per-buffer pixmap backgrounds), which puts it far ahead of everything else out there. (after all, what is a post without a bit of polemic?)
Explorer is no longer Exploder.
Excel and Access are no longer Edsel and Absess.
And One Day, they will even get IIS and Exchange right.
The main difference from other players being that they can afford to get it wrong for a Long Time, both in money and mindshare. By making sure that ever version is better than the last, and making a big marketing hoopla about that, they're all set.
In a way, it suprises me that they haven't spun off their brand management division. These guys are Fscking geniuses.
How is mozilla for news? It appears that netscape has an O(n^2) algorithm for sorting messages in newsgroups, and very poor multipart support. Things like this are important in the post napster age.
I'm suprised. I has just assumed that they were using some dynamic feedback system, so that when the peice moves in the right direction, they keep it up, and if it moves in the wrong direction, they do something else.
I'd expect that a neural net driving a set of vibration generators would work quite well.
yes.
arguably, bob has been negligent in letting his computer be infected. A very clear analogy is Bob keeping the confidential documents in a physically insecure place, where a casual visitor can easily read them.
It is then up to the courts to decide to which extent Bob has been negligent. Has he been negligent in running an OS which is known to have many security holes? Is he responisble for keeping it secure?
Guru Bruce Schneier predicts that computer security will only become a concern for people like bob when their insurance premiums and legal risk of prosecution hurt them where it counts.
This is a commonly recurring theme on comp.risks (well recommended for friday afternoon reading).
1) Mr. Lessig is a lawyer, so I assume he has read up on this stuff, but his claims that the US were alone in having DCMA laws suprised me. I was under the impression that europe was in fact drafting even more sweeping ones?
2) What outrage? This trial will never make to the front page, much less the evening news. The concepts are too abstract for your average joe to grasp in the 10 seconds attention span he/the newscasters will alot to the news item.
3) The only hope is (erm, I THINK this is the process, but std discl) is that he is convicted, and the appeal makes it all the way up to the supreme court which can then strike the law down as unconstitional. However, you'll recall that Dubya is appointing a couple of supreme judges this term, and we can expect that Big Business will bend the ear of the administration, so I wouldn't hold my breath waiting for the supremes to be agree with techolibertarians like us.
4) I would LOVE to be proved wrong, but... oink flap oink flap.
No big difference, you'd think. But energy varies as the square of the velocity, so this reduces yeild to (v_i^2/2) Joules per Kg (~ just shy of 1 MegaJoule per kg). This is 1/220 th of your value. Assume we could drop a KiloTonne (assuming metric tonnes: 1 tonne == 1000 Kg), that would give us 1000 Gigajoules, or 1/4 as effective as good old TNT.
Hrm. Is that so? I mean, you can easily calculate the kinetic energy per Kg. What is the terminal velocity of a warhead shaped thingie? A google search for joule kilotonne gives 4200 GigaJoule per kilotonne. The maximum payload of the x33 is 25000 kg, so counting backwards, we see that the payload needs to reach ~ 18300 m/s just to get a kilotonne yeild (and that's for a 40th of a kilotonne, which isn't great). Furthermore, that is mach 61, which I suspect is beyond the terminal velocity of anything in freefall (ie, without incoming velocity).
However, if you want to knock out infrastructure without serious casualties, dropping 25000 unpiloted slugs might do the trick. or 50000 1/2 kilo ones that have explosives. Especially if you can drop them in piloted packages to disperse at lower altitudes. Terminal velocity is your friend, since it basically means you loose nothing by dispersing late (presumably the smaller slugs have higher terminal velocity than the delivery package).
So actually, I see this as a fairly humane rapid suppression device.
and I ended up posting it twice. The browser crashed the first time, so I thought it hit the bit bucket. So if there ever was call for a post to be marked redundant, that'd be me.
Slashdot Mirror
I dunno. That site consists of a bunch of people with a REALLY good sense of humor, or complete and blithering lunatics. Articles about how you should traumatize cattle so the meat tastes good!? I don't want to take it seriously, but I got the impression that several of the posters there do. I'm a bit unsettled by that.
But, given that their main poll had 137 votes, I guess most people have already voted with their [metaphorical] feet.
fscking genius! Bravo, my good man.
Following your lead, I have deduced that Othello must be a game-theoretic treatsie, while the Comedy of Errors is clearly a C compiler.
I suspect that All's Well that Ends Well must be an early formulation of the halting problem, but I'm still waiting for the mechanical proof solver to come back with results. It may be that the last word of the title is a transcription error.
heh.
slashdot links to browser forkbomb.
cute.
Me, I had to reboot; I was unable to get control before the windows had spawned to infinity, and beyond!
The one thing that sets VOD appart from the normal uses of MP3 or MPEG is that the VOD can require a cray to compress, as long as the decompression is easy. Just like fractal compression, which is theoretically possible, just REALLY time consuming to compress.
One way might be to send executable code: stored procedures that manipulate image regions. Think of it as motion prediction on steriods. Now, I don't have a clue as to how the compresssor would figure out these code snippets. Exhaustive search? Mind you, it doesn't necessarily have to be turing complete, but perhaps a very advanced command langauge.
In fact, VOD compression doesn't have to be 100% automatic. You can very well justify having an operator select regions where the compressor should try harder to optimize, and what kind of optimization is likely to be needed. Say, you mark _this_ as background for the next X frames, _this_ is foreground, _that_ is a repetetive element, so store that in our image cache -- see _there_ it is again.
These sort of human driven superoptimisations might be able to achieve a very high level of compression with acceptable results.
A somewhat constructive question, but along the lines of the above:
How will this 25X be programmed? It seems to be like the CM-1. You have all these tiny processors comunicating quickly, but with almost no local instructions or data.
I tend to think that the compactness of Forth and the ~ 1 KiloInstruction availible to each processor will be enough to store a useful program in each core, while still leaving room for housekeeping chores. I'll be willing to accept that IO won't be a complete killer.
But! I'm dammed if I'm going to sit and hand optimize the communication between 25 cores. What if I decide that my application really needs 25X25 cores? I mean, Occam wasn't that fun, nor was Lisp*. Ask Transputer or Danny Hillis.
So for the question:
Do you have any programming tools where I can express my algorithm in a communication neutral way, and then have it tuned for the architecture at hand? Or is it not that hard to make this architecture fly?
ah.
where to start? Java has not actually RUN on a virtual machine for many years. Compiled to it, yes, but JITted to native code in pretty much all cases.
So you can see the JVM as a convenient intermediate stack format.
Translating to registers is notso hard, if you have a few registers handy. You can view the register file as an array implementing a circular list (the stack). When you catch up to the tail, start flushing to memory. Using this technique, caller saves is perfectly implementable (ie, the caller guarantees that the callee will not overwrite the caller's data. It is up to the callee to not overwrite its own).
I believe the compiler long ago surpassed the programmer as the best suited to optimize whole programs. Flow analysis, specialisation, profiling feedback... all of these just are too much work for the programmer to optimise.
When humans optimize, some can do a better job than most compilers for short sequences of code, but they spend too much time optimizing the wrong thing, at the expense of bugs, development time, and maintenance.
You could never fit a show inside the message size limits most email has. However, if I have Broadband and you have broadband, then I could send you a somewhat secure URL that points to my net-enabled PVR where you could stream it from.
I'd imagine that the URL is all you are emailing.
Of course, this feature alone is going to keep the lawyers busy busy. And for once, I kinda see their point.
The thing that sets this device appart from VCRs is that it can record the entire commercial and not incurr a penalty for fastforwarding through it. So the PVR has the luxury of only needing to decide whether that was a commercial, not whether this is one.
The way I would implement it is to record everything and use a user-tunable heuristic to mark blocks as likely commercials which are then skipped during playback. If you get it wrong, the user can view the block w/o skipping it. For example, commercials tend to be a bit louder than average programming. You know that there will be a big change (got that from another post) in picture before and after the block, AND you know that it will be a multiple of 30 seconds.
The first and last of these criteria, in conjunction with post-facto marking rather than the pre-commercial guessing makes the PVR much better suited to the task of identifying commercials than VCRs.
Only digital product placement is likely to be able to foil these sorts of heuristics, esp if the user is able to write their own rules and assign levels of certainty to them.
the DMCA says very little about breaking encryption. Rather it speaks about circumventing copyright protection devices. I don't see what copyright the studios would be breaking if they reverse engineer the replay heuristics.
On the contrary; All studios need to do is to create a licence that their programming is licenced for viewing as a whole. Any attempt to select parts of the programming could be argued to be an attempt to circumvent licencing, and hence by extension copyright.
Not a watertight case in either direction, but it seems that the DMCA would once again play into the big studios hand if it has any bearing at all.
what a surprise
ya know... this is turning into a real pet peeve of mine:
100,000 times thinner, reduced by 200%, 3 times less power...
I see what they mean, most of the time (although the 200% is kinda vague) but if you mean 1/100,000th as thick, that is not really the same as 100,000 times thinner. I tend to view thin and thick as complements under addition, not reciprocals.
mutter mutter. Pedants are never bored, but tend to be boring.
Say I release versions 1,2,3, and 4 of my Program under the GPL. These versions, by being "released under the GPL", are exactly what you say : GPL 4Life.
I am under no obligation of releasing version 5 under the GPL. I can also re-release versions 1 - 4 under someother licence. For example, I can sell them as closed source to M$. However, these actions in no way invalidate the licence on the existing released code. And only I can take these actions.
You on the other hand, are free and encouraged to build on my GPLed code, but if you do so, it MUST be released under the GPL 4Life as well. You have no rights to resell/relicence your _derivative_ work. ( I guess I could sell/transfer these rights to you tho, once again, without invalidating the licence on already released code).
great idea; but it fails on the glacially slow refresh times for LCDs. Many people don't realise that when an LCD pixel turns on or off, the molecules in the pixel physically reconfigure themselves. It takes time for them to react to the... erm guessing here ... elecrical field applied to them.
/. a while back.
However, TI had a display-on-a-chip (using mirrors mounted on micro-actuators, iirc) that used exactly this color cycling technique. As always, this was discussed on
But search is down, so I can't point you to the discussion. Make do with the original site instead:
http://www.dlp.com/dlp/default.asp
click on the "see how it works" for a flash animation. The "color" tab describes the process. With a physically spinning color disk. How oldskool!
hrm.
The problem is that even if the DCMA gets overturned in the US as being unconstutional or just plain stupid, now you have to overturn it in all of these other countries that have been forced to pass it.
Likely, they will be unwilling to look like complete marionettes, and thus resist repealing the law immediately, arguing that they actually passed it because they thought it was a good idea, not because the US forced them to.
just a stupid question:
was it 8MB or 8 Mword? I seem to recall crays using some non-standard wordsize.
While I'm at it, here's another:
How fast were those 160 MFlops; I suspect that sustained throughput would play a big part in it. Is that about as fast -- in real world speed, not peak tight loop speed -- as today's desktops, or have we finally caught up to that?
Multiple Document Interface.
I HATE that. I want the speed and simplicity of Opera, but I just can't use those MDI programs.
Even MS-Word, for which I think microsoft invented/developed the whole MDI thing, has given up on that.
I mean, let the window manager take care of windows. Mind you, I love emacs's buffers, but that's different...
dictionary?!
when did you last use a dictionary for anything other than an authored document, as opposed to an email or news posting?
Speaking of news, how is it for that? Is NS 6.1 better than the crap 4.7? I swear that thing has an O(N^2) algorithm for loading group overviews. Anthing over 1000 msgs, and you're out to lunch before its done. *mp3 has 60000 msgs these days..
absolutely.
M$ almost never gets it right, the first, second, or even third time (it is widely acknowledged even among their most ardent fans, that Windows -- their flagship -- was only usable in its FOURTH incarnation: WfWG).
But they always get it right in the end.
Word is almost half as good as emacs (well, it's not perfect, while emacs only lacks multithreading and per-buffer pixmap backgrounds), which puts it far ahead of everything else out there. (after all, what is a post without a bit of polemic?)
Explorer is no longer Exploder.
Excel and Access are no longer Edsel and Absess.
And One Day, they will even get IIS and Exchange right.
The main difference from other players being that they can afford to get it wrong for a Long Time, both in money and mindshare. By making sure that ever version is better than the last, and making a big marketing hoopla about that, they're all set.
In a way, it suprises me that they haven't spun off their brand management division. These guys are Fscking geniuses.
ramble
rambe
How is mozilla for news? It appears that netscape has an O(n^2) algorithm for sorting messages in newsgroups, and very poor multipart support. Things like this are important in the post napster age.
Mathematics!?
I'm suprised. I has just assumed that they were using some dynamic feedback system, so that when the peice moves in the right direction, they keep it up, and if it moves in the wrong direction, they do something else.
I'd expect that a neural net driving a set of vibration generators would work quite well.
hrm?
yes.
arguably, bob has been negligent in letting his computer be infected. A very clear analogy is Bob keeping the confidential documents in a physically insecure place, where a casual visitor can easily read them.
It is then up to the courts to decide to which extent Bob has been negligent. Has he been negligent in running an OS which is known to have many security holes? Is he responisble for keeping it secure?
Guru Bruce Schneier predicts that computer security will only become a concern for people like bob when their insurance premiums and legal risk of prosecution hurt them where it counts.
This is a commonly recurring theme on comp.risks (well recommended for friday afternoon reading).
A couple of points:
... oink flap oink flap.
1) Mr. Lessig is a lawyer, so I assume he has read up on this stuff, but his claims that the US were alone in having DCMA laws suprised me. I was under the impression that europe was in fact drafting even more sweeping ones?
2) What outrage? This trial will never make to the front page, much less the evening news. The concepts are too abstract for your average joe to grasp in the 10 seconds attention span he/the newscasters will alot to the news item.
3) The only hope is (erm, I THINK this is the process, but std discl) is that he is convicted, and the appeal makes it all the way up to the supreme court which can then strike the law down as unconstitional. However, you'll recall that Dubya is appointing a couple of supreme judges this term, and we can expect that Big Business will bend the ear of the administration, so I wouldn't hold my breath waiting for the supremes to be agree with techolibertarians like us.
4) I would LOVE to be proved wrong, but
Thanks for the heads up about drag and friction. You made a small error in the velocity calculation: the max theoretical impact speed is
v = g*t_i
where t_i of impact, in seconds from release.
So we have:
v(t) = g*t
d(t) = (1/2)g*t^2
so all we have to do is solve d(t) = 100000, for t
t = sqrt(2*100000/g) (~= 145 sec)
so we plug this into v(t):
v_i = g*t = g*sqrt(2*100000/g) = sqrt(2*100000 * g) (~= 1453 m/s)
No big difference, you'd think. But energy varies as the square of the velocity, so this reduces yeild to (v_i^2/2) Joules per Kg (~ just shy of 1 MegaJoule per kg). This is 1/220 th of your value. Assume we could drop a KiloTonne (assuming metric tonnes: 1 tonne == 1000 Kg), that would give us 1000 Gigajoules, or 1/4 as effective as good old TNT.
Not so great.
Near-space altitude obviates the need for warhead
Hrm. Is that so? I mean, you can easily calculate the kinetic energy per Kg. What is the terminal velocity of a warhead shaped thingie? A google search for joule kilotonne gives 4200 GigaJoule per kilotonne. The maximum payload of the x33 is 25000 kg, so counting backwards, we see that the payload needs to reach ~ 18300 m/s just to get a kilotonne yeild (and that's for a 40th of a kilotonne, which isn't great). Furthermore, that is mach 61, which I suspect is beyond the terminal velocity of anything in freefall (ie, without incoming velocity).
However, if you want to knock out infrastructure without serious casualties, dropping 25000 unpiloted slugs might do the trick. or 50000 1/2 kilo ones that have explosives. Especially if you can drop them in piloted packages to disperse at lower altitudes. Terminal velocity is your friend, since it basically means you loose nothing by dispersing late (presumably the smaller slugs have higher terminal velocity than the delivery package).
So actually, I see this as a fairly humane rapid suppression device.
and I ended up posting it twice. The browser crashed the first time, so I thought it hit the bit bucket. So if there ever was call for a post to be marked redundant, that'd be me.