GPL fanatics have wanted to encourage commercial use all along. Obviously it's a tough sell since it goes completely against the status quo in the software industry, but I think most GPL proponents would like nothing more than to see the software industry adopt the GPL on a large scale.
Remember, the GPL is not anti-business. It's anti-closedness.
I implemented the Blowfish cipher when I was 17. My implementation used a 128-bit key, but it could easily go up to 448 bits. If a 17 year old kid could do this in his bedroom with a cheap laptop and free software, is there ANY reason to think someone with malicious intent and a bit more education couldn't do something much more powerful?
Just got a shiny new PowerPC system. Most of the examples worked immediately, and the one I could find that didn't is due to a bug in SDL, as far as I can tell. The next release of the listings archive will include full PowerPC support.:)
-John
Re:Games and Linux, isn't that what my Windows is
on
Programming Linux Games
·
· Score: 3, Interesting
To some extent, this is already the case.
1. Minor quirks aside, OpenGL is the same on any platform. OpenGL games can usually be ported to Linux with relatively little effort. Although GL has recently lost some ground to Direct3D, hopefully the OpenGL 1.3 specification will improve its acceptance (1.3 was intended specifically to make GL's model of operation more compatible with Direct3D).
2. DRM (support for DRI (3D acceleration)) and fbdev are accepted kernel-level interfaces. You really don't WANT something like SDL or DX in the kernel -- that's just bad design. Much better to use the kernel for minimal hardware interfacing code and leave the rest to userspace, which is exactly what DRI and SDL do.
There's really nothing stopping anyone from implementing a DX-compatible API for Linux on top of the provided kernel services, and in fact WINE is working on just that. I'd rather use SDL myself, though; it's simpler and much lighter than DX, and still has just about everything you need for game programming.
Linux will never be a better Windows than Windows, nor do I think our role as developers should be reduced to simply trying to copy MS. We can do better, even if it means breaking compatibility.
For the record, SDL is a highly portable API. It works on Windows, Mac OS (incl. OS X), BeOS, and just about every modern UNIX.
It's an incredibly addictive robot battle game. You generally build robots with a GUI interface, but for serious hackers there is an object oriented definition language called Ice that compiles into the same VM code as the GUI builder.
For the most part, it's a bunch of artists most people will never hear about. That's unfortunate for the artists (some of them are quite good), but it does mean that they're not "mass produced" alternative, for the most part. MP3.com carries a huge variety of stuff that you'd never find on a radio station.
Musicians often have to make compromises to get signed by a record label. Internet artists do not, for the most part.
Yes, but that's not the point. I don't think anyone is arguing that they shouldn't be able to profit from their work. The problems are a) that they started out as a GPLed project and switched their license, after many people contributed with the impression that future releases would be GPLed also, and b) it's not clear that they even have a right to use a non-GPL license at this point, due to the way they've patched the kernel.
Why would I want a restrictive eBook when I can go to Project Gutenberg, grab plain text files of any number of classic works, and dump them on my Palm Pilot?
This is presuming that I'm interested in an electronic book at all. Cheap paperbacks are so much more pleasant to read.
You can get the beta HipZip codec from the Vorbis developers. Easiest place to catch them is probably irc.openprojects.net, #vorbis.
The HipZip runs a modified version of Cygnus' eCos, apparently. I'm sure you could get Linux on there if you really wanted to, but you'd probably have trouble getting enough information out of Iomega to do it.
The Ogg Vorbis add-on is just that, an add-on. It doesn't affect the existing MP3 or WMA (ugh) support. It does have a few bugs (occasional lockups between tracks, etc), but nothing showstopping, and apparently they'll be fixed in the final release.
Aside from the issue of patents (mainly annoying if you're trying to do something commercial with MP3), there are several good reasons to use Ogg:
-Performance. On certain types of audio, Ogg spins circles around MP3. I'm sure MP3 has its own best cases, but I've yet to find them. In the general case, Ogg holds its own against MP3, usually producing slightly smaller streams at comparable quality.
-Flexibility. Ogg streams are very easy to manipulate. To join two streams, just concatenate the files. Streaming software can arbitrarily reduce a stream's size by chopping off the ends of packets, since the less important information is stored near the end. It's also possible to store multiple logical streams of Vorbis audio in one Ogg stream.
-Quality. Older encoders did have some serious bugs, but the newer releases produce excellent results. I added the Vorbis codec to my HipZip portable player, and I use it for almost all of my music, unless it's already stored in MP3 (in practice, I usually encode my own stuff, so that's not a problem).
And no, I'm not an Ogg Vorbis developer. I've just taken an interest in the project.
What gives? I've been told that Red Hat 7.1 installs flawlessly out of the box on Dell's current top of the line laptops. If you want Linux, pop in a CD and go.
The only real issue I see is that you have to pay for a Windows license. My guess is that you probably had to pay for that anyway with Dell's previous Linux offerings (I believe they were dual boot, though I could be mistaken).
In my experience, Dell makes some of the best laptops out there. I won't go off on my Sony rant again, but there's just no comparison there. They're not an evil company.
I completely agree. I started learning C in 6th grade, and it was a very slow process because I had almost no reference material, other than the (lousy) books that came with Turbo C++ and a couple of C library references. The K&R book is exactly the kind of no-bullshit book I was looking for.
O'Reilly's _Learning Python_ would also have been very popular at my grade school's library if it had been around then. Basically anything to help kids get a solid, self-guided start in programming, without the boring structure of a computer class. I frequently talk to high schoolers online who want to learn more advanced programming but don't have any materials available to do so. One guy I know ended up writing a simple 3D engine in mIRC script of all things because he didn't know where to start with C.
Remember: the ancient Egyptians, for the most part, were about as intelligent as humans today. They lacked our level of technology, but they managed to do some really funky stuff with large stones, and that shows they had more than a bit of mechanical skill and raw labor to throw around.
True, it's unlikely that they could build the equivalent of today's metal pulleys, ropes, and fabrics, but that's what we think of mainly because that's what we have available to us today.
Perhaps they did use kites, and employed an alternate system to attach them? And built the kites out of whatever materials they had at hand?
Actually I've heard that one of the main reasons for the switch is lower power consumption, not necessarily speed. 33 MHz (Dragonball VZ, based on the ubiquitous M68k ColdFire core) is plenty for most Palm apps -- but the ARM will allow them to improve battery life, which is very important for handhelds.
That said, I've recharged my Palm m505 three times during the entire summer, and I use it quite a bit... They're not really hurting in the battery life department. But perhaps this would allow Palm to apply more of that power to the display instead of the processor... The m505 could certainly use a brighter backlight. Every bit helps.
As a hobbyist Palm hacker, I appreciate Palm's fairly open attitude toward PalmOS. This kind of thing is likely to encourage Palm to clamp down on PalmOS with more restrictive licenses.
Maybe this could actually further PalmOS' acceptance, but if I were a Palm Computing exec, I would be very nervous about seeing PalmOS run on non-Palm-approved PDAs (ie, without hefty licensing fees).
Virus writers are a rare breed these days, imho. Maybe the first few Outlook virii were cute hacks, but nowadays they're repetitive and boring, and hardly an exhibition of skill. I had a little bit of respect for the guy who wrote Neuroquila (a DOS virus written in assembly which infected a bunch of computers I was trying to upgrade back in '97 or so)... That took some talent to write. But these script trojans just aren't of the same caliber. In my mind, the people who write these are just script kiddies, and vandals at that.
(Not to excuse Microsoft for failing to secure their product; I consider that criminal negligence at this point. But we all know how likely that is to change.)
These virii do show that a company can get away with maintaining a defective commercial product for an indefinite period of time, as long as they have a sufficient marketing department and pander to weak-kneed sysadmins who are too concerned about their job security to do anything but cover up the problem.
It's dubious whether or not he violated the DMCA. People are upset at Adobe for filing a DMCA complaint to the FBI instead of admitting that their product is faulty, advising its customers of this (since many of them rely on Adobe eBooks for their businesses), and fixing the problem. Gestapo tactics won't make their fundamentally flawed software any better, and until that happens, they're knowingly selling a defective product to their customers. One of the major problems with the DMCA is that it encourages companies to rely on litigation rather than solid engineering to protect their data.
(And yes, I've written several letters to various congresspeople about this. I encourage everyone to do the same.)
He gave a talk at Defcon on the (lack of) security in Adobe eBook documents. Adobe markets eBooks as a secure way for publishers to distribute books, and it turns out that they aren't so secure after all. Dimitry demonstrated this, and Adobe complained to the FBI that he was violating the DMCA. He's now in jail.
I don't own a TV (except a WinTV card with the open source Linux driver), and this kind of crap doesn't make me want to run out and get one. So screw them. The local bookstore doesn't have annoying copy control, and there's less advertising.
I posted this yesterday in another comment, but my feelings on MagicGate Memory Sticks are relevant to this too:
Slashdot Mirror
GPL fanatics have wanted to encourage commercial use all along. Obviously it's a tough sell since it goes completely against the status quo in the software industry, but I think most GPL proponents would like nothing more than to see the software industry adopt the GPL on a large scale.
Remember, the GPL is not anti-business. It's anti-closedness.
-John
I implemented the Blowfish cipher when I was 17. My implementation used a 128-bit key, but it could easily go up to 448 bits. If a 17 year old kid could do this in his bedroom with a cheap laptop and free software, is there ANY reason to think someone with malicious intent and a bit more education couldn't do something much more powerful?
-John
Thanks Wade. But since when were YOU into game programming? :)
-John
Just got a shiny new PowerPC system. Most of the examples worked immediately, and the one I could find that didn't is due to a bug in SDL, as far as I can tell. The next release of the listings archive will include full PowerPC support. :)
-John
To some extent, this is already the case.
1. Minor quirks aside, OpenGL is the same on any platform. OpenGL games can usually be ported to Linux with relatively little effort. Although GL has recently lost some ground to Direct3D, hopefully the OpenGL 1.3 specification will improve its acceptance (1.3 was intended specifically to make GL's model of operation more compatible with Direct3D).
2. DRM (support for DRI (3D acceleration)) and fbdev are accepted kernel-level interfaces. You really don't WANT something like SDL or DX in the kernel -- that's just bad design. Much better to use the kernel for minimal hardware interfacing code and leave the rest to userspace, which is exactly what DRI and SDL do.
There's really nothing stopping anyone from implementing a DX-compatible API for Linux on top of the provided kernel services, and in fact WINE is working on just that. I'd rather use SDL myself, though; it's simpler and much lighter than DX, and still has just about everything you need for game programming.
Linux will never be a better Windows than Windows, nor do I think our role as developers should be reduced to simply trying to copy MS. We can do better, even if it means breaking compatibility.
For the record, SDL is a highly portable API. It works on Windows, Mac OS (incl. OS X), BeOS, and just about every modern UNIX.
-John
http://www.cognitoy.com
And, of course, Loki ported it to Linux.
It's an incredibly addictive robot battle game. You generally build robots with a GUI interface, but for serious hackers there is an object oriented definition language called Ice that compiles into the same VM code as the GUI builder.
-John
http://archive.nytimes.com/2001/09/04/business/04D EAL.html
Just replace www with archive.
-John
For the most part, it's a bunch of artists most people will never hear about. That's unfortunate for the artists (some of them are quite good), but it does mean that they're not "mass produced" alternative, for the most part. MP3.com carries a huge variety of stuff that you'd never find on a radio station.
Musicians often have to make compromises to get signed by a record label. Internet artists do not, for the most part.
-John
Yes, but that's not the point. I don't think anyone is arguing that they shouldn't be able to profit from their work. The problems are a) that they started out as a GPLed project and switched their license, after many people contributed with the impression that future releases would be GPLed also, and b) it's not clear that they even have a right to use a non-GPL license at this point, due to the way they've patched the kernel.
-John
Why would I want a restrictive eBook when I can go to Project Gutenberg, grab plain text files of any number of classic works, and dump them on my Palm Pilot?
This is presuming that I'm interested in an electronic book at all. Cheap paperbacks are so much more pleasant to read.
-John
You can get the beta HipZip codec from the Vorbis developers. Easiest place to catch them is probably irc.openprojects.net, #vorbis.
The HipZip runs a modified version of Cygnus' eCos, apparently. I'm sure you could get Linux on there if you really wanted to, but you'd probably have trouble getting enough information out of Iomega to do it.
The Ogg Vorbis add-on is just that, an add-on. It doesn't affect the existing MP3 or WMA (ugh) support. It does have a few bugs (occasional lockups between tracks, etc), but nothing showstopping, and apparently they'll be fixed in the final release.
-John
Aside from the issue of patents (mainly annoying if you're trying to do something commercial with MP3), there are several good reasons to use Ogg:
-Performance. On certain types of audio, Ogg spins circles around MP3. I'm sure MP3 has its own best cases, but I've yet to find them. In the general case, Ogg holds its own against MP3, usually producing slightly smaller streams at comparable quality.
-Flexibility. Ogg streams are very easy to manipulate. To join two streams, just concatenate the files. Streaming software can arbitrarily reduce a stream's size by chopping off the ends of packets, since the less important information is stored near the end. It's also possible to store multiple logical streams of Vorbis audio in one Ogg stream.
-Quality. Older encoders did have some serious bugs, but the newer releases produce excellent results. I added the Vorbis codec to my HipZip portable player, and I use it for almost all of my music, unless it's already stored in MP3 (in practice, I usually encode my own stuff, so that's not a problem).
And no, I'm not an Ogg Vorbis developer. I've just taken an interest in the project.
-John
What gives? I've been told that Red Hat 7.1 installs flawlessly out of the box on Dell's current top of the line laptops. If you want Linux, pop in a CD and go.
The only real issue I see is that you have to pay for a Windows license. My guess is that you probably had to pay for that anyway with Dell's previous Linux offerings (I believe they were dual boot, though I could be mistaken).
In my experience, Dell makes some of the best laptops out there. I won't go off on my Sony rant again, but there's just no comparison there. They're not an evil company.
-John
I completely agree. I started learning C in 6th grade, and it was a very slow process because I had almost no reference material, other than the (lousy) books that came with Turbo C++ and a couple of C library references. The K&R book is exactly the kind of no-bullshit book I was looking for.
O'Reilly's _Learning Python_ would also have been very popular at my grade school's library if it had been around then. Basically anything to help kids get a solid, self-guided start in programming, without the boring structure of a computer class. I frequently talk to high schoolers online who want to learn more advanced programming but don't have any materials available to do so. One guy I know ended up writing a simple 3D engine in mIRC script of all things because he didn't know where to start with C.
-John
Remember: the ancient Egyptians, for the most part, were about as intelligent as humans today. They lacked our level of technology, but they managed to do some really funky stuff with large stones, and that shows they had more than a bit of mechanical skill and raw labor to throw around.
True, it's unlikely that they could build the equivalent of today's metal pulleys, ropes, and fabrics, but that's what we think of mainly because that's what we have available to us today.
Perhaps they did use kites, and employed an alternate system to attach them? And built the kites out of whatever materials they had at hand?
-John
Sam ported SDL to the PS2 graphics chip not too long ago. The code is in CVS.
:)
Anyone up for porting Mesa?
-John
Actually I've heard that one of the main reasons for the switch is lower power consumption, not necessarily speed. 33 MHz (Dragonball VZ, based on the ubiquitous M68k ColdFire core) is plenty for most Palm apps -- but the ARM will allow them to improve battery life, which is very important for handhelds.
That said, I've recharged my Palm m505 three times during the entire summer, and I use it quite a bit... They're not really hurting in the battery life department. But perhaps this would allow Palm to apply more of that power to the display instead of the processor... The m505 could certainly use a brighter backlight. Every bit helps.
-John
As a hobbyist Palm hacker, I appreciate Palm's fairly open attitude toward PalmOS. This kind of thing is likely to encourage Palm to clamp down on PalmOS with more restrictive licenses.
Maybe this could actually further PalmOS' acceptance, but if I were a Palm Computing exec, I would be very nervous about seeing PalmOS run on non-Palm-approved PDAs (ie, without hefty licensing fees).
Nice hack, though.
-John
Virus writers are a rare breed these days, imho. Maybe the first few Outlook virii were cute hacks, but nowadays they're repetitive and boring, and hardly an exhibition of skill. I had a little bit of respect for the guy who wrote Neuroquila (a DOS virus written in assembly which infected a bunch of computers I was trying to upgrade back in '97 or so)... That took some talent to write. But these script trojans just aren't of the same caliber. In my mind, the people who write these are just script kiddies, and vandals at that.
(Not to excuse Microsoft for failing to secure their product; I consider that criminal negligence at this point. But we all know how likely that is to change.)
These virii do show that a company can get away with maintaining a defective commercial product for an indefinite period of time, as long as they have a sufficient marketing department and pander to weak-kneed sysadmins who are too concerned about their job security to do anything but cover up the problem.
-John
It's dubious whether or not he violated the DMCA. People are upset at Adobe for filing a DMCA complaint to the FBI instead of admitting that their product is faulty, advising its customers of this (since many of them rely on Adobe eBooks for their businesses), and fixing the problem. Gestapo tactics won't make their fundamentally flawed software any better, and until that happens, they're knowingly selling a defective product to their customers. One of the major problems with the DMCA is that it encourages companies to rely on litigation rather than solid engineering to protect their data.
(And yes, I've written several letters to various congresspeople about this. I encourage everyone to do the same.)
-John
That's the whole point.
He gave a talk at Defcon on the (lack of) security in Adobe eBook documents. Adobe markets eBooks as a secure way for publishers to distribute books, and it turns out that they aren't so secure after all. Dimitry demonstrated this, and Adobe complained to the FBI that he was violating the DMCA. He's now in jail.
-John
In stores in a few weeks. Hopefully. :)
-John
-John
I don't own a TV (except a WinTV card with the open source Linux driver), and this kind of crap doesn't make me want to run out and get one. So screw them. The local bookstore doesn't have annoying copy control, and there's less advertising.
I posted this yesterday in another comment, but my feelings on MagicGate Memory Sticks are relevant to this too:
http://treklink.net/~overcode/copy-rant.txt
-John
I have a few thoughts on the Memory Stick, especially the MagicGate "technology"...
http://treklink.net/~overcode/copy-rant.txt
-John