Rubbish. It's a prime example of how a corporation saw something for nothing, took it, and is renegging on the deal it engaged it. Few people, if any, in the Linux community got anything out of them using our work, but they got a complete operating system that would have otherwise cost them a shitload to develop or buy. You can bitch about the GPL being restrictive if you want, but all that's really being asked is to give as well as take. Except instead of just asking nicely, it's a legal requirement.
I'm guessing your not a Linux user or contributor, judging by your last remark. Most of us don't give a toss about Linux widespread acceptance, we just want Linux to work better for us - if others like it too, then that's a bonus. We're certainly not doing it to make other companies wealthy on our backs with nothing in return, even if it is a token gesture in comparison.
Some years ago before I became a geek and thus unappealing to women, I was seeing an otherwise lovely girl who had a wart on her thumb. She had a cough and a few other ailments that happen when you're a poor student and stressed out too. Anyway, I was going though an "alternative" phase, and suggested she try colliodial silver. And strangely enough it worked. The wart was gone in a couple of weeks. Being much more cynical and practical now, I look back and think WTF? But if anyone's got warts that need killin' and they don't want N2 scars, I'd say it's worth a try.
Re:Writing is bad enough, testing is worse
on
Exploiting Software
·
· Score: 1
That would be a very good - and extraordinarily boring - Linux kernel sub-project. A testing framework that runs continuously over the source tree and flags when things are broken/out-of-limits/unsafe etc, makes a lot of sense. It's the sort of thing that should be on the kernel.org page, showing that state of say, stable, pre-patch, development, etc. I'd sure feel safer *knowing* that my OS is secure and reliable, not just that smart folks *think* it is.
The Ford Racing "high flow" Mustang injectors I bought for my Range Rover recently are actually made by Denso of Japan. Simply upping the injector flow isn't much good without increasing the airflow too, otherwise you'll find the computer learns about the excess fuel and reduces the pulsewidth anyway - except on openloop like hard acceleration, so it'll be running richer and most likely be producing more power (and wasting gas). Another mod might be swapping the fuel pressure regulator for something a little higher, say go from 2.7 bar up to 3.5 bar, to increase fuel flow a little. Probably cheaper and just as effective as swapping the injectors themselves, as long as the computer doesn't get pissy about the higher fuel pressure and produce a warning code.
But yes, you're right about the service issue. It's not as bad as this cat makes out though, as much of the info is actaully available in various standards anyway. Check out SAE or ISO, and a lot of it's there for the buying. I know because I'm buying some of it to produce an aftermarket engine management system. Sure, having a cost is a pain for the small garage, but last time I checked, good documentation was rarely free.
Yeh, the average street SUV would be toast out there. Having said that, some of them are fairly capable once they've been suitable prepared. I'm really not sure if I'd rather be in a new Range Rover or my worked '82 model when things got "interesting" off-road. Certainly the new one would be better for high speed cross country travel. Maybe I'll rethink that once the electronic traction and stability control is finished...
I checked SciAutronics, and I dunno... It looks agile and able, but a larger design would be better able to soak up the rocks and bumps I think. I prefer the CyberRider buggy with big tires and airbag suspension (but no 4wd:-( ). Gives more leeway if your suspension can handle desert terrain rather than making the electronic systems dodge football sized rocks.
I reckon it'll be a while before computers are able to rockcrawl and make the hard choices about where to place their wheels to get past an obstacle, so my money is on building a system with enough sense of self (position, abilities, state) and objective (goals, risks) to find the easiest path, then let a supple chassis soak up much of the details. Current 4wd off-road racer tech is plenty capable of this, so as usual it comes down to the software.
I'm happy with ads next to the results, as long as it filters the results themselves. AFAIK, Google doesn't get paid for regular search hits, just the "marked as sponsored" stuff, so that arrangement is fine with me. To be fair, I'd probably pay a trivial sum for access to a better-than-Google-and-no-ads search engine, as I do for Safari Bookshelf and some of my J2EE tools and docs, along with car reference manuals even. OT, but I'd say the age of mainstream payed-subscription content is well and truely upon us.
Sources like Lexis-Nexis appear to be aimed at businesses anyway, and most punters won't have even heard of them, let alone pay the large sums asked for them.
I think the "next big thing" will be information only search engines. Filtering through the plethora of advertising and crud is getting more tiresome as the punters learn how to optimise their rankings. Something like Google but with a Bayesian spam filter attached to the front end to filter the results for me...
Of course they shouldn't, but they will anyway. Australia is pretty good at bending over for the United States, and sending one man to PITA prison is a sacrifice Australian politicians will happily make to stay in favour for the next round of trade talks.
True, and we do it without laser rangefinders and a dazzling array of sensors. We do it with a fairly basic kit of visual, audio, and accelerometer inputs, a little bit of force-feedback (depending on how luxo the car is), plus a bunch of rules and a heirachy of goals for any given situation. I can't help thinking that some of these guys approached it the wrong way. I don't know enough about the systems used to say, but I'll guess more focus should be put on the system's self-awareness than on being a big ol' sensor rack with some DSP processors hooked up to some wheels.
Also, the rules say the course can be driven by an SUV. So why invest time and money building a bunch of oddball off-roaders that fail with mechanical problems? Granted, some made sense (CyberRider, for example) and RedTeam's choice of base was a good one, but I wonder if people have forgotten the KISS principal in all the excitement.
I didn't forget that at all. You're mis-reading my post. The point is that if someone further up the GPL distribution chain breaches the copyright of *non* GPL'd software, those further down are still distributing it. Whether it's a problem for them or not is up to the courts to decide.
There's nothing magic about the GPL that lets this happen, other than that the distribution chain can be much larger. Of course the correct result wil be something like my second argument, but I wouldn't put it past MS (or *AA, etc) to try to argue the first one. Like I posted earlier, it smells like FUD to me. Anything targeting the end user is most often bluster and bullying.
No, I mean if Person A rips off some of Company B's source code, and includes it in a GPL'd product, which is then distributed by Person C.
There's two obvious arguments: 1) Person C is an IP Thief(tm) distributing Company B's copyrighted work (which they are) and must be sued and made an example of. 2) Person C is an innocent party and received the copy in good faith, and the liability is deferred back to Person A, who actually did the nasty deed.
Hopefully the latter argument would win, but once you get laywers involved, anything can go wrong...
C&D's don't mean anything - they're just scare tactics. The thing about the GPL which I suspect he's alluding to, is that you are allowed to redistributed copyrighted works if you stick to the terms. However, the act of redistributing GPL'd works could be argued as also breaching traditional copyright, if a previous contributor has included someone else's copyrighted work. Whether it would stick I don't know, but that's the obvious end-user attack - if they are redistributing the code in question. It doesn't affect Joe User who downloads Mandrake and runs it on his box at home.
First paragraph, yes. Second one, not so sure. I've never seen it written that receiving copyrighted material was against the law. Certainly distributing it is, but I'm not so sure about it being a copyright violation to obtain them.
I think there's a difference between waiving your consumer rights (eg put wide wheels on your car, don't go crying GM if the axle breaks) and it being illegal to use products you've bought in any manner you see fit (as long as you're not messing with others rights). Eg, running your old games on an emulator == OK, but putting them on the 'net for others to run == BAD.
Rubbish. Unless there is a written and legally binding agreement between two parties, there is no "license" in effect. When Joe Public buys a Nintendo cartridge he's buying a product, and the manufacturer writing arbitrary restrictions on or inside the box does not change this. It's no different to buying a book from a book store - and good luck enforcing a book's EULA saying you can't read it on the train because someone might read over your shoulder.
I agree to an extent, which is why I mentioned rendered Jue from Final Flight Of The Osiris in the Animatrix (probably the best 3D rendered person I've seen yet, better than Dawn, IMO). As with most things 3D a lot depends on the artists... so I'm guessing a decent 3D scanner and a willing model will make this attainable to the average porn overlord (when 3D scanner prices drop a bit more). I was aware that it has been around for a long time, and quite possibly stereo 3D doesn't work for most people, but I think it makes an interesting feature for a site. Eitherway, interactivity is a must.
Rubbish. It's a prime example of how a corporation saw something for nothing, took it, and is renegging on the deal it engaged it. Few people, if any, in the Linux community got anything out of them using our work, but they got a complete operating system that would have otherwise cost them a shitload to develop or buy. You can bitch about the GPL being restrictive if you want, but all that's really being asked is to give as well as take. Except instead of just asking nicely, it's a legal requirement.
I'm guessing your not a Linux user or contributor, judging by your last remark. Most of us don't give a toss about Linux widespread acceptance, we just want Linux to work better for us - if others like it too, then that's a bonus. We're certainly not doing it to make other companies wealthy on our backs with nothing in return, even if it is a token gesture in comparison.
Some years ago before I became a geek and thus unappealing to women, I was seeing an otherwise lovely girl who had a wart on her thumb. She had a cough and a few other ailments that happen when you're a poor student and stressed out too. Anyway, I was going though an "alternative" phase, and suggested she try colliodial silver. And strangely enough it worked. The wart was gone in a couple of weeks. Being much more cynical and practical now, I look back and think WTF? But if anyone's got warts that need killin' and they don't want N2 scars, I'd say it's worth a try.
That would be a very good - and extraordinarily boring - Linux kernel sub-project. A testing framework that runs continuously over the source tree and flags when things are broken/out-of-limits/unsafe etc, makes a lot of sense. It's the sort of thing that should be on the kernel.org page, showing that state of say, stable, pre-patch, development, etc. I'd sure feel safer *knowing* that my OS is secure and reliable, not just that smart folks *think* it is.
The Ford Racing "high flow" Mustang injectors I bought for my Range Rover recently are actually made by Denso of Japan. Simply upping the injector flow isn't much good without increasing the airflow too, otherwise you'll find the computer learns about the excess fuel and reduces the pulsewidth anyway - except on openloop like hard acceleration, so it'll be running richer and most likely be producing more power (and wasting gas). Another mod might be swapping the fuel pressure regulator for something a little higher, say go from 2.7 bar up to 3.5 bar, to increase fuel flow a little. Probably cheaper and just as effective as swapping the injectors themselves, as long as the computer doesn't get pissy about the higher fuel pressure and produce a warning code.
But yes, you're right about the service issue. It's not as bad as this cat makes out though, as much of the info is actaully available in various standards anyway. Check out SAE or ISO, and a lot of it's there for the buying. I know because I'm buying some of it to produce an aftermarket engine management system. Sure, having a cost is a pain for the small garage, but last time I checked, good documentation was rarely free.
Yeh, the average street SUV would be toast out there. Having said that, some of them are fairly capable once they've been suitable prepared. I'm really not sure if I'd rather be in a new Range Rover or my worked '82 model when things got "interesting" off-road. Certainly the new one would be better for high speed cross country travel. Maybe I'll rethink that once the electronic traction and stability control is finished...
:-( ). Gives more leeway if your suspension can handle desert terrain rather than making the electronic systems dodge football sized rocks.
I checked SciAutronics, and I dunno... It looks agile and able, but a larger design would be better able to soak up the rocks and bumps I think. I prefer the CyberRider buggy with big tires and airbag suspension (but no 4wd
I reckon it'll be a while before computers are able to rockcrawl and make the hard choices about where to place their wheels to get past an obstacle, so my money is on building a system with enough sense of self (position, abilities, state) and objective (goals, risks) to find the easiest path, then let a supple chassis soak up much of the details. Current 4wd off-road racer tech is plenty capable of this, so as usual it comes down to the software.
I'm happy with ads next to the results, as long as it filters the results themselves. AFAIK, Google doesn't get paid for regular search hits, just the "marked as sponsored" stuff, so that arrangement is fine with me. To be fair, I'd probably pay a trivial sum for access to a better-than-Google-and-no-ads search engine, as I do for Safari Bookshelf and some of my J2EE tools and docs, along with car reference manuals even. OT, but I'd say the age of mainstream payed-subscription content is well and truely upon us.
Sources like Lexis-Nexis appear to be aimed at businesses anyway, and most punters won't have even heard of them, let alone pay the large sums asked for them.
I think the "next big thing" will be information only search engines. Filtering through the plethora of advertising and crud is getting more tiresome as the punters learn how to optimise their rankings. Something like Google but with a Bayesian spam filter attached to the front end to filter the results for me...
Think about it, nowadays nearly everyone is a criminal, either he shared some files, or unknowingly infringed some patents.
Yep, and all you have to do is prove it. It's 2004. Are you a criminal yet?
Of course they shouldn't, but they will anyway. Australia is pretty good at bending over for the United States, and sending one man to PITA prison is a sacrifice Australian politicians will happily make to stay in favour for the next round of trade talks.
True, and we do it without laser rangefinders and a dazzling array of sensors. We do it with a fairly basic kit of visual, audio, and accelerometer inputs, a little bit of force-feedback (depending on how luxo the car is), plus a bunch of rules and a heirachy of goals for any given situation. I can't help thinking that some of these guys approached it the wrong way. I don't know enough about the systems used to say, but I'll guess more focus should be put on the system's self-awareness than on being a big ol' sensor rack with some DSP processors hooked up to some wheels.
Also, the rules say the course can be driven by an SUV. So why invest time and money building a bunch of oddball off-roaders that fail with mechanical problems? Granted, some made sense (CyberRider, for example) and RedTeam's choice of base was a good one, but I wonder if people have forgotten the KISS principal in all the excitement.
Word. What's wrong with Video Lan Server or something? Must be NDH syndrome.
Meh. I think they're just making numbers up now. It's gone back down to 7 miles.
SciAutronics II seems to be advancing anyway, despite being disabled. Is this the beginning of the end???
I didn't forget that at all. You're mis-reading my post. The point is that if someone further up the GPL distribution chain breaches the copyright of *non* GPL'd software, those further down are still distributing it. Whether it's a problem for them or not is up to the courts to decide.
Anyone know how to let ShockWave run on Linux?
There's nothing magic about the GPL that lets this happen, other than that the distribution chain can be much larger. Of course the correct result wil be something like my second argument, but I wouldn't put it past MS (or *AA, etc) to try to argue the first one. Like I posted earlier, it smells like FUD to me. Anything targeting the end user is most often bluster and bullying.
No, I mean if Person A rips off some of Company B's source code, and includes it in a GPL'd product, which is then distributed by Person C.
There's two obvious arguments:
1) Person C is an IP Thief(tm) distributing Company B's copyrighted work (which they are) and must be sued and made an example of.
2) Person C is an innocent party and received the copy in good faith, and the liability is deferred back to Person A, who actually did the nasty deed.
Hopefully the latter argument would win, but once you get laywers involved, anything can go wrong...
C&D's don't mean anything - they're just scare tactics. The thing about the GPL which I suspect he's alluding to, is that you are allowed to redistributed copyrighted works if you stick to the terms. However, the act of redistributing GPL'd works could be argued as also breaching traditional copyright, if a previous contributor has included someone else's copyrighted work. Whether it would stick I don't know, but that's the obvious end-user attack - if they are redistributing the code in question. It doesn't affect Joe User who downloads Mandrake and runs it on his box at home.
Sounds like FUD to me.
First paragraph, yes. Second one, not so sure. I've never seen it written that receiving copyrighted material was against the law. Certainly distributing it is, but I'm not so sure about it being a copyright violation to obtain them.
I think it's not illegal to download it, but rather it's illegal to distribute it (eg, upload it). Small distinction, but an important one.
I think there's a difference between waiving your consumer rights (eg put wide wheels on your car, don't go crying GM if the axle breaks) and it being illegal to use products you've bought in any manner you see fit (as long as you're not messing with others rights). Eg, running your old games on an emulator == OK, but putting them on the 'net for others to run == BAD.
Rubbish. Unless there is a written and legally binding agreement between two parties, there is no "license" in effect. When Joe Public buys a Nintendo cartridge he's buying a product, and the manufacturer writing arbitrary restrictions on or inside the box does not change this. It's no different to buying a book from a book store - and good luck enforcing a book's EULA saying you can't read it on the train because someone might read over your shoulder.
You *are* allowed to make backups and fair-use copies. Wailing lawyers don't change this fact.
I agree to an extent, which is why I mentioned rendered Jue from Final Flight Of The Osiris in the Animatrix (probably the best 3D rendered person I've seen yet, better than Dawn, IMO). As with most things 3D a lot depends on the artists... so I'm guessing a decent 3D scanner and a willing model will make this attainable to the average porn overlord (when 3D scanner prices drop a bit more). I was aware that it has been around for a long time, and quite possibly stereo 3D doesn't work for most people, but I think it makes an interesting feature for a site. Eitherway, interactivity is a must.