do what Kanasta said, but flip one image left-right. now affix a mirror inbetween the two, perpendicular to the screen. press your nose agains the mirror, and voila, fusing the images becomes a snap.
The one drawback is that all images appear to be offcenter (behind the reflective side of the mirror, as it were).
Ok, so replying to myself is kinda gimpy, but I had what I thought was a kick ass idea:
given that SOO many of us have notebooks with fairly standard dot-pitch LCDs, wouldn't it be cool if a game manufacturer shipped a fresnel lens cover for your LCD with their 1-st person shoot-em-up?
Every other row or better yet column of pixels would be focussed the in the proper direction, towards each eye. Basically do optically what the MIT people do with fancy schmancy beam splitters.
Ok, so you'd have to keep your head still while playing, and the quality might not be up there, but it would be cool and cheap.
Some problems with the idea might be that getting the lens to line up properly with the pixels would be a hassle. But here's the cool part. LCDs have subpixels, baby! yeah! just shift the image left or right 1 subpixel at a time until the display comes out right.
You might have to sacrifice every other column of pixels too (LxRxLxR...) in order for it to come out nice, but that still leaves you 330 odd horisontal. good enough for a shootie, Ithink.
This'd suck processor like you wouldn't believe, but I will bet you anything thing that true 3D at low framerate (and since you have an LCD, you're kinda limited in that regard anyway) over 2D at 100fps. and if you cut resolution, you have less to do anyway.
ok, so I thought that was pretty cool, but I'm biased. I think these ideas are so cool that I'm gonna have to claim copyright to them right now.
From the MITwebsite (http://www.media.mit.edu/groups/spi/new_AutoM1.ht ml),
it appears that the system actually projects different images into each eye. This one appears to force you to keep your head still, while the MIT one tracks you via IR. Anyway, knowing where your head is, it uses special magic (erm... polarising filter/beamsplitter) to direct the proper image into each eye.
I think they're taking advantage of the narrow viewing angle of the lcd, and tu(r)ning every other row of pixels to each eye.
As far as I can read from the article, the machine takes the 3-10 Mb/s input uncompressed rate and makes decompresses it to... 10 Mb/s?
Doesn't that seem a little slow?If you want to drive a megapixel display, that's something like 1/3 fps at 24 bit color.
So if we assume that the New Scientist got the output wrong, and its really outputting 270 Mb/s, that's more reasonable. Then you get a megapixel display at 30 fps if you go for [slightly] lower color resolution (and that's doable).
But then the article goes on to say how he wants to connect 27 of these dohickeys to saturate the SDI pipe, which again brings them down to 10 Mb/s.
Almost. SMP and DSM (Distributed Shared Memory, for those not up on acronyms) would be more correct. Some memory is local, some is not. By non-local, that means it could be in another box, or maybe just off another bus.
The important thing is that all memory is accessed as if it were local (just ask for it by address). The hard part is of course getting all processes that want to stomp over the same memory to share it nicely. This involves both page migration (of DSMfame) to processes, and process migration. And of course, at all times maintaining coherency between all copies of a page.
Ultimately, you want all processes that use the same memory to be on the same CPU that memory, or at least near by.
This is hard stuff, as all by scientific applications have hard-to-predict paging behavior. But then scientific apps is what pays the bills for people who need these machines.
A munition is much heavier than the arms that the 2nd ammendment allows. Munitions include shells for heavy artillery and bombs, both of which you most definately are not allowed to own.
As far as Iknow, CD Spins here in boston will let you listen to any CD in the store. Now that they have new CDs, Inever end up going to tower for some reason... strange.
And for those of you who think MHz is a consideration, bzzzt! In the world of 24x7 server farms, torque counts more than horsepower. The UltraSparc cpu can handle a much greater load without sweat than most other cpus. The UltraSparc was designed to handle massive loads.
Not to be abtuse here, but I don't get your analogy. Straight line speed is one thing, data throughput is another thing. So I thought maybe you meant that while Sparcs are slower that the PIII, they typically come with better I/O infrastructures. I'm all set to agree with you there. After all IBM sells big iron on this argument alone.
But then you said load, and I'm thinking maybe you meant straight line speed anyway.
I read the partitioning to mean virtualizable -- ie able to run operating systems as processes, nested arbitrarily. This is what the s390 guy did who had 40000 odd linux kernels running simultaneously.
Note that vmware does a variant of this, but because they do it on the i386, they have to jump trough hoops. The i386 isn't virtualisable, so they have to scan for instructions that would expose this and rewrite them to a trap to some meta-level code.
It stands to reason that if you have the ability to run two OSs on one CPU, the capability carries over to SMP configurations too.
He's got a point. Having the display driver which knows about monitor size and resoulution figure out the pixels from the point size is the only logical way to go.
And it really shouldn't be that hard to retrofit. Designing towards the lowest common denominator is ok, as long as the non-lowest aren't unduly harmed, or the lowest aren't to far below the next rung up. In this case, there's little excuse for the lowest rung even existing.
Hrm. I often forget about opera, 'cause last I looked it didn't run on unix and used those damn windows -in- windows. Horrible. Doesn't really matter how nice the inside of the windows are when the rest of the browser doesn't make it to the starting line
I'm always wary of such estimations. How do you go about tallying up damages?
"oh, our department was shut down for 24 hours, so our damages are wages paid for no productivity?"
That seems really close to the lost income arguments. The great thing about lost income is that it is impossible to justify; "the virus stopped me from having time to buy a lottery ticket, and I was going to pick exactly the numbers that won!"
"Given the global impact of the I Love You virus and the growing threat of malicious hackers, we strongly believe we must take the unprecedented step of limiting certain popular functionality in Outlook to provide a significant, additional security option for our customers,"
...which, of course, has probably triggered thousands of email gateway scanners to throw the message back as containing a worm...duh!
I should hope that any firewall worth its salt is cofigured to allow discussion of viruses whilst still blocking the virus itself. Any coorporation that makes this sort of discussion impossible deserves whatever fate its ignorance leads it to.
I believe that in orbit tethers are a hassle and a half because of the coriolis force -- basically both ends are in different orbits, so they want to orbit at different speeds. This of course is hard to do if they are connected by a tether. Add to this any magnetic drag (which was the point of the excersize) and you get a compicated tangle.
However, the more I think about it, the less I see why centripital force doesn't keep tension.
Actually, you can. I forget exactly how... it had something to do with using most of the sail as a gigantic mirror to focus rays back onto a reversed braking sail. The big reflecting sail is of course cut loose, and accellerates away even faster, which keeps the momentum totals equal.
Of course, going to saturn you could just use the sail as an anchor and fish for asteroids to slow you down, or even bleed off speed in the atomsphere.
A kilowatt aint negilible (is it? I'm talkin' outa ma ass hea) if you're gonna keep going for a couple of years.
Does anyone have an energy rating for one of those nuclear batteries? For how long can it supply a kilowatt continuously?I'm assuming that if you turn the power off, you loose your plasma gas.
Hrm. Why is HTML copyrightable? I mean the tags, the structure -- I completely agree that the information being marked up is copyrightable, but HTML is so limited that there is only one (or a very limited number) way to mark something up in a certain way. This builds on the non-copyrightability (mentioned in an earlier post) of formatting.
"hey! your table code looks just like mine. You must have stolen it!"
For HTML code to be copyrightable, I should hope that it has some novel ideas or non-markup elements used in a non-trivial manner.
So is it true that when you say that HTML is copyrighable, you mean that the design choices (layout) behind the formatting is copyrightable, but in many cases, the actual implementation (making emphasised text using the EM tag) is not?
Some of them seem to be caused by bad contacts. I have an el-cheapo laptop w a 1024x768 screen and one or two "naugty" pixels (prone to go bad, but if you give 'em a tap they shape up).
Actually, it's better to give the screen a tap about 2 inches away from the bad pixel, horizontally or vertically. That has yet to fail to turn the pixel on.
Anyway, after that experience, from now on, I'm buying all my laptops at stores, so I can check out the screen before I take them home.
I will not spend $2K for anything that is defective, regardless of what they conscider acceptable.
Re:Silly mechanical engineers
on
NASA Snake-Bots
·
· Score: 1
The main problem that I see (and it was mentioned earlier) is temperature. You want to use this thing in space, right? So then you have to take a wide temperature swings into account. Esp if this thing is supposed to be in deep space (what is that, past mars orbit?) where there isn't an apprciable ammount of heat from the sun.
So the actuators would have to work at close to absolute zero (hrm, space is at what +0.3 kelvin or was it 3?) so anything based on deformation is going to have a hard time of it. Or liquid lubrication for that matter.
Anyone know how spacesuits work? Internal heatingf for the gaskets around the joints?
Let's try to be just a little consistent here. Under any other circumstances, slashdot posters would deride those TOS as completely ludicrous -- of course you should be allowed to use bots and reverse engineer the system.
But now you want to enforce such restricive terms!? With a fine !?. That must be the most stupid thing I've heard all day.
Slashdot Mirror
Cheap man's 3d:
do what Kanasta said, but flip one image left-right. now affix a mirror inbetween the two, perpendicular to the screen. press your nose agains the mirror, and voila, fusing the images becomes a snap.
The one drawback is that all images appear to be offcenter (behind the reflective side of the mirror, as it were).
But it's cheap.
and it works.
That zippy thing is what I (perhaps mistakenly, please correct) call a fresnel lens below.
Johan
Ok, so replying to myself is kinda gimpy, but I had what I thought was a kick ass idea:
...) in order for it to come out nice, but that still leaves you 330 odd horisontal. good enough for a shootie, Ithink.
given that SOO many of us have notebooks with fairly standard dot-pitch LCDs, wouldn't it be cool if a game manufacturer shipped a fresnel lens cover for your LCD with their 1-st person shoot-em-up?
Every other row or better yet column of pixels would be focussed the in the proper direction, towards each eye. Basically do optically what the MIT people do with fancy schmancy beam splitters.
Ok, so you'd have to keep your head still while playing, and the quality might not be up there, but it would be cool and cheap.
Some problems with the idea might be that getting the lens to line up properly with the pixels would be a hassle. But here's the cool part. LCDs have subpixels, baby! yeah! just shift the image left or right 1 subpixel at a time until the display comes out right.
You might have to sacrifice every other column of pixels too (LxRxLxR
This'd suck processor like you wouldn't believe, but I will bet you anything thing that true 3D at low framerate (and since you have an LCD, you're kinda limited in that regard anyway) over 2D at 100fps. and if you cut resolution, you have less to do anyway.
ok, so I thought that was pretty cool, but I'm biased. I think these ideas are so cool that I'm gonna have to claim copyright to them right now.
Johan
From the MITwebsite (http://www.media.mit.edu/groups/spi/new_AutoM1.ht ml),
it appears that the system actually projects different images into each eye. This one appears to force you to keep your head still, while the MIT one tracks you via IR. Anyway, knowing where your head is, it uses special magic (erm... polarising filter/beamsplitter) to direct the proper image into each eye.
I think they're taking advantage of the narrow viewing angle of the lcd, and tu(r)ning every other row of pixels to each eye.
Does that make sense?
As far as I can read from the article, the machine takes the 3-10 Mb/s input uncompressed rate and makes decompresses it to... 10 Mb/s?
Doesn't that seem a little slow?If you want to drive a megapixel display, that's something like 1/3 fps at 24 bit color.
So if we assume that the New Scientist got the output wrong, and its really outputting 270 Mb/s, that's more reasonable. Then you get a megapixel display at 30 fps if you go for [slightly] lower color resolution (and that's doable).
But then the article goes on to say how he wants to connect 27 of these dohickeys to saturate the SDI pipe, which again brings them down to 10 Mb/s.
Help!
can anyone explain so that the numbers add up?
middle step between SMP and MPP cluster
Almost. SMP and DSM (Distributed Shared Memory, for those not up on acronyms) would be more correct. Some memory is local, some is not. By non-local, that means it could be in another box, or maybe just off another bus.
The important thing is that all memory is accessed as if it were local (just ask for it by address). The hard part is of course getting all processes that want to stomp over the same memory to share it nicely. This involves both page migration (of DSMfame) to processes, and process migration. And of course, at all times maintaining coherency between all copies of a page.
Ultimately, you want all processes that use the same memory to be on the same CPU that memory, or at least near by.
This is hard stuff, as all by scientific applications have hard-to-predict paging behavior. But then scientific apps is what pays the bills for people who need these machines.
bzzt.
A munition is much heavier than the arms that the 2nd ammendment allows. Munitions include shells for heavy artillery and bombs, both of which you most definately are not allowed to own.
As far as Iknow, CD Spins here in boston will let you listen to any CD in the store. Now that they have new CDs, Inever end up going to tower for some reason... strange.
Not to be abtuse here, but I don't get your analogy. Straight line speed is one thing, data throughput is another thing. So I thought maybe you meant that while Sparcs are slower that the PIII, they typically come with better I/O infrastructures. I'm all set to agree with you there. After all IBM sells big iron on this argument alone.
But then you said load, and I'm thinking maybe you meant straight line speed anyway.
So could you clarify?
I read the partitioning to mean virtualizable -- ie able to run operating systems as processes, nested arbitrarily. This is what the s390 guy did who had 40000 odd linux kernels running simultaneously.
Note that vmware does a variant of this, but because they do it on the i386, they have to jump trough hoops. The i386 isn't virtualisable, so they have to scan for instructions that would expose this and rewrite them to a trap to some meta-level code.
It stands to reason that if you have the ability to run two OSs on one CPU, the capability carries over to SMP configurations too.
Erm, posted to soon.
Would anyone care to comment on the ability to write a denial of (printer/cpu) service attack with postscript?
Yes, 'cause having a turing complete language to make pages isn't smart. Or rather, it's too smart.
I'm actually a bit suprised that no-one's gone and written a postscript (font!) based virus or backdoor. Should be possible, no?
Or is postscript evaluated in a really tight sandbox?
He's got a point. Having the display driver which knows about monitor size and resoulution figure out the pixels from the point size is the only logical way to go.
And it really shouldn't be that hard to retrofit. Designing towards the lowest common denominator is ok, as long as the non-lowest aren't unduly harmed, or the lowest aren't to far below the next rung up. In this case, there's little excuse for the lowest rung even existing.
Hrm. I often forget about opera, 'cause last I looked it didn't run on unix and used those damn windows -in- windows. Horrible. Doesn't really matter how nice the inside of the windows are when the rest of the browser doesn't make it to the starting line
I'm always wary of such estimations. How do you go about tallying up damages?
"oh, our department was shut down for 24 hours, so our damages are wages paid for no productivity?"
That seems really close to the lost income arguments. The great thing about lost income is that it is impossible to justify; "the virus stopped me from having time to buy a lottery ticket, and I was going to pick exactly the numbers that won!"
hogwash Itell you.
I should hope that any firewall worth its salt is cofigured to allow discussion of viruses whilst still blocking the virus itself. Any coorporation that makes this sort of discussion impossible deserves whatever fate its ignorance leads it to.
I believe that in orbit tethers are a hassle and a half because of the coriolis force -- basically both ends are in different orbits, so they want to orbit at different speeds. This of course is hard to do if they are connected by a tether. Add to this any magnetic drag (which was the point of the excersize) and you get a compicated tangle.
However, the more I think about it, the less I see why centripital force doesn't keep tension.
Eh?
Actually, you can. I forget exactly how... it had something to do with using most of the sail as a gigantic mirror to focus rays back onto a reversed braking sail. The big reflecting sail is of course cut loose, and accellerates away even faster, which keeps the momentum totals equal.
Of course, going to saturn you could just use the sail as an anchor and fish for asteroids to slow you down, or even bleed off speed in the atomsphere.
Hey!
A kilowatt aint negilible (is it? I'm talkin' outa ma ass hea) if you're gonna keep going for a couple of years.
Does anyone have an energy rating for one of those nuclear batteries? For how long can it supply a kilowatt continuously?I'm assuming that if you turn the power off, you loose your plasma gas.
Hrm. Why is HTML copyrightable? I mean the tags, the structure -- I completely agree that the information being marked up is copyrightable, but HTML is so limited that there is only one (or a very limited number) way to mark something up in a certain way. This builds on the non-copyrightability (mentioned in an earlier post) of formatting.
"hey! your table code looks just like mine. You must have stolen it!"
For HTML code to be copyrightable, I should hope that it has some novel ideas or non-markup elements used in a non-trivial manner.
So is it true that when you say that HTML is copyrighable, you mean that the design choices (layout) behind the formatting is copyrightable, but in many cases, the actual implementation (making emphasised text using the EM tag) is not?
yes. I too was suprised that the poster apparently had failed to read the article.
Now that has to be the height of laziness
must have been freedatabasing when he wrote that.
*groan*
Some of them seem to be caused by bad contacts. I have an el-cheapo laptop w a 1024x768 screen and one or two "naugty" pixels (prone to go bad, but if you give 'em a tap they shape up).
Actually, it's better to give the screen a tap about 2 inches away from the bad pixel, horizontally or vertically. That has yet to fail to turn the pixel on.
Anyway, after that experience, from now on, I'm buying all my laptops at stores, so I can check out the screen before I take them home.
I will not spend $2K for anything that is defective, regardless of what they conscider acceptable.
The main problem that I see (and it was mentioned earlier) is temperature. You want to use this thing in space, right? So then you have to take a wide temperature swings into account. Esp if this thing is supposed to be in deep space (what is that, past mars orbit?) where there isn't an apprciable ammount of heat from the sun.
So the actuators would have to work at close to absolute zero (hrm, space is at what +0.3 kelvin or was it 3?) so anything based on deformation is going to have a hard time of it. Or liquid lubrication for that matter.
Anyone know how spacesuits work? Internal heatingf for the gaskets around the joints?
Let's try to be just a little consistent here. Under any other circumstances, slashdot posters would deride those TOS as completely ludicrous -- of course you should be allowed to use bots and reverse engineer the system.
But now you want to enforce such restricive terms!? With a fine !?. That must be the most stupid thing I've heard all day.
*shakes head, sighs, and turns away*