SSHHH, music execs should be making boat loads [literally] of cash a year... It's um the stress... This is the same with the movie industry and sports [well American especially].
Maybe if they lowered their salary to reasonable limits they'd find that they're not so strapped for cash.
Well that and if they spent less on advertising and more on music lessons we'd have people who can actually sing and perform well.
People seem to dismiss outright that a classical music academy is not useful for even a rock, alternative, pop music performer... I remember when I was taking private piano lessons the teacher would complain that the younger students wanted to play pop music instead of classical pieces.
Now there are some good modern pieces [e.g. that have tricky fingering or timings that develop the talent] but most are just repetitive and basic. That's why they use classical music because it has new styles [e.g. technique] that help the student learn how to be expressive with the instrument...
But now you get people like Ashlee or Lindsey who have probably never taken a formal lesson [outside of B.J. Mechanics 101] in music and couldn't produce an original tune to save their lives [I mean really, does Lohan honestly think her songs are original? Hello, Britney [who is a copy of Madonna] did the same songs the year before...]
Yes, AMD has more gates todo the same thing a RISC could do [e.g. decoders] but they also add a bunch of gates to out-of-order execution, register renaming, that the typical RISC [e.g. ARM/MIPS] does not have.
Those are tricks you need EVEN in the RISC world [although perhaps less as the # of registers increases].
While I agree x86_64 is a kludge it seems to work well. The ideal would be AMD making a new ISA, replacing their existing decoders but keeping the rest the same. That would remove the prefix bytes/etc nonsense they have. Personally I could go for AMD dropping 32-bit native support [by that I mean protected mode, keep longmode [which can run 32-bit binaries] but drop the 16-bit VM nonsense].
Though code density is not such a horible problem. More registers usually means fewer instructions [less load/store] which means the net affect is usually zero or positive.
My big beef with PPC such as the G5 is that they're expensive and generate a shit load of heat. The AMD64 by comparison is a lot cheaper and doesn't generate nearly as much heat.
I think you'll find the gains from 16 extra registers is less than what the [for example] AMD gains from having three pipelines, register file, etc...
It's like cache, throwing more registers pays off big to start [say going from 1 to 2, 2 to 4,...] but dies off quick after that.
Take apart that 5% of your program that takes 95% of the time and see how many registers it actually uses in the inner loop.
With bignum math for instance, inner loops usually amount to 3 registers for an accumulator, 1 for a step counter, 2 for source pointers and 1 for an outer loop counter, 7 registers in total...
Take the EM64T case, it implements x86_64 as well but AMD still pwnz it bad. Why? Well let's see, three [not one] dedicated decoders, three ALU pipelines with 8-step schedulers [re parallelism], etc...
Intel still pwnz AMD when it comes to SSE2 and memory ops but that gap has been closing with every new AMD release [AMD64 for instance has more SSE2 opcodes implemented as directpath instead of MicroROM] where in the Intel camp their cpus haven't really been getting ANY better...
Something like ReiserFS doesn't scale well at the low end. On a 128M memory stick [iirc] the file system takes something like 16-20MB off the top. That's 15% of the drive gone instantly [compared to the 5M that FAT takes].
Granted if you have a 1GB memory stick the 20M or so that Reiser takes is less of a pain and the gains you get from the stability are more worth it...
Not really, just tell your customer "you're not using it right, read the damn manual" and be done with.
What really makes companies that provide hardware look bad is when you use THEIR drivers and the device is still flaky and unreliable.
What hurts Microsoft is the low quality of the software they write [well depends, some tools are decent but as a whole they're pretty bad]. I've never called their techsupport, I have called their activation drone before though...
If a customer calls and says "I tried to l33t mod the device and it borked" and then gets upset when the dude on the other and says "STBU"... well too bad really. And in terms of "word of mouth" damage it's low.
If any of my friends tried to "home brew upgrade" their motherboards or processors or something and it blew up I'd just laugh [at them] and lend them money for a new box in the meantime.
Again you missed the point. The police officer can give you that ticket EVEN if the box ends up being defective.
What is at stake here is the right to mount a defense [in most cases you can't refuse or deny being given a ticket, but you can fight it in court and get it dismissed].
Citing economic reasons to not disclose how it works is stupid. That's the line of work they were getting into they should have known and worked around that from the start.
This is another "different value" market. The value [to the police] isn't so much in the proprietary DSP algorithms they use [or whatever] but in the assemblage. It's put together and tested to meet some minimal standards so the police don't have to. That's the value.
The fact that it uses publicly available algorithms and designs doesn't take away from the value.
"I realize Slashdot is the land of the tinfoil hats, but stop for a second and realize this isn't a story about police corruption. It's about criminals who were obviously felt to be intoxicated by a trained officer of the law getting away with it because of a sleazy loophole."
Criminals? Says who? You're INNOCENT until PROVEN guilty. And that loophole is the american justice system.
Suppose you just finished your PhD, you got a nice job in the government and part of that was a mandatory drug test. So you pee into the cup, admire it's bounty and pass it off. In the lab they accidently cross-contaminate your urine sample with that of an inmate in a local prison. Now they bring it back and it turns out your a drug addict. Well kiss your government job goodbye. You now also have a drug offense on your permanent record.
Oh, but you should [by your logic] have no right to question in which manner your "presumed guilt" was determined. So now you vested 150k into an education that will basically make you the most intelligent burger flipper in the world.
Congradulations, civil rights are just a "sleezy loophole".
The problem is brushing the proverbial mosquito off your arm also knocks off um... fuck these metaphors...
You piss off customers by making flaky non-portable drivers. ATI for instance doesn't really share the user base in Linux as nvidia does and it's solely because their drivers suck.
But think about it this way, if the company thinks their TOP OF THE LINE product can be easily replicated today using part-time volunteers spread out across the globe... of how much value is their product anyways?
Look at the x86 CPU [or even ARM]. Look at GPUs, etc..
There isn't a lot of competition in high end products because just knowing how to "jigger the doohickey" isn't enough to figure out how the doohickey works.
If your trying to sell something like a 16550 UART in 2005 as a standalone IC... well... um you got issues.
Complex enough products [to which the value is greater because the ability to solve it on ones own is lower] are hard to mimic to a level of similarly demanded by the customer.
Look at Transmeta and VIA. They're both x86 processors [well sorta] that did little to dent the x86 world [of which AMD which produce awesome cpus is even a small part of].
It's not enough that you're functionality equivalent you have to be just as efficient and cost effective.
I mean, I could buy that 100$ slow as shit Transmeta processor or I could buy that 200$ very fast and power efficient AMD64 processor... to me the extra 200$ is worth it since I plan to own the processor for several years the average cost compared to the time savings is huge...
So no, I don't think companies revealing the interface reveal their meaningful valuable secrets. It's just a matter of controlling how and when and where the user uses the device.
This, ladies and gentlemen is why I own a LG superdrive.
I put it in my AMD64 Gentoo based box, booted up and it "just worked". No drivers, no special CD burning software [outside of cdrecord and growisofs], etc...
There is no value for me in commercial CDR tools since free [and decently working ones] exist already. The sooner...HARDWARE... manufacturers realize that... the better.
The "worth" of the card in that case factors into several camps...
1. Is it efficient, can I get high FPS without killing the CPU?
2. Does it look good? Is there quality?
3. Does it work reliably? Can I use this card to develop on and trust that my customer sees the same thing?
#1 depends on the hardware
#2 and #3 depend moreso on the software.
Even still it's a 3D graphics library that works only with the nVidia [or ATI or...] interface. You'd have to spend time and energy changing it to work with a different piece of hardware...
That said..
What's stopping giving out the INTERFACE and not the software side of things? You still have to BUY the card. Software alone can't emulate that.
Think of the value proposition in terms of things you can't easily provide yourself. What does nvidia do that makes me want to give them money?
It certainly isn't writing drivers because we have competent people at xorg-x11 to do that. What they do that others [in the OSS world] cannot do is produce a GPU capable of pushing polygons.
In otherwords I don't want to pay nvidia to write drivers. I want to pay them to design efficient hardware that makes the nippomatics in ut2k4 even more realistic.
BTW, do you pay Intel to write an OS for you to use with your Intel processor? Did you pay sony money to produce music to listen with your sony CD player? Did you pay Maytag to make the dishes you wash with your dishwasher and did you pay Ford for the roads you want to drive your ford truck on?
You could just not document things that are not 100%?
Or am I missing something?
I mean I understand that products are usually evolutionary [e.g. the current model may have the beginnings of stuff that isn't ready yet but will be in the "next model"] but just don't document them.
I dunno about you but pretty much every piece of hardware from CD drive to network card, etc.. has had some piece of windows software which I promptly tossed in the garbage. I didn't buy the drivers [cuz I run hardware that I know works in Linux] I bought the hardware.
If the hardware didn't work out of the box with the Linux kernel the hardware has ZERO value to ME period. So if they want to push the hardware they have to provide the drivers [or specs or both].
The drivers have zero commercial value since they can only be used with that hardware. It isn't as if someone can copy it and use it for their software [unless they mimic the interface].
Again, the hobby groups of the 70s/80s/90s haven't disappeared. You don't even need to spend time on writing drivers if you just give people the specs.
Here's how you do it.
1. Make 4 cards [of whatever].
2. Email the lkml and offer them to people willing to write drivers.
3. Send them the card and the specs and tell them they can keep the card if they write a working driver under the GPL.
4. Take the driver and bundle it with said hardware [or get it merged into the kernel].
5. Profit!
All this costs you is 4 copies of your product [for a 50$ network card this is a mere 200$], postage and a copies of the interface specification.
If the hardware is worth anything the developers have incentive since they get cool hardware for essentially free. So they have incentive to participate...
The problem I have with GPU manufacters like ATI and nvidia is...
Don't you think they both understand what a pixel pipeline is? Or what a vertex shader is? Or... It's not like they're really that different technologies...
Also since when does interface dictate implementation? I mean look at AMD and Intel. Both implement x86 processors, both give out cycle timings, opcode formats, etc... I don't see cheap AMD clones on the market today...
I don't know how a GPU interface works exactly but chances are it has some uniform data format for various things like vertex lists, textures, vertex programs, etc. None of which explains how the GPU pipeline was implemented in silicon...
Again it's just more "our drivers are worth $$$"....
Imagine how crippled the x86 world would be if you had to use Intels compiler and were never told what the opcodes are or how they worked... We would have an exponentially smaller scale computer world I'd say. Specially since the hobby driven folk of the 80s and 90s are what really made the explosion possible.
Specially since you buy the DRIVE not the software. The software is the COST of doing business. The problem is you have ignorant marketting and investors who think "everything has commercial value".
Look at Broadcom. They hold their hardware specs a closely guarded secret [for the most part] and the net affect is you can only use their wifi stuff [reliably] in windows... The problem is without the drivers the hardware has zero customer value. But giving out free drivers lets you SELL hardware since it now has value.
The sad thing it isn't even that you have to write the damn drivers. In the OSS world of BSD/Linux the kernel contributors would GLADLY write a driver for free if it meant they could use some quality hardware with the respective OS. So all it costs the hardware manufacturer is describing the interface [at the high level] of how to talk to the hardware. Since these documents are ROUTINELY produced internally so the software teams can write their windows drivers all it means is you re-brand the.doc file and give it out.
You do realize that just living near the pesticides is enough to get it into your body right?
You and your chemlawn using motherfucking yuppy friends are why I hate society. You get some idiotic believe that something that totally rapes and murders plants magically will have no effect on humans and if you can't immediately see it you must not be exposed to it.
The minute people realize that everything, and I mean EVERYTHING we do is interelated the better.
That and my front and back lawn are decently healthy and we do nothing but spray water on it and cut it regularly. Sure we're not "lawnman" green but it's still a healthy lawn. Most of my neighbours that get their lawns sprayed often neglect their lawns and they yellow half way through the summer...
That's a stupid argument because it has no bearing on the conversation. What, are you 8 years old or something?
Knowing how a jet fighter works has no bearing on whether a little box that says "guilty" is actually working properly. And it certainly has nothing to say about my right to question how and if the box actually works in the first place.
Why is "how" important. It could have false positives [e.g. medicines may trigger it for instance] that make that particular test invalid. It's happened before and I'd be surprised if it never comes up again.
Why is "if" important? It may be well outside acceptable error tolerances and the results are not significant or useful in a court of law.
You guys keep saying "getting off"... Prove to me he was actually DUI. That's what this "little box" is supposed to do. "Getting off" implies he's actually guilty and was not convicted anyways. You think getting arrested is the same as "being guilty" because you have no appreciation for what the law actually says and stands for.
Just wait till your on the wrong end of private corporations wet dream and see how much sympathy you get from your peers who will assume your guilty without having been given a fair chance to defend yourself. Of course you can't imagine this because you're an 8 year old brat skipping second period in school to post on slashdot... Go to class you little brat!
Nobody [certainly not myself] are saying that drinking and driving is a good thing. What I am trying to say is at least have your ducks in a row.
What you call "getting off on technicalities" I call upholding your rights.
I mean having a lawyer present is just a technicality too right? If you're guilty why should you have a defense? That just means you can get "off" with the crime!
Questioning witnesses? Why? They're under oath therefore all they speak is verbatim truth.... etc...
Yeah it would be nice if these drunk drivers just fessed up and did their time. That doesn't mean they don't have a right to question the way they were deemed to have violated the law.
You seemed to be along the line of "if you were caught you must be guilty". As if law enforcement really deserve that unquestionability status.
I say just wait till someone cries rape in your direction and see how much fun the "guilty before innocent" status you're so desperately trying to stick on others feels on yourself.
I mean suppose the guy really wasn't drunk and he was held for a day in jail. He'd miss the next day of work, have to explain to his boss, have the social stigma of being known for being arrested DUI... all before he's been able to even question if the little device that has stung him so actually works within the expected margins of error.
mmm, you can't question if something works unless you know how it works.
For instance, maybe the radar gun works different at night? [making that up but you never know]. If you didn't know how a radar actually worked you wouldn't know to test [or demand that be tested].
Also it's the right of the defense to perform independent testing [just like they can question witnesses...].
I don't see what the big deal is. As a company providing law enforcement equipment/software/etc you should have known WELL in advance that you'd be called upon to explain how it works.
If you're business model depends on selling law enforcement tools that cannot withstand scrutiny... you SHOULD be out of business.
Say something as simple as distilled water... whatever.
The point is we're all humans. When a mistake is made you either throw it out or keep it. In both cases you account for the mistake. It could be you did something that wouldn't affect [or at least you reasonably believe it won't affect] the net result. I mean, say the time for culturing bacteria [of a certain type] has been determined to be optimally at 37C for exactly 37 hours 43 minutes 29 seconds.
Chances are if you're at 37.1C for 37 hours 43 minutes 21 seconds the result won't be damaged enough [significantly] as to require it to be dismissed. Yet I imagine this happens all the time.
The real threat portrayed in this story though is that there is zero accountability. If you didn't have to account for how you came up with the results what's saying you didn't just make up the outcome?
"They are guilty. You know it, and I know it. It seems that they are not disputing the fact that they had been drinking. They are focusing on the way they got caught."
You seem to have no clue what the legal process is. Your INNOCENT until... PROVEN... guilty. Without scrutinizing the box you can't prove guilt.
There is no more discussion worth having on this point.
And if you call upholding your rights "whining" then again, you're living in a bubble that has yet to be pierced...
Slashdot Mirror
How about you look to music academy graduates instead of busty plastic surgeon patrons for MUSICAL talent?
I still have to question what they think of themselves, then what they think of themselves in 30 years... should be interesting.
Tom
SSHHH, music execs should be making boat loads [literally] of cash a year... It's um the stress... This is the same with the movie industry and sports [well American especially].
Maybe if they lowered their salary to reasonable limits they'd find that they're not so strapped for cash.
Well that and if they spent less on advertising and more on music lessons we'd have people who can actually sing and perform well.
People seem to dismiss outright that a classical music academy is not useful for even a rock, alternative, pop music performer... I remember when I was taking private piano lessons the teacher would complain that the younger students wanted to play pop music instead of classical pieces.
Now there are some good modern pieces [e.g. that have tricky fingering or timings that develop the talent] but most are just repetitive and basic. That's why they use classical music because it has new styles [e.g. technique] that help the student learn how to be expressive with the instrument...
But now you get people like Ashlee or Lindsey who have probably never taken a formal lesson [outside of B.J. Mechanics 101] in music and couldn't produce an original tune to save their lives [I mean really, does Lohan honestly think her songs are original? Hello, Britney [who is a copy of Madonna] did the same songs the year before...]
Tom
Yes, AMD has more gates todo the same thing a RISC could do [e.g. decoders] but they also add a bunch of gates to out-of-order execution, register renaming, that the typical RISC [e.g. ARM/MIPS] does not have.
Those are tricks you need EVEN in the RISC world [although perhaps less as the # of registers increases].
While I agree x86_64 is a kludge it seems to work well. The ideal would be AMD making a new ISA, replacing their existing decoders but keeping the rest the same. That would remove the prefix bytes/etc nonsense they have. Personally I could go for AMD dropping 32-bit native support [by that I mean protected mode, keep longmode [which can run 32-bit binaries] but drop the 16-bit VM nonsense].
Though code density is not such a horible problem. More registers usually means fewer instructions [less load/store] which means the net affect is usually zero or positive.
My big beef with PPC such as the G5 is that they're expensive and generate a shit load of heat. The AMD64 by comparison is a lot cheaper and doesn't generate nearly as much heat.
Tom
x86_64 has 16 GPRs and 16 XMM [simd] registers.
...] but dies off quick after that.
I think you'll find the gains from 16 extra registers is less than what the [for example] AMD gains from having three pipelines, register file, etc...
It's like cache, throwing more registers pays off big to start [say going from 1 to 2, 2 to 4,
Take apart that 5% of your program that takes 95% of the time and see how many registers it actually uses in the inner loop.
With bignum math for instance, inner loops usually amount to 3 registers for an accumulator, 1 for a step counter, 2 for source pointers and 1 for an outer loop counter, 7 registers in total...
Take the EM64T case, it implements x86_64 as well but AMD still pwnz it bad. Why? Well let's see, three [not one] dedicated decoders, three ALU pipelines with 8-step schedulers [re parallelism], etc...
Intel still pwnz AMD when it comes to SSE2 and memory ops but that gap has been closing with every new AMD release [AMD64 for instance has more SSE2 opcodes implemented as directpath instead of MicroROM] where in the Intel camp their cpus haven't really been getting ANY better...
Tom
Buy an AMD64 [for half the cost] and install Gentoo on it [for free].
;-)
There ya go, efficient high performance 64-bit desktop platform which leaves you enough money for a good road trip with some buddies.
Tom
Something like ReiserFS doesn't scale well at the low end. On a 128M memory stick [iirc] the file system takes something like 16-20MB off the top. That's 15% of the drive gone instantly [compared to the 5M that FAT takes].
Granted if you have a 1GB memory stick the 20M or so that Reiser takes is less of a pain and the gains you get from the stability are more worth it...
Tom
I said my lawn was green. I did say it was because of one species of grass.
We got odd ball weeds and things here and there growing. The point is it's not just some yellow'ed out field with a chemlawn poster...
Tom
Correct saying would be "swing and a miss".
Tom
Not really, just tell your customer "you're not using it right, read the damn manual" and be done with.
... well too bad really. And in terms of "word of mouth" damage it's low.
What really makes companies that provide hardware look bad is when you use THEIR drivers and the device is still flaky and unreliable.
What hurts Microsoft is the low quality of the software they write [well depends, some tools are decent but as a whole they're pretty bad]. I've never called their techsupport, I have called their activation drone before though...
If a customer calls and says "I tried to l33t mod the device and it borked" and then gets upset when the dude on the other and says "STBU"
If any of my friends tried to "home brew upgrade" their motherboards or processors or something and it blew up I'd just laugh [at them] and lend them money for a new box in the meantime.
Tom
So why would they provide support for that? It's clearly not a documented use of the product. So they're not obligated to support users doing it.
Tom
Again you missed the point. The police officer can give you that ticket EVEN if the box ends up being defective.
What is at stake here is the right to mount a defense [in most cases you can't refuse or deny being given a ticket, but you can fight it in court and get it dismissed].
Citing economic reasons to not disclose how it works is stupid. That's the line of work they were getting into they should have known and worked around that from the start.
This is another "different value" market. The value [to the police] isn't so much in the proprietary DSP algorithms they use [or whatever] but in the assemblage. It's put together and tested to meet some minimal standards so the police don't have to. That's the value.
The fact that it uses publicly available algorithms and designs doesn't take away from the value.
"I realize Slashdot is the land of the tinfoil hats, but stop for a second and realize this isn't a story about police corruption. It's about criminals who were obviously felt to be intoxicated by a trained officer of the law getting away with it because of a sleazy loophole."
Criminals? Says who? You're INNOCENT until PROVEN guilty. And that loophole is the american justice system.
Suppose you just finished your PhD, you got a nice job in the government and part of that was a mandatory drug test. So you pee into the cup, admire it's bounty and pass it off. In the lab they accidently cross-contaminate your urine sample with that of an inmate in a local prison. Now they bring it back and it turns out your a drug addict. Well kiss your government job goodbye. You now also have a drug offense on your permanent record.
Oh, but you should [by your logic] have no right to question in which manner your "presumed guilt" was determined. So now you vested 150k into an education that will basically make you the most intelligent burger flipper in the world.
Congradulations, civil rights are just a "sleezy loophole".
Tom
The problem is brushing the proverbial mosquito off your arm also knocks off um ... fuck these metaphors...
You piss off customers by making flaky non-portable drivers. ATI for instance doesn't really share the user base in Linux as nvidia does and it's solely because their drivers suck.
But think about it this way, if the company thinks their TOP OF THE LINE product can be easily replicated today using part-time volunteers spread out across the globe... of how much value is their product anyways?
Tom
This premise is proven false time and time again.
... well ... um you got issues.
... to me the extra 200$ is worth it since I plan to own the processor for several years the average cost compared to the time savings is huge...
Look at the x86 CPU [or even ARM]. Look at GPUs, etc..
There isn't a lot of competition in high end products because just knowing how to "jigger the doohickey" isn't enough to figure out how the doohickey works.
If your trying to sell something like a 16550 UART in 2005 as a standalone IC
Complex enough products [to which the value is greater because the ability to solve it on ones own is lower] are hard to mimic to a level of similarly demanded by the customer.
Look at Transmeta and VIA. They're both x86 processors [well sorta] that did little to dent the x86 world [of which AMD which produce awesome cpus is even a small part of].
It's not enough that you're functionality equivalent you have to be just as efficient and cost effective.
I mean, I could buy that 100$ slow as shit Transmeta processor or I could buy that 200$ very fast and power efficient AMD64 processor
So no, I don't think companies revealing the interface reveal their meaningful valuable secrets. It's just a matter of controlling how and when and where the user uses the device.
Tom
This, ladies and gentlemen is why I own a LG superdrive.
...HARDWARE... manufacturers realize that ... the better.
I put it in my AMD64 Gentoo based box, booted up and it "just worked". No drivers, no special CD burning software [outside of cdrecord and growisofs], etc...
There is no value for me in commercial CDR tools since free [and decently working ones] exist already. The sooner
Tom
The "worth" of the card in that case factors into several camps...
...] interface. You'd have to spend time and energy changing it to work with a different piece of hardware...
1. Is it efficient, can I get high FPS without killing the CPU?
2. Does it look good? Is there quality?
3. Does it work reliably? Can I use this card to develop on and trust that my customer sees the same thing?
#1 depends on the hardware
#2 and #3 depend moreso on the software.
Even still it's a 3D graphics library that works only with the nVidia [or ATI or
That said..
What's stopping giving out the INTERFACE and not the software side of things? You still have to BUY the card. Software alone can't emulate that.
Think of the value proposition in terms of things you can't easily provide yourself. What does nvidia do that makes me want to give them money?
It certainly isn't writing drivers because we have competent people at xorg-x11 to do that. What they do that others [in the OSS world] cannot do is produce a GPU capable of pushing polygons.
In otherwords I don't want to pay nvidia to write drivers. I want to pay them to design efficient hardware that makes the nippomatics in ut2k4 even more realistic.
BTW, do you pay Intel to write an OS for you to use with your Intel processor? Did you pay sony money to produce music to listen with your sony CD player? Did you pay Maytag to make the dishes you wash with your dishwasher and did you pay Ford for the roads you want to drive your ford truck on?
Tom
You could just not document things that are not 100%?
Or am I missing something?
I mean I understand that products are usually evolutionary [e.g. the current model may have the beginnings of stuff that isn't ready yet but will be in the "next model"] but just don't document them.
Tom
I dunno about you but pretty much every piece of hardware from CD drive to network card, etc.. has had some piece of windows software which I promptly tossed in the garbage. I didn't buy the drivers [cuz I run hardware that I know works in Linux] I bought the hardware.
If the hardware didn't work out of the box with the Linux kernel the hardware has ZERO value to ME period. So if they want to push the hardware they have to provide the drivers [or specs or both].
The drivers have zero commercial value since they can only be used with that hardware. It isn't as if someone can copy it and use it for their software [unless they mimic the interface].
Again, the hobby groups of the 70s/80s/90s haven't disappeared. You don't even need to spend time on writing drivers if you just give people the specs.
Here's how you do it.
1. Make 4 cards [of whatever].
2. Email the lkml and offer them to people willing to write drivers.
3. Send them the card and the specs and tell them they can keep the card if they write a working driver under the GPL.
4. Take the driver and bundle it with said hardware [or get it merged into the kernel].
5. Profit!
All this costs you is 4 copies of your product [for a 50$ network card this is a mere 200$], postage and a copies of the interface specification.
If the hardware is worth anything the developers have incentive since they get cool hardware for essentially free. So they have incentive to participate...
Tom
The problem I have with GPU manufacters like ATI and nvidia is ...
... It's not like they're really that different technologies...
Don't you think they both understand what a pixel pipeline is? Or what a vertex shader is? Or
Also since when does interface dictate implementation? I mean look at AMD and Intel. Both implement x86 processors, both give out cycle timings, opcode formats, etc... I don't see cheap AMD clones on the market today...
I don't know how a GPU interface works exactly but chances are it has some uniform data format for various things like vertex lists, textures, vertex programs, etc. None of which explains how the GPU pipeline was implemented in silicon...
Again it's just more "our drivers are worth $$$"....
Imagine how crippled the x86 world would be if you had to use Intels compiler and were never told what the opcodes are or how they worked... We would have an exponentially smaller scale computer world I'd say. Specially since the hobby driven folk of the 80s and 90s are what really made the explosion possible.
Tom
Specially since you buy the DRIVE not the software. The software is the COST of doing business. The problem is you have ignorant marketting and investors who think "everything has commercial value".
.doc file and give it out.
Look at Broadcom. They hold their hardware specs a closely guarded secret [for the most part] and the net affect is you can only use their wifi stuff [reliably] in windows... The problem is without the drivers the hardware has zero customer value. But giving out free drivers lets you SELL hardware since it now has value.
The sad thing it isn't even that you have to write the damn drivers. In the OSS world of BSD/Linux the kernel contributors would GLADLY write a driver for free if it meant they could use some quality hardware with the respective OS. So all it costs the hardware manufacturer is describing the interface [at the high level] of how to talk to the hardware. Since these documents are ROUTINELY produced internally so the software teams can write their windows drivers all it means is you re-brand the
Tom
You do realize that just living near the pesticides is enough to get it into your body right?
You and your chemlawn using motherfucking yuppy friends are why I hate society. You get some idiotic believe that something that totally rapes and murders plants magically will have no effect on humans and if you can't immediately see it you must not be exposed to it.
The minute people realize that everything, and I mean EVERYTHING we do is interelated the better.
That and my front and back lawn are decently healthy and we do nothing but spray water on it and cut it regularly. Sure we're not "lawnman" green but it's still a healthy lawn. Most of my neighbours that get their lawns sprayed often neglect their lawns and they yellow half way through the summer...
Hypocrites...
Tom
That's a stupid argument because it has no bearing on the conversation. What, are you 8 years old or something?
Knowing how a jet fighter works has no bearing on whether a little box that says "guilty" is actually working properly. And it certainly has nothing to say about my right to question how and if the box actually works in the first place.
Why is "how" important. It could have false positives [e.g. medicines may trigger it for instance] that make that particular test invalid. It's happened before and I'd be surprised if it never comes up again.
Why is "if" important? It may be well outside acceptable error tolerances and the results are not significant or useful in a court of law.
You guys keep saying "getting off"... Prove to me he was actually DUI. That's what this "little box" is supposed to do. "Getting off" implies he's actually guilty and was not convicted anyways. You think getting arrested is the same as "being guilty" because you have no appreciation for what the law actually says and stands for.
Just wait till your on the wrong end of private corporations wet dream and see how much sympathy you get from your peers who will assume your guilty without having been given a fair chance to defend yourself. Of course you can't imagine this because you're an 8 year old brat skipping second period in school to post on slashdot... Go to class you little brat!
Tom
Nobody [certainly not myself] are saying that drinking and driving is a good thing. What I am trying to say is at least have your ducks in a row.
... etc ...
... all before he's been able to even question if the little device that has stung him so actually works within the expected margins of error.
What you call "getting off on technicalities" I call upholding your rights.
I mean having a lawyer present is just a technicality too right? If you're guilty why should you have a defense? That just means you can get "off" with the crime!
Questioning witnesses? Why? They're under oath therefore all they speak is verbatim truth.
Yeah it would be nice if these drunk drivers just fessed up and did their time. That doesn't mean they don't have a right to question the way they were deemed to have violated the law.
You seemed to be along the line of "if you were caught you must be guilty". As if law enforcement really deserve that unquestionability status.
I say just wait till someone cries rape in your direction and see how much fun the "guilty before innocent" status you're so desperately trying to stick on others feels on yourself.
I mean suppose the guy really wasn't drunk and he was held for a day in jail. He'd miss the next day of work, have to explain to his boss, have the social stigma of being known for being arrested DUI
Tom
mmm, you can't question if something works unless you know how it works.
For instance, maybe the radar gun works different at night? [making that up but you never know]. If you didn't know how a radar actually worked you wouldn't know to test [or demand that be tested].
Also it's the right of the defense to perform independent testing [just like they can question witnesses...].
I don't see what the big deal is. As a company providing law enforcement equipment/software/etc you should have known WELL in advance that you'd be called upon to explain how it works.
If you're business model depends on selling law enforcement tools that cannot withstand scrutiny... you SHOULD be out of business.
Tom
I picked it at random because IANAB ;-)
Say something as simple as distilled water... whatever.
The point is we're all humans. When a mistake is made you either throw it out or keep it. In both cases you account for the mistake. It could be you did something that wouldn't affect [or at least you reasonably believe it won't affect] the net result. I mean, say the time for culturing bacteria [of a certain type] has been determined to be optimally at 37C for exactly 37 hours 43 minutes 29 seconds.
Chances are if you're at 37.1C for 37 hours 43 minutes 21 seconds the result won't be damaged enough [significantly] as to require it to be dismissed. Yet I imagine this happens all the time.
The real threat portrayed in this story though is that there is zero accountability. If you didn't have to account for how you came up with the results what's saying you didn't just make up the outcome?
Tom
"They are guilty. You know it, and I know it. It seems that they are not disputing the fact that they had been drinking. They are focusing on the way they got caught."
... PROVEN ... guilty. Without scrutinizing the box you can't prove guilt.
You seem to have no clue what the legal process is. Your INNOCENT until
There is no more discussion worth having on this point.
And if you call upholding your rights "whining" then again, you're living in a bubble that has yet to be pierced...
Tom