I thought that 'glass flow' was a myth and that the reason that old windows are thicker at the base is a result of imprecise manufacturing processes ~400 years ago. If you made a sheet of glass that was thicker on one side, which side would you put at the bottom of the window?
Anyone know about glass flow that can confirm/deny this?
HH
Yellow tigers crouched in jungles in her dark eyes.
You're right - it's not a big deal. At least not for the average Linux user. But for companies who produce computer graphics for TV and/or film, the prospect of low-cost render farms is probably the greatest deal they've ever had. So don't knock it.
HH
Yellow tigers crouched in jungles in her dark eyes.
AMD have dropped the 500Mhz Athlon already. Will they still make 600Mhz chips in 18 months? Wonder what AMD's Athlon product line will look like then? Possibly 2Ghz, 1.8Ghz, 1.5GHz, 1.2Ghz, 600Mhz?????
HH
Yellow tigers crouched in jungles in her dark eyes.
But you could also argue that it's a PC masquerading as a console. In 18 months time, if Moore's Law holds, PC's will be up to 2GHz. Cutting edge PC games will be starting to take advantage of this power and may well run slowly on a 600Mhz machine. Graphics chipsets will also have moved forwards significantly. I suspect that the Emotion Engine performs significantly better than it's 300Mhz would suggest. MHz, as you probably know, is a useless indicator of the relative performance of different processor architectures.
HH
Yellow tigers crouched in jungles in her dark eyes.
Hopefully companies like Creative Labs will set a good example for others to follow. Creative released a closed-source driver for the soundblaster live. It worked (just) but MP3 playback was noticably worse than in Windows. They got a lot of flack from the Linux community over their refusal to open-source the driver and eventually gave in and released the source under the GPL. The soundblaster live driver suddenly improved beyond recognitions (thanks Alan & everyone else) and now plays MP3's better than the Windows driver on my PC.
If companies realise that there are real benefits to open-sourcing drivers, then they might just do so.
HH
Yellow tigers crouched in jungles in her dark eyes.
Interesting questions. Firstly a disclaimer. I know bugger-all about digital watermarking, but I've got some ideas on how to implement it.
I've no idea about patents. Hopefully someone else will be able to answer that one.
Do they involve scary maths? I don't think so, unless you consider error-checking/correction algorithms to be scary.
My guess is that these algorithms work by using the least significant bit of the RGB values of the image to embed some kind of digital signature (what that signature contains is up to you). This will add a small amount of 'noise' to the image, but my guess is that this will be imperceptible, since the human eye/brain is very forgiving of noise (that is why floyd-steinberg error diffusion dithering works so well).
For a watermark, you could almost certainly get away with just using the LSB of one byte of the RGB value. I'd recommend using Blue, since it's the darkest of the three primary colours.
So far, that would be very trivial to implement. Your biggest problem would be preserving the watermark when lossy compression algorithms, such as JPEG, are used. This may (I'm guessing here) be solvable by repeating the watermark thoughout the image and using an error-correction algorithm to reconstruct the watermark. This should be possible, but you may need to experiment and possibly devise your own error-correction algorithm.
I've no idea if anyone has implemented an open-source digital watermark. I thought that it would make an interesting project, but got no further than thinking about it.
If you're interested in writing one and want to discuss it further with me, my email address is wyrd(a)dial.pipex.com (remember to replace (a) with @)
HH
Yellow tigers crouched in jungles in her dark eyes.
generate the appropriate texture volume and then put that in the image cache with the standard 2D versions
As I mentioned elsewhere in this discussion, precomputed 3D texture maps would take up vast amounts of memory on your video card. IMHO it would probably be better to let the CPU compute the procedural textures and transfer them to the card using AGP.
Better still, provide a texture compiler that produces bytecode that can be executed directly by the card. Now that would be cool. Procedural displacement mapping (like RenderMan uses) would be ultra-cool.
So the first 3D card that can execute RenderMan shader bytecode will get my money:-)
HH
Yellow tigers crouched in jungles in her dark eyes.
I think that the HP890c is one of those "winprinters." OF course I could be talking out of my ass. If it is I'm pleasantly surprised that it works at all.
Well, I won't accuse you of talking out of your ass, but the 890c uses PCL3+ and TIFF run-length encoding for compression. I spent a few hours of my precious free time last weekened examining hex dumps of the output from the Windows and Linux drivers (Time I'd reserved for filling in my tax returns!) The Windows driver seems to use some undocumented PCL3 escape sequences, but I think that I can improve the Linux driver to give similar performance, if I can find a day or two to dedicate to it.
HH
Yellow tigers crouched in jungles in her dark eyes.
How about Phong instead of Gourard shading? Fast Phong algorithms for hardware implementation have been about for years. They're still more computationally intensive than Gourard but remove the need for specular texture maps and reduce mach-banding artifacts.
Real-time radiosity? Not for a very long time, methinks. Radiosity is usually pre-computed. I remember reading one of John Carmack's.plans where he said that ID had bought a supercomputer (a Sequent?) to perform the Quake radiosity calculations.
HH
Yellow tigers crouched in jungles in her dark eyes.
Nice to see support for 3D textures. These are very cool. The article says:
Traditionally, polygons are used to represent 3D objects. However, with 3D textures, volumes of texels (textured pixels) may also be used. In a 2D texture map (the kind that we see "glued" to a wall for instance) indexing occurs via two texture coordinates, whereas in a 3D texture, there are three coordinates.
One good example of 3D texture use would be that of a marble cube. If the corner of the cube were to be chipped off, any veins running through the marble would already be defined and visible without any additional textures being generated.
This means that you could chop a block of wood up, and have the wood grain on the cut surfaces rendered correcly. However, the article then goes on to say:
Unfortunately, we feel 3D textures will have to be used incredibly sparingly because in order to implement the marble cube example explained above, an artist would have to draw the entire 3D surface (including the veins inside the cube which may never be seen!).
This is incorrect. How can an artist draw the inside of a solid cube of marble or wood? I've never heard of a 3D texture being created in this way. They are normally generated procedurally, where you have a function that mathematically calculates the texture colour given the x,y,z coordinates within the texture. This does mean that you can't store 3D textures on the card, unless you pre-calculate an array of texels, but this would require vast amounts of texture memory on the card.
HH
Yellow tigers crouched in jungles in her dark eyes.
One of the best things about Windows, in fact, one of the only truly wonderful things about Windows is the printer support. It's very easy to install and configure printers. When I print, to my HP890c I can hit a button in a dialog box that selects 'econofast' mode. This prints quickly (12 secs for a test page), with pretty good (300x300dpi) quality.
The same test page printed from Red Hat 6.1 takes almost THREE AND A HALF MINUTES to print. This means that a 10 page document would print in two minutes from Windows and over half an hour in Linux. It was quicker to reboot into Windows, print and then boot Linux again, than to print in Linux. This sucks hard.
I finally managed to work out how to get Linux to print quicker, but it took a very long time and I had to download, compile and read the source to Ghostscript to do it. Your average user couldn't/wouldn't do this.
I'd love to take time off work to help out with one of the Linux printing projects, but am far too busy at the moment trying to get an interactive TV project on air (I'll try to fix some of the bugs in the print driver though). Maybe in the summer...
But until Linux printing is as good as Windows, it really doesn't have a chance on the desktop. </rant>
HH
Yellow tigers crouched in jungles in her dark eyes.
You really want one with a digital (DVI) connector. However, these (& cards that use them) are not very common yet. Your best bets are to either stick with your 21" monitor or buy a really cheap analog flat panel and wait for the better DVI ones to become widely available. I went for the latter route. The display is OK but not amazing, but it has a silver surround and looks very cool.
Also try the monitor before you buy - many (most?) of them have some dead pixels. Bright dead pixels are much more noticable than dark ones.
HH
Yellow tigers crouched in jungles in her dark eyes.
There's a lot of extremely academic stuff that's required to produce games. Most games use 3D graphics. That requires some pretty advanced mathematics - I know, I've got a masters degree in it. How about user interface design? A good game with a crap user interface could be an unplayable flop. Colour theory - that's another important aspect of CG. How about programming? Game programmers have to be good to extract the last bit of performance from the hardware. This course seems to be designed to prepare people for a career, unlike your example of folkdancing (not that there's anything actually wrong with folkdancing), and it's a damn sight more academic than sociology IMHO.
HH
Yellow tigers crouched in jungles in her dark eyes.
Alias/Wavefront is very high-end 3D-graphics modelling/rendering software. It's used for stuff like special effects for films. The article is incorrect when it says Alias is a 'game maker'. Licences for their software are extremely expensive, it's not something you'd run on your PC at home, it's more the kind of software that's people buy SGI workstations for.
Wish I'd had it when I was studying computer graphics at university.
HH
Yellow tigers crouched in jungles in her dark eyes.
"What do all the big sites have in common? Minimalist design."
I've just looked at Nielson's page on the top 10 mistakes in web design and one thing strikes me as slightly strange. His list of 'popular websites' (as opposed to 'prominent websites') is AOL.com, Yahoo, MSN, Go, GeoCities, Netscape, Excite, Microsoft, Lycos, and Angelfire. He says this about the 'popular websites' It is not an accident that the sites with the most traffic have an uncommonly low rate of violations of the top ten mistakes of Web design. On the contrary, it is because these sites are easy to use that they get so much traffic.
Is it? Netscape's site is very hard to navigate. It gets a lot of hits because it's the default home page for Netscape's browsers. AOL, Microsoft and MSN probably also get large numbers of hits for the same reason. Microsoft must get lots of hits because it's Microsoft, not because it's site is minimalist. Most of the other popular sites are search engines/portals which again are hit because of their function, rather than their design. Having said that, I love Google's minimalist design.
He also says "I know that Slashdot readers don't want to hear this, but the very first question is whether it is even possible to create a truly good user experience on top of Linux. Many other companies have tried to make Unix easy to use and many very talented designers have worked hard on these projects for several years without very good results. The only data points we have say that it can't be done." to which I say "MacOS X"
HH Yellow tigers crouched in jungles in her dark eyes.
He's gradually turning up the heat on Amazon. First it was a private exchange of email. Now it's an open letter and a petition. If that fails, maybe he will stop supplying Amazon. I doubt that it would hurt his business very much, as people buy O'Reilly because of their reputation and if Amazon don't stock them, they'll go elsewhere.
Linux can't just drop USB. If I go to the PC World next to my office I can buy USB keyboards, mices, scanners, printers, mp3 players, digital cameras etc. Most (if not all) new PC's and motherboards come with built-in USB. Linux needs to support USB. Therefore the developers need the specs. Whilst the specs should be freely available, $2,500 is peanuts to companies like IBM, SGI, Corel, VA, Red Hat etc.
HH
Yellow tigers crouched in jungles in her dark eyes.
Don't know why you can't read the article (slashdotted maybe?) but youre general impressions are completely wrong. With the S/390 mainframe, multiple operating systems can be run simultaneously. You can even debug unstable, crashing OS's without taking the machine down. Very cool.
And are IBM 'brain dead' for porting Linux to their own mainframes? I don't think so.
I can't believe how this post has been moderated up to 4 (at time of writing) when the poster obviously hasn't even read the article. Read it. It's very informative and interesting, especially if you don't know what modern mainframes are capable of.
Can you blame them? I turn that fscking paperclip off, but if I'm typing a long technical document and start a paragraph with the word 'To', the paperclip jumps up and says "It looks like you're writing a letter". Even if I'm 20 pages into the document. Artificial Stupidity at its worst. Microsoft should either fix this 'feature' or remove it.
The development process descibed in the article really does account for the freeping creaturism found in MS software.
So Microsoft is one big, happy family;-) I think that the problem is maybe that there are a few people within Microsoft who are far too aggressive towards competitors. As the article says, the majority of M$ employees are probably quite normal, likeable people who are not bent on total world domination at the expense of everybody else.
I was a bit suprised that there was no mention of Linux in this article. I would assume that it would be a topic of conversation amongst Microsofties. From the tone of the article, I doubt very much that James Fallows will donate his unused ideas to the open-source office suites, though it would be very noble of him if he did.
Slashdot Mirror
Microsoft is developing a new version of Windows, specifically for the X-Box. It's going to be called X-Windows!
HH
Yellow tigers crouched in jungles in her dark eyes.
I thought that 'glass flow' was a myth and that the reason that old windows are thicker at the base is a result of imprecise manufacturing processes ~400 years ago. If you made a sheet of glass that was thicker on one side, which side would you put at the bottom of the window?
Anyone know about glass flow that can confirm/deny this?
HH
Yellow tigers crouched in jungles in her dark eyes.
You're right - it's not a big deal. At least not for the average Linux user. But for companies who produce computer graphics for TV and/or film, the prospect of low-cost render farms is probably the greatest deal they've ever had. So don't knock it.
HH
Yellow tigers crouched in jungles in her dark eyes.
AMD have dropped the 500Mhz Athlon already. Will they still make 600Mhz chips in 18 months? Wonder what AMD's Athlon product line will look like then? Possibly 2Ghz, 1.8Ghz, 1.5GHz, 1.2Ghz, 600Mhz?????
HH
Yellow tigers crouched in jungles in her dark eyes.
But you could also argue that it's a PC masquerading as a console. In 18 months time, if Moore's Law holds, PC's will be up to 2GHz. Cutting edge PC games will be starting to take advantage of this power and may well run slowly on a 600Mhz machine. Graphics chipsets will also have moved forwards significantly. I suspect that the Emotion Engine performs significantly better than it's 300Mhz would suggest. MHz, as you probably know, is a useless indicator of the relative performance of different processor architectures.
HH
Yellow tigers crouched in jungles in her dark eyes.
Hopefully companies like Creative Labs will set a good example for others to follow. Creative released a closed-source driver for the soundblaster live. It worked (just) but MP3 playback was noticably worse than in Windows. They got a lot of flack from the Linux community over their refusal to open-source the driver and eventually gave in and released the source under the GPL. The soundblaster live driver suddenly improved beyond recognitions (thanks Alan & everyone else) and now plays MP3's better than the Windows driver on my PC.
If companies realise that there are real benefits to open-sourcing drivers, then they might just do so.
HH
Yellow tigers crouched in jungles in her dark eyes.
Are you sure about a "Christmas 2001" release? Don't you mean Christmas 2000? A 600MHz box will be seriously out of date by December 2001.
HH
Yellow tigers crouched in jungles in her dark eyes.
Interesting questions. Firstly a disclaimer. I know bugger-all about digital watermarking, but I've got some ideas on how to implement it.
I've no idea about patents. Hopefully someone else will be able to answer that one.
Do they involve scary maths? I don't think so, unless you consider error-checking/correction algorithms to be scary.
My guess is that these algorithms work by using the least significant bit of the RGB values of the image to embed some kind of digital signature (what that signature contains is up to you). This will add a small amount of 'noise' to the image, but my guess is that this will be imperceptible, since the human eye/brain is very forgiving of noise (that is why floyd-steinberg error diffusion dithering works so well).
For a watermark, you could almost certainly get away with just using the LSB of one byte of the RGB value. I'd recommend using Blue, since it's the darkest of the three primary colours.
So far, that would be very trivial to implement. Your biggest problem would be preserving the watermark when lossy compression algorithms, such as JPEG, are used. This may (I'm guessing here) be solvable by repeating the watermark thoughout the image and using an error-correction algorithm to reconstruct the watermark. This should be possible, but you may need to experiment and possibly devise your own error-correction algorithm.
I've no idea if anyone has implemented an open-source digital watermark. I thought that it would make an interesting project, but got no further than thinking about it.
If you're interested in writing one and want to discuss it further with me, my email address is wyrd(a)dial.pipex.com (remember to replace (a) with @)
HH
Yellow tigers crouched in jungles in her dark eyes.
generate the appropriate texture volume and then put that in the image cache with the standard 2D versions
:-)
As I mentioned elsewhere in this discussion, precomputed 3D texture maps would take up vast amounts of memory on your video card. IMHO it would probably be better to let the CPU compute the procedural textures and transfer them to the card using AGP.
Better still, provide a texture compiler that produces bytecode that can be executed directly by the card. Now that would be cool. Procedural displacement mapping (like RenderMan uses) would be ultra-cool.
So the first 3D card that can execute RenderMan shader bytecode will get my money
HH
Yellow tigers crouched in jungles in her dark eyes.
I think that the HP890c is one of those "winprinters." OF course I could be talking out of my ass. If it is I'm pleasantly surprised that it works at all.
Well, I won't accuse you of talking out of your ass, but the 890c uses PCL3+ and TIFF run-length encoding for compression. I spent a few hours of my precious free time last weekened examining hex dumps of the output from the Windows and Linux drivers (Time I'd reserved for filling in my tax returns!) The Windows driver seems to use some undocumented PCL3 escape sequences, but I think that I can improve the Linux driver to give similar performance, if I can find a day or two to dedicate to it.
HH
Yellow tigers crouched in jungles in her dark eyes.
There must be others
.plans where he said that ID had bought a supercomputer (a Sequent?) to perform the Quake radiosity calculations.
How about Phong instead of Gourard shading? Fast Phong algorithms for hardware implementation have been about for years. They're still more computationally intensive than Gourard but remove the need for specular texture maps and reduce mach-banding artifacts.
Real-time radiosity? Not for a very long time, methinks. Radiosity is usually pre-computed. I remember reading one of John Carmack's
HH
Yellow tigers crouched in jungles in her dark eyes.
Nice to see support for 3D textures. These are very cool. The article says:
Traditionally, polygons are used to represent 3D objects. However, with 3D textures, volumes of texels (textured pixels) may also be used. In a 2D texture map (the kind that we see "glued" to a wall for instance) indexing occurs via two texture coordinates, whereas in a 3D texture, there are three coordinates.
One good example of 3D texture use would be that of a marble cube. If the corner of the cube were to be chipped off, any veins running through the marble would already be defined and visible without any additional textures being generated.
This means that you could chop a block of wood up, and have the wood grain on the cut surfaces rendered correcly. However, the article then goes on to say:
Unfortunately, we feel 3D textures will have to be used incredibly sparingly because in order to implement the marble cube example explained above, an artist would have to draw the entire 3D surface (including the veins inside the cube which may never be seen!).
This is incorrect. How can an artist draw the inside of a solid cube of marble or wood? I've never heard of a 3D texture being created in this way. They are normally generated procedurally, where you have a function that mathematically calculates the texture colour given the x,y,z coordinates within the texture. This does mean that you can't store 3D textures on the card, unless you pre-calculate an array of texels, but this would require vast amounts of texture memory on the card.
HH
Yellow tigers crouched in jungles in her dark eyes.
One of the best things about Windows, in fact, one of the only truly wonderful things about Windows is the printer support. It's very easy to install and configure printers. When I print, to my HP890c I can hit a button in a dialog box that selects 'econofast' mode. This prints quickly (12 secs for a test page), with pretty good (300x300dpi) quality.
The same test page printed from Red Hat 6.1 takes almost THREE AND A HALF MINUTES to print. This means that a 10 page document would print in two minutes from Windows and over half an hour in Linux. It was quicker to reboot into Windows, print and then boot Linux again, than to print in Linux. This sucks hard.
I finally managed to work out how to get Linux to print quicker, but it took a very long time and I had to download, compile and read the source to Ghostscript to do it. Your average user couldn't/wouldn't do this.
I'd love to take time off work to help out with one of the Linux printing projects, but am far too busy at the moment trying to get an interactive TV project on air (I'll try to fix some of the bugs in the print driver though). Maybe in the summer...
But until Linux printing is as good as Windows, it really doesn't have a chance on the desktop.
</rant>
HH
Yellow tigers crouched in jungles in her dark eyes.
You really want one with a digital (DVI) connector. However, these (& cards that use them) are not very common yet. Your best bets are to either stick with your 21" monitor or buy a really cheap analog flat panel and wait for the better DVI ones to become widely available. I went for the latter route. The display is OK but not amazing, but it has a silver surround and looks very cool.
Also try the monitor before you buy - many (most?) of them have some dead pixels. Bright dead pixels are much more noticable than dark ones.
HH
Yellow tigers crouched in jungles in her dark eyes.
There's a lot of extremely academic stuff that's required to produce games. Most games use 3D graphics. That requires some pretty advanced mathematics - I know, I've got a masters degree in it. How about user interface design? A good game with a crap user interface could be an unplayable flop. Colour theory - that's another important aspect of CG. How about programming? Game programmers have to be good to extract the last bit of performance from the hardware. This course seems to be designed to prepare people for a career, unlike your example of folkdancing (not that there's anything actually wrong with folkdancing), and it's a damn sight more academic than sociology IMHO.
HH
Yellow tigers crouched in jungles in her dark eyes.
Alias/Wavefront is very high-end 3D-graphics modelling/rendering software. It's used for stuff like special effects for films. The article is incorrect when it says Alias is a 'game maker'. Licences for their software are extremely expensive, it's not something you'd run on your PC at home, it's more the kind of software that's people buy SGI workstations for.
Wish I'd had it when I was studying computer graphics at university.
HH
Yellow tigers crouched in jungles in her dark eyes.
"What do all the big sites have in common? Minimalist design."
I've just looked at Nielson's page on the top 10 mistakes in web design and one thing strikes me as slightly strange. His list of 'popular websites' (as opposed to 'prominent websites') is AOL.com, Yahoo, MSN, Go, GeoCities, Netscape, Excite, Microsoft, Lycos, and Angelfire. He says this about the 'popular websites'
It is not an accident that the sites with the most traffic have an uncommonly low rate of violations of the top ten mistakes of Web design. On the contrary, it is because these sites are easy to use that they get so much traffic.
Is it? Netscape's site is very hard to navigate. It gets a lot of hits because it's the default home page for Netscape's browsers. AOL, Microsoft and MSN probably also get large numbers of hits for the same reason. Microsoft must get lots of hits because it's Microsoft, not because it's site is minimalist. Most of the other popular sites are search engines/portals which again are hit because of their function, rather than their design. Having said that, I love Google's minimalist design.
He also says "I know that Slashdot readers don't want to hear this, but the very first question is whether it is even possible to create a truly good user experience on top of Linux. Many other companies have tried to make Unix easy to use and many very talented designers have worked hard on these projects for several years without very good results. The only data points we have say that it can't be done."
to which I say "MacOS X"
HH
Yellow tigers crouched in jungles in her dark eyes.
He's gradually turning up the heat on Amazon. First it was a private exchange of email. Now it's an open letter and a petition. If that fails, maybe he will stop supplying Amazon. I doubt that it would hurt his business very much, as people buy O'Reilly because of their reputation and if Amazon don't stock them, they'll go elsewhere.
And please sign the petition
HH
Yellow tigers crouched in jungles in her dark eyes.
Linux can't just drop USB. If I go to the PC World next to my office I can buy USB keyboards, mices, scanners, printers, mp3 players, digital cameras etc. Most (if not all) new PC's and motherboards come with built-in USB. Linux needs to support USB. Therefore the developers need the specs. Whilst the specs should be freely available, $2,500 is peanuts to companies like IBM, SGI, Corel, VA, Red Hat etc.
HH
Yellow tigers crouched in jungles in her dark eyes.
Don't know why you can't read the article (slashdotted maybe?) but youre general impressions are completely wrong. With the S/390 mainframe, multiple operating systems can be run simultaneously. You can even debug unstable, crashing OS's without taking the machine down. Very cool.
And are IBM 'brain dead' for porting Linux to their own mainframes? I don't think so.
HH
I can't believe how this post has been moderated up to 4 (at time of writing) when the poster obviously hasn't even read the article. Read it. It's very informative and interesting, especially if you don't know what modern mainframes are capable of.
HH
Yeah, alt.slashdot (altdot?) - like we don't have enough trolls & flamers already.
Seriously, if there is ever an nntp gateway (and I can't see any valid reasons for needing one), it should be read-only.
15,000 rpm. That's 500 rpm SLOWER than the redline of a Yamaha R6 motorbike.
HH
Can you blame them? I turn that fscking paperclip off, but if I'm typing a long technical document and start a paragraph with the word 'To', the paperclip jumps up and says "It looks like you're writing a letter". Even if I'm 20 pages into the document. Artificial Stupidity at its worst. Microsoft should either fix this 'feature' or remove it.
The development process descibed in the article really does account for the freeping creaturism found in MS software.
HH
So Microsoft is one big, happy family ;-)
I think that the problem is maybe that there are a few people within Microsoft who are far too aggressive towards competitors. As the article says, the majority of M$ employees are probably quite normal, likeable people who are not bent on total world domination at the expense of everybody else.
I was a bit suprised that there was no mention of Linux in this article. I would assume that it would be a topic of conversation amongst Microsofties. From the tone of the article, I doubt very much that James Fallows will donate his unused ideas to the open-source office suites, though it would be very noble of him if he did.
HH