Thanks for the help and sorry for the ignorance. But, still, it wouldn't be that difficult to have a compiler build floating-point operation code that runs on processors that don't support it. Yes, it would be slower, but it would still work.
It seems that, tho I like what they did, it could have been better.
The problem with the "no software implementation" for sending the images to different monitors? Tho still grounded slightly in hardware, wouldn't it have been easier to just use two cards?
And, tho i understand they can afford it, when I get around to trying this out, I'm making the projectors myself! A nice google search tells you all you need to know.
The first point I made solves this one: they didn't need to do so much image editing. Not to mention, aren't the image files timestamped? That should help with the ordering... And most camera software lets you specify filename prefixes.
Seriously, whats so tough about doing floating point math that these devices can't do it? It isn't difficult to see that floating point operations dont have to be so special.
I dont see why hardware needs to recognize floating points at all. Look at the following example.
The only thing to handle is that the fractional part of the value must be converted from its decimal to the binary by the ratio of 16/10. Multiply the "binimal" by 16 and then divide by 10.
whats so difficult about that? At the least, these float-less processors could include support for floats in their compilers can then just do the conversions in software. Tough, tough, a whole 3 or 4 instructions!
I really gotta say.. this is just stupid. Not to be insulting to them, but what kind of a contest is this? It really doesn't seem all that great or even challenging, from a programming standpoint. These kinds of contests are much better when they're about pure data processing; thats a better challenge. Plus, the entries could be useful. How about a stream compression contest? Challenging and the results could be very useful. Win the contest and license the technology to companies.
There's alot of talk about what is still legal with all the digital copyright laws and what is illegal. However, there are, as always, things forgotten. Laws can be illegal. Like it or not, just because a piece of paper with a few signitures on it is sitting in some book or some case in Washington doesn't mean its a law. If we are really a democracy, then a law will only be such when the majority of those it imposses its rules on want to be impossed in such a way, for whatever reason. And the majority of those people can and will be convinced of their rights in this matter. We have the right to copy, move, time and spaceshift. We have the right to convert and to use the data we posses in the ways we want. And, although I don't see MP3/Og-sharing-type as stealing (because to steal something, you have to take something away from someone and them then not posses that thing anymore), however, there are more reasons to copy! I would love to make a copy of all my DVDs so I can use the copies. I don't want to worry about ruining the films I value so much. (Perhaps, sometimes, a little too much:) ) I hope I'm helpful.
How about Microsoft? Sure, it would take ten million of us several hundred thousand a piece but damn wouldnt that be nice? Even just getting someone onto the board...
I've been pondering the same thing for a while. I propose something is done. I want to start something called "The M Initiative" to move for standard APIs for modularity and app convergence, eventually creating a whole new kind of system.
Anyone interested, please contact me and lets get this off the ground. I've already got most of a very nice modular framework designed, called Mafia. Completely P2P from the ground up, anonomous and also with security and authentication protocols.
I just think this whole idea is great and I would love to get some people in on it. I don't even care if you have no skill, we need people who want this to get together. So, lets get together and collaborate.
Well, in that case, maybe we should also work towards keeping the developement of one part of the kernel (or any project for that matter) to not affect the developement of another.
For example, look at DirectX, although I doubt many heavy Linux Kernel developers know much about DirectX. When a function, method, class, etc is changed in a way to be incompatible with the old version, the new form is added a number allowing it to be accessed seperately from the old version. This could be helpful.
Example:
void Function(int, char);// old version
int Function2(char, int);// new version
Many times, the old version would just take its arguments and pass them to the new one in a different way. Of course, you are saying, bloat bloat bloat... Of course. thats why these backward compatibility libraries would be patches of their own, automatically applied for patches which need them, and ditched when not needed.
Of course, this would take alot of work, a very very very advanced file revision system with probably a good amount of user help (until the system matures more), and lots more headaches i could name.
But, sometimes, isn't a little headache worth the results?
Would it be possible to work with a keyboard that changed as you typed? Constantly evaluting what you press to move keys around to better suit you? If it did this predictably, you could learn to keep up with the movement without looking. Coding? Nice when those {} and switch places.
The merging of Patch A into the stable kernel would not affect Patch B or Patch C because if they are dependant on Patch A, then the developers are already using the most up-to-date kernel patched with Patch A. Of course, something like this would work alot better with a really good file revision system. Not CVS or any of those, but something where each patch can have its own repository, easily updating from the current stable kernel. developers of Patch B would update from two repositories, stable and Patch A. Patch C devs would also update from B. When A is merged, the code the other developers work from is unchanged.
Although I do see the need for keeping old kernels maintained (I even have some old systems lying around I want to install 2.0.x on), I see this as one of the problems with Linux.
Why do we have to have so many kernels maintained at the same time? Even just the current "stable" and "unstable" release system is a little strange to me. I mean, why spread the work among two kernels when we could be doing twice the work on just one?
I would propose things differently. A single kernel, the latest release, is the only one maintained (officially. anyone can maintain old kernels if they wish). the patches would, however, be marked stable and unstable. Test patches and work on them until they are stable enough for what would be a stable release, then merge them permanently into the main source. Until the patch is stable, it remains just a patch, being tested and worked upon.
I admit, I'm no kernel hacker (yet) but I do think this would be a much better solution. Linux would advance much faster with all the effort focusing on one kernel, no more.
for my mom to call and ask how to run that funny attatchment she just got from her sister in email. or how to install that "Friends" screen saver she saw on the website.
Linux should be a platform where developers write apps for themselves
Yes, exactly. Whats wrong with that? Linux is a UNIX-like system, and both are tailored for developers. im not saying, however, that work shouldnt be done to do a mirror system based on linux but for end users. i would be slower, but be much easier. things like full compatibility between kernels and simple installs. and, installing can just be one script they double click thay compiles the app.
i agree with alot of that, but there's a problem...
so what if linux sucks for desktop? cant ppl realize it isnt meant for that and shouldnt be. linux is and should be an OS in development for developers. thats what it is. i dont care if its easy for joe user, he can use xp if he wants and he should.
some mistakes: source code. it is good for more than modifying programs. compile code on your machine and get a more optimized product. look at sourceror. and there are windows apps that come as sourcecode. rare, but they exist.
and yeah, X does suck. it should be replaced. firstly we need a system doing nothing but giving access to video. no windows or anything. just video. build a seperate window system over that. toss in a network layer (default should be local, so its fast!) and there you go.
packages. we try to standardize, but there are always choices. we dont all want the same thing. solution? a system that allows all package systems to be used the same; rpm, deb, and tgz.
good luck in xp. but, id recomend win98. ive yet to see a system run xp faster. and joe user doesnt need to be super. single use is best for him.
Almost forgot
Taco, why even post this? It's not anything really important.. at all even. Either its a slow day, you made a mistake, or/. crowd is geekier than I thought.
Lets post a story about Spielbergs favorite porn sites next. It would just as relevent to us as this piece.
I've actually lost some respect for Spielberg because of this. Even asking... I mean, come on! I don't care if he is a hot-shot director, it seems rude. What? Doesn't he have enough movies of his own?
The big problem, I believe, with all the talk about video games and their many aspects, is what a video game is. As gamers, many of us either know that the selection is extremely varied, or think that the only game is quake3/gta3/starcraft/etc.
One comment said that the only skill video games give us are the ability to rapidly press the buttons. Anyone play Silent Hill 2 lately? In a fight, sure I'm firing away every last shot I have in a second if it takes that; but, most of the time, I'm calming going down the halls, checking doors, re-reading riddles, and trying to figure out what the hell is going on.
Of course, I used to play Quake all the time. Now that was a button-pusher, but not much of a thinker. Get the routine down and there isn't much too it. Same with games like Soul Calibur on Dreamcast, which I was addicted too. Hell, I had that down so much I could beat anyone with the controler under my toes, not fingers, while eating a pizza and talking. No thought involved, lots of buttons being destroyed.
What's my point? There is a balence in video games we need to consider with this sort of discussion and research. Thought versus habit. How much of the gameplay can be trained (Quake) and how much is a surprise, always requiring and even tuning thought processes (Silent Hill 2).
Think about it. And if you have trouble thinking, play some video games and you're brain will work a little better.
I just do in XHTML, which forces me to follow standards. Works like a charm and if your browser doesn't read the XHTML right, thats your browser's fault and I am not doing a damn thing to fix it.
Can't stand how people like him, who should be so thankful for the freedom of thought and opinion, could shun so many others for having that same freedom.
Slashdot Mirror
Thanks for the help and sorry for the ignorance. But, still, it wouldn't be that difficult to have a compiler build floating-point operation code that runs on processors that don't support it. Yes, it would be slower, but it would still work.
It seems that, tho I like what they did, it could have been better.
The problem with the "no software implementation" for sending the images to different monitors? Tho still grounded slightly in hardware, wouldn't it have been easier to just use two cards?
And, tho i understand they can afford it, when I get around to trying this out, I'm making the projectors myself! A nice google search tells you all you need to know.
The first point I made solves this one: they didn't need to do so much image editing. Not to mention, aren't the image files timestamped? That should help with the ordering... And most camera software lets you specify filename prefixes.
Seriously, whats so tough about doing floating point math that these devices can't do it? It isn't difficult to see that floating point operations dont have to be so special.
I dont see why hardware needs to recognize floating points at all. Look at the following example.
4.5 + 0.5
0100.1000
+ 0000.1000
-----------
0101.0000
The only thing to handle is that the fractional part of the value must be converted from its decimal to the binary by the ratio of 16/10. Multiply the "binimal" by 16 and then divide by 10.
whats so difficult about that? At the least, these float-less processors could include support for floats in their compilers can then just do the conversions in software. Tough, tough, a whole 3 or 4 instructions!
I really gotta say.. this is just stupid. Not to be insulting to them, but what kind of a contest is this? It really doesn't seem all that great or even challenging, from a programming standpoint. These kinds of contests are much better when they're about pure data processing; thats a better challenge. Plus, the entries could be useful. How about a stream compression contest? Challenging and the results could be very useful. Win the contest and license the technology to companies.
There's alot of talk about what is still legal with all the digital copyright laws and what is illegal. However, there are, as always, things forgotten. Laws can be illegal. Like it or not, just because a piece of paper with a few signitures on it is sitting in some book or some case in Washington doesn't mean its a law. If we are really a democracy, then a law will only be such when the majority of those it imposses its rules on want to be impossed in such a way, for whatever reason. And the majority of those people can and will be convinced of their rights in this matter. We have the right to copy, move, time and spaceshift. We have the right to convert and to use the data we posses in the ways we want. And, although I don't see MP3/Og-sharing-type as stealing (because to steal something, you have to take something away from someone and them then not posses that thing anymore), however, there are more reasons to copy! I would love to make a copy of all my DVDs so I can use the copies. :) )
I don't want to worry about ruining the films I value so much. (Perhaps, sometimes, a little too much
I hope I'm helpful.
this is a repost. i read this before. same site i think. sheesh...
How about Microsoft? Sure, it would take ten million of us several hundred thousand a piece but damn wouldnt that be nice? Even just getting someone onto the board...
a good thing too that it isnt linux. this poster said they run on a Mac.
and my friend with broad band -just- finished downloading the previous stable release for me! 6 CDs... (3 binary, 3 source). release often, eh?
I've been pondering the same thing for a while. I propose something is done. I want to start something called "The M Initiative" to move for standard APIs for modularity and app convergence, eventually creating a whole new kind of system.
Anyone interested, please contact me and lets get this off the ground. I've already got most of a very nice modular framework designed, called Mafia. Completely P2P from the ground up, anonomous and also with security and authentication protocols.
I just think this whole idea is great and I would love to get some people in on it. I don't even care if you have no skill, we need people who want this to get together. So, lets get together and collaborate.
Contact me.
For example, look at DirectX, although I doubt many heavy Linux Kernel developers know much about DirectX. When a function, method, class, etc is changed in a way to be incompatible with the old version, the new form is added a number allowing it to be accessed seperately from the old version. This could be helpful.
Many times, the old version would just take its arguments and pass them to the new one in a different way. Of course, you are saying, bloat bloat bloat... Of course. thats why these backward compatibility libraries would be patches of their own, automatically applied for patches which need them, and ditched when not needed.
Of course, this would take alot of work, a very very very advanced file revision system with probably a good amount of user help (until the system matures more), and lots more headaches i could name.
But, sometimes, isn't a little headache worth the results?
Would it be possible to work with a keyboard that changed as you typed? Constantly evaluting what you press to move keys around to better suit you? If it did this predictably, you could learn to keep up with the movement without looking. Coding? Nice when those {} and switch places.
The merging of Patch A into the stable kernel would not affect Patch B or Patch C because if they are dependant on Patch A, then the developers are already using the most up-to-date kernel patched with Patch A. Of course, something like this would work alot better with a really good file revision system. Not CVS or any of those, but something where each patch can have its own repository, easily updating from the current stable kernel. developers of Patch B would update from two repositories, stable and Patch A. Patch C devs would also update from B. When A is merged, the code the other developers work from is unchanged.
Why do we have to have so many kernels maintained at the same time? Even just the current "stable" and "unstable" release system is a little strange to me. I mean, why spread the work among two kernels when we could be doing twice the work on just one?
I would propose things differently. A single kernel, the latest release, is the only one maintained (officially. anyone can maintain old kernels if they wish). the patches would, however, be marked stable and unstable. Test patches and work on them until they are stable enough for what would be a stable release, then merge them permanently into the main source. Until the patch is stable, it remains just a patch, being tested and worked upon.
I admit, I'm no kernel hacker (yet) but I do think this would be a much better solution. Linux would advance much faster with all the effort focusing on one kernel, no more.
for my mom to call and ask how to run that funny attatchment she just got from her sister in email. or how to install that "Friends" screen saver she saw on the website.
Yes, exactly. Whats wrong with that? Linux is a UNIX-like system, and both are tailored for developers. im not saying, however, that work shouldnt be done to do a mirror system based on linux but for end users. i would be slower, but be much easier. things like full compatibility between kernels and simple installs. and, installing can just be one script they double click thay compiles the app.
i agree with alot of that, but there's a problem...
so what if linux sucks for desktop? cant ppl realize it isnt meant for that and shouldnt be. linux is and should be an OS in development for developers. thats what it is. i dont care if its easy for joe user, he can use xp if he wants and he should.
some mistakes: source code. it is good for more than modifying programs. compile code on your machine and get a more optimized product. look at sourceror. and there are windows apps that come as sourcecode. rare, but they exist.
and yeah, X does suck. it should be replaced. firstly we need a system doing nothing but giving access to video. no windows or anything. just video. build a seperate window system over that. toss in a network layer (default should be local, so its fast!) and there you go.
packages. we try to standardize, but there are always choices. we dont all want the same thing. solution? a system that allows all package systems to be used the same; rpm, deb, and tgz.
good luck in xp. but, id recomend win98. ive yet to see a system run xp faster. and joe user doesnt need to be super. single use is best for him.
if we need it enough, someone will take over and keep the thing going. thats the beauty of open source. this is a positive point, not negative.
Almost forgot /. crowd is geekier than I thought.
Taco, why even post this? It's not anything really important.. at all even. Either its a slow day, you made a mistake, or
Lets post a story about Spielbergs favorite porn sites next. It would just as relevent to us as this piece.
I've actually lost some respect for Spielberg because of this. Even asking... I mean, come on! I don't care if he is a hot-shot director, it seems rude. What? Doesn't he have enough movies of his own?
One comment said that the only skill video games give us are the ability to rapidly press the buttons. Anyone play Silent Hill 2 lately? In a fight, sure I'm firing away every last shot I have in a second if it takes that; but, most of the time, I'm calming going down the halls, checking doors, re-reading riddles, and trying to figure out what the hell is going on.
Of course, I used to play Quake all the time. Now that was a button-pusher, but not much of a thinker. Get the routine down and there isn't much too it. Same with games like Soul Calibur on Dreamcast, which I was addicted too. Hell, I had that down so much I could beat anyone with the controler under my toes, not fingers, while eating a pizza and talking. No thought involved, lots of buttons being destroyed.
What's my point? There is a balence in video games we need to consider with this sort of discussion and research. Thought versus habit. How much of the gameplay can be trained (Quake) and how much is a surprise, always requiring and even tuning thought processes (Silent Hill 2).
Think about it. And if you have trouble thinking, play some video games and you're brain will work a little better.
...I really don't think this idea will fly with me.
I just do in XHTML, which forces me to follow standards. Works like a charm and if your browser doesn't read the XHTML right, thats your browser's fault and I am not doing a damn thing to fix it.
Can't stand how people like him, who should be so thankful for the freedom of thought and opinion, could shun so many others for having that same freedom.
Quantum-hologram scope (Read something about them somewhere) to watch your favorite internal organ.