Does anyone understand the methodology behind overclocking, or is it just trial and error? Any recommendations?
My recommendation is not to bother trying. XScale and a lot of other embedded processors are highly integrated units - all the stuff in a PC that you plug in cards for is in the chip (audio, display, etc.)
To run all of these internal peripherals, there are internal registers that divide the clock down. If you change the crystal, you screw up the clock for all these internal gadgets, and they probably won't work at all. No audio, no display, no pcmcia...it would all glitch.
Can't help but notice that you're ".Bruce Perens", not "Bruce Perens". You might not be Bruce Prime. But I'll respond here anyways.
There are measures to protect again this sort of abuse.
I didn't realize that it was abuse to want to keep your code yours. I'm not intending to abuse anyone. Here's my problem - upper mgt. is ok with using Linux, but doesn't want to release IP. That's it. If we have to release IP, then the whole project switches to Windows CE. And having written for 3 years for Windows CE, I can honestly say I'd rather endure an all-night root canal.
I would advice you to be very careful with your assumption that drivers are non-GPL. It can be done, but it isn't my postion to tell you how to do it today.
If it's so difficult, what are MODULE_LICENSE and EXPORT_SYMBOL_GPL for?
I think his terminology was incorrect. I think what he was saying is that it's a stock linux kernel with non-GPL'd kernel drivers.
Yes, that's exactly what I had in mind - thanks. Take a stock kernel, write your own drivers specific to your widget, and release a binary.
In this case the kernel would be GPL, but stock. You can get it anywhere. Your drivers would be non-GPL. They're your own business. Aren't they?
I'm asking because I work in the embedded field, and recently talked my engineering team into Linux for our next target. And this is the development path we intend to follow. We aren't modifying anything, we're using stock kernels and adding our drivers in.
I hope Bruce keeps reading this thread. I'm not trolling - honest! I'd really like to not wind up on "the list".
Seriously, the GPL only applies to derivitave works, right? How do they know it's not a stock kernel with no changes and hand-written non-derivative drivers added on to it?
Well, ISPs could block a known port or range of ports. IIRC, gnutella uses 6346 as a default. Block all traffic on 6346 and that'd stop gnutella traffic for the most part.
And users could get past that by changing the default port. I'm not on gnutella but I'm pretty sure that's possible. If it isn't currently, it could be with a quick patch.
Then, ISPs could sniff traffic and look for mp3-ish content. And block that.
And the next gen file swapper would simply encrypt packets, making sniffing computationally unfeasable.
So the short answer is no, ISPs could not enforce this. They could throw up roadblocks, but they would eventually just be speedbumps.
Well, I'm not really commenting on the case, I'm commenting on their business tactic. The case could be 100% valid and I'd still say it's a stupid move. Their latest SEC filing would seem to bear that out.
That being said, I also think the case is stupid. If their case is so rock solid, why not tell everyone what the offending code is? Why hide it, if it's so rock-solid? The truth should be able to stand up to any examination, and I believe that's what they fear. Hence, no disclosure. Gambling your entire business on a bluff, is...stupid.
There's no mention of SCO flushing themselves down the toilet. That's classic stupidity. "Let's not make product anymore, let's make lawsuits instead."
It'll make an interesting read once it's all over.
Really? Thanks - that's good to know. I wasn't sure if the programmer guy was passing good info. I always get suspicious when spammers are involved, and that includes people who work for 'em.
Still though - I wouldn't recommend everyone go sign up for g'e'ner!c vi-agr-a and try the chargeback trick. All you're likely to get for your trouble is a few bogus charges, a bottle of sawdust shaped like pills and no refund - and that's if you're lucky.
Absolutely the best post in this whole thread. Bravo.
The need to process credit cards is the weak link in much of the spam business, and it is very hard for them to work around an inability to obtain the services of a merchant credit card account.
Let's look this post a bit and do a little translation:
Part of my companies' income is from sales of various and sundry products sold via soley online "stores." Part of that traffic is via banner ads, text links, etc, and another portion is via bulk mail (spam)
Translation: I am a spammer.
If you really want to hurt a spammer, get thousands of people to order a product, then send it back and charge-back the order on their cards.
Translation: Give me your credit card number.
Spammers are the wise guys and con men of the digital age. DO NOT TRUST THEM. I mean really - if this guy makes his living this way is he honestly going to give you a stick to beat him with???
It's more likely he'll take your credit card number, charge it to the hilt and take off to Zaire.
Give me your credit card number and I'll be hurt. Please!
I wouldn't worry too much about *eraldo's good name. When I think of journalistic integrity, he's not the first guy to spring to mind. I know I'm not the only guy who was rooting for the skinhead that plowed him with a chair.
But since you're so worried about it - I hereby promise to submit all future postings with *eraldo's name in it to the New York Times for proofreading before I hit {Submit}, so as not to offend the delicate sensibilities of all *eraldo fans who for some reason also read Slashdot.
You forget, this is America, where the dollar reigns supreme. Not only will they dig them up, they'll broadcast it live with Heraldo Riviera doing the commentary. And I'll betcha there will be at least 2 made-for-tv movies out of this before it's all over.
This is nothing surprising. It's SCO's standard delay-as-long-as-we-can tactic. SCO knows that as soon as they actually have to offer up their proof that copied code is in Linux, it won't stand up to analysis, their case will be rejected, and their stock will drop like a stone. It'll be game over for SCO as a company. Their current business model depends on not offering any proof of their claims.
So why not subpoena everyone, to make things as slow and difficult as possible? I'm surprised Elvis and Bigfoot aren't on the list.
Kinda reminds me of the negotiations at the end of the Korean war. Every last detail of how the talks were to proceed were argued to death before the talks could begin. There were even provisions as to which direction the delegates sat and how high their chairs were in relation to each other before they'd talk. And the reason was, was that the delegates simply didn't want to be there. Same for SCO.
Ah, thanks. I skimmed the article and didn't get it either. And I'm a BSEE. =)
That being said, it sounds like a pretty decent idea. But since you're actually in the know, any idea why they'd want to use oscillators in the first place?
My first thought would be to store the static charge in some kind of switched capacitor type thing - like maybe a single bit of DRAM.
If banner ads fail, more and more sites will be forced into a pay model, and the days of the "Free Internet" will be almost over.
First of all, when the internet started it was free. Advertising sleazed in years after the whole structure was in place. The internet didn't need it then, and it doesn't need it now. I know my inbox sure doesn't need any more "advertising".
Second, not everybody uses Norton Antivirus. Or even Windows for that matter. And not only Norton blocks popups. You can do that yourself with a simple hosts file.
So to sum up - no, Norton Antivirus will not destroy the internet.
Considering that Microsoft says it is focused on security, the comments from the Chief Software Architect aren't inspiring... What kind of message does that send to the developers who work for Gates?
The same message M$ has always sent.
"The left hand does not know what the right hand is doing, so once again you're on your own, folks."
but if I write a piece of software (or come up with a really good idea) and decide to patent it, I should expect (and recieve) fair price from anyone who decides to use it. Period.
I agree. But only if you invent it yourself. The patent in question covers plug-in technology. It's old news, there's a lot of prior art, and the patent holder certainly didn't invent it himself. This is why it's an abusive patent, and will be a good thing when it goes away.
Allow me to offer a counterexample. What if I were to get a patent granted for posting anonymously on the internet? I didn't invent it, but you'd owe me a nickel.
I almost never click the "Patent Pending" Slashdot links, because they are almost always depressing. Nice to see a change of pace!
I really hope this wins. I get so sick and tired of people filing abusive patents to strangle the whole world just so they can make a quick buck. My company is currently fighting something similar, so it's a pet peeve of mine.
If W3C wins this, they should keep the momentum going and go after Jeff Bezos. He's a prime example of this kind of abuse.
Normally I wouldn't respond, but you don't sound like you're trolling so why not? AC, you shall have your answers.
So, Weaselmancer my friend, are you saying that it takes a certain psychological makeup to avoid becoming addicted to cocaine?
Nope. What I'm saying is that some kinds of people are more susceptible to addiction than others. Any twelve stepper would tell you the same. And this guy sounds like an addictive personality. The whole "keep at it until it's fun again" is textbook. People who OD usually are thinking something similar.
What about crack?
The author shouldn't do crack, either.:^)
Are you speaking from experience, or just trying to make a joke?
I was modded 70% funny, 30% insightful and that reflects what I was feeling when posted pretty accurately.
And I am not speaking from direct experience - I've never done the coke/crack thing. But a lot of my friends growing up did. Oddly enough, most of them now (you guessed it) play EQ. Compulsively. They think nothing of "camping" some imaginary monster for 2 weeks at a time just to get the Toenail of Yendor that gives you +1 to your knitting skill. It boggles me.
But you're a trooper. You stiffen that upper lip and press on, certain that if you can only hang in there the good times will arrive and the game will be FUN again.
They've switched to saying that only fortune 1000 companies can get a license in order to stop people from trying to get one.
That's a really interesting idea.
So what you're saying is that the whole SCO fiasco boils down to stock inflation only. They know their claim is bogus - and selling a license would be a different offense (fraud), with a higher penalty. They're unwilling to risk real jail time over this, so they don't sell any licenses to anyone on purpose, to maintain the scam as long as possible while they cash out.
I wish I could split my mod points with you - good call.
Makes sense to me. SCO is playing the lottery here, and hoping one of the tickets is a winner.
Why go after Joe Consumer? SCO knows their odds of even finding private citizens using Linux are next to zero. Private citizens hardly ever get busted using a pirated copy of Windows, and Redmond has cash to burn to go looking for them. And even if they were to nail a few guys, so what? They're looking for a big payoff here, not nickel-and-dime end users.
But Fortune 1000 companies, ah! Big bucks. They're hoping to intimidate some huge company with the threat of audits and huge legal expenses vs. the relatively low cost of a site license.
And they know they only have a limited time to try their horseshit before some judge somewhere finally makes them show the "you can't see it yet" infringing code, and that'll be game over when it happens. So they're in a hurry - no time for small potatoes.
So please, don't bother SCO unless you have obscene piles of cash lying around and a panicky board of directors!
Slashdot Mirror
Does anyone understand the methodology behind overclocking, or is it just trial and error? Any recommendations?
My recommendation is not to bother trying. XScale and a lot of other embedded processors are highly integrated units - all the stuff in a PC that you plug in cards for is in the chip (audio, display, etc.)
To run all of these internal peripherals, there are internal registers that divide the clock down. If you change the crystal, you screw up the clock for all these internal gadgets, and they probably won't work at all. No audio, no display, no pcmcia...it would all glitch.
Weaselmancer
Ah! I got it. That makes perfect sense. And it's something I might have forgotten when the time comes. Thanks for the help.
Weaselmancer
Can't help but notice that you're ".Bruce Perens", not "Bruce Perens". You might not be Bruce Prime. But I'll respond here anyways.
There are measures to protect again this sort of abuse.
I didn't realize that it was abuse to want to keep your code yours. I'm not intending to abuse anyone. Here's my problem - upper mgt. is ok with using Linux, but doesn't want to release IP. That's it. If we have to release IP, then the whole project switches to Windows CE. And having written for 3 years for Windows CE, I can honestly say I'd rather endure an all-night root canal.
I would advice you to be very careful with your assumption that drivers are non-GPL. It can be done, but it isn't my postion to tell you how to do it today.
If it's so difficult, what are MODULE_LICENSE and EXPORT_SYMBOL_GPL for?
Weaselmancer
I think his terminology was incorrect. I think what he was saying is that it's a stock linux kernel with non-GPL'd kernel drivers.
Yes, that's exactly what I had in mind - thanks. Take a stock kernel, write your own drivers specific to your widget, and release a binary.
In this case the kernel would be GPL, but stock. You can get it anywhere. Your drivers would be non-GPL. They're your own business. Aren't they?
I'm asking because I work in the embedded field, and recently talked my engineering team into Linux for our next target. And this is the development path we intend to follow. We aren't modifying anything, we're using stock kernels and adding our drivers in.
I hope Bruce keeps reading this thread. I'm not trolling - honest! I'd really like to not wind up on "the list".
Weaselmancer
Seriously, the GPL only applies to derivitave works, right? How do they know it's not a stock kernel with no changes and hand-written non-derivative drivers added on to it?
Maybe they are GPL compliant.
Weaselmancer
Well, ISPs could block a known port or range of ports. IIRC, gnutella uses 6346 as a default. Block all traffic on 6346 and that'd stop gnutella traffic for the most part.
And users could get past that by changing the default port. I'm not on gnutella but I'm pretty sure that's possible. If it isn't currently, it could be with a quick patch.
Then, ISPs could sniff traffic and look for mp3-ish content. And block that.
And the next gen file swapper would simply encrypt packets, making sniffing computationally unfeasable.
So the short answer is no, ISPs could not enforce this. They could throw up roadblocks, but they would eventually just be speedbumps.
Weaselmancer
Well, I'm not really commenting on the case, I'm commenting on their business tactic. The case could be 100% valid and I'd still say it's a stupid move. Their latest SEC filing would seem to bear that out.
That being said, I also think the case is stupid. If their case is so rock solid, why not tell everyone what the offending code is? Why hide it, if it's so rock-solid? The truth should be able to stand up to any examination, and I believe that's what they fear. Hence, no disclosure. Gambling your entire business on a bluff, is...stupid.
There's no mention of SCO flushing themselves down the toilet. That's classic stupidity. "Let's not make product anymore, let's make lawsuits instead."
It'll make an interesting read once it's all over.
Weaselmancer
Too bad you posted as an AC, that was a good post.
If it takes $30k to get the law involved, then that needs to be changed too. Maybe we could add 3 feds to the 3 lawyer/3 geek team.
Weaselmancer
Really? Thanks - that's good to know. I wasn't sure if the programmer guy was passing good info. I always get suspicious when spammers are involved, and that includes people who work for 'em.
Still though - I wouldn't recommend everyone go sign up for g'e'ner!c vi-agr-a and try the chargeback trick. All you're likely to get for your trouble is a few bogus charges, a bottle of sawdust shaped like pills and no refund - and that's if you're lucky.
Weaselmancer
Absolutely the best post in this whole thread. Bravo.
The need to process credit cards is the weak link in much of the spam business, and it is very hard for them to work around an inability to obtain the services of a merchant credit card account.
Let's look this post a bit and do a little translation:
Part of my companies' income is from sales of various and sundry products sold via soley online "stores." Part of that traffic is via banner ads, text links, etc, and another portion is via bulk mail (spam)
Translation: I am a spammer.
If you really want to hurt a spammer, get thousands of people to order a product, then send it back and charge-back the order on their cards.
Translation: Give me your credit card number.
Spammers are the wise guys and con men of the digital age. DO NOT TRUST THEM. I mean really - if this guy makes his living this way is he honestly going to give you a stick to beat him with???
It's more likely he'll take your credit card number, charge it to the hilt and take off to Zaire.
Give me your credit card number and I'll be hurt. Please!
I wouldn't worry too much about *eraldo's good name. When I think of journalistic integrity, he's not the first guy to spring to mind. I know I'm not the only guy who was rooting for the skinhead that plowed him with a chair.
But since you're so worried about it - I hereby promise to submit all future postings with *eraldo's name in it to the New York Times for proofreading before I hit {Submit}, so as not to offend the delicate sensibilities of all *eraldo fans who for some reason also read Slashdot.
Weaselmancer
You forget, this is America, where the dollar reigns supreme. Not only will they dig them up, they'll broadcast it live with Heraldo Riviera doing the commentary. And I'll betcha there will be at least 2 made-for-tv movies out of this before it's all over.
All that being said, you're exactly right.
Weaselmancer
This is nothing surprising. It's SCO's standard delay-as-long-as-we-can tactic. SCO knows that as soon as they actually have to offer up their proof that copied code is in Linux, it won't stand up to analysis, their case will be rejected, and their stock will drop like a stone. It'll be game over for SCO as a company. Their current business model depends on not offering any proof of their claims.
So why not subpoena everyone, to make things as slow and difficult as possible? I'm surprised Elvis and Bigfoot aren't on the list.
Kinda reminds me of the negotiations at the end of the Korean war. Every last detail of how the talks were to proceed were argued to death before the talks could begin. There were even provisions as to which direction the delegates sat and how high their chairs were in relation to each other before they'd talk. And the reason was, was that the delegates simply didn't want to be there. Same for SCO.
Weaselmancer
Ah, thanks. I skimmed the article and didn't get it either. And I'm a BSEE. =)
That being said, it sounds like a pretty decent idea. But since you're actually in the know, any idea why they'd want to use oscillators in the first place?
My first thought would be to store the static charge in some kind of switched capacitor type thing - like maybe a single bit of DRAM.
Weaselmancer
If banner ads fail, more and more sites will be forced into a pay model, and the days of the "Free Internet" will be almost over.
First of all, when the internet started it was free. Advertising sleazed in years after the whole structure was in place. The internet didn't need it then, and it doesn't need it now. I know my inbox sure doesn't need any more "advertising".
Second, not everybody uses Norton Antivirus. Or even Windows for that matter. And not only Norton blocks popups. You can do that yourself with a simple hosts file.
So to sum up - no, Norton Antivirus will not destroy the internet.
Weaselmancer
Considering that Microsoft says it is focused on security, the comments from the Chief Software Architect aren't inspiring... What kind of message does that send to the developers who work for Gates?
The same message M$ has always sent.
"The left hand does not know what the right hand is doing, so once again you're on your own, folks."
Weaselmancer
AC, or Jess,
but if I write a piece of software (or come up with a really good idea) and decide to patent it, I should expect (and recieve) fair price from anyone who decides to use it. Period.
I agree. But only if you invent it yourself. The patent in question covers plug-in technology. It's old news, there's a lot of prior art, and the patent holder certainly didn't invent it himself. This is why it's an abusive patent, and will be a good thing when it goes away.
Allow me to offer a counterexample. What if I were to get a patent granted for posting anonymously on the internet? I didn't invent it, but you'd owe me a nickel.
Weaselmancer
I almost never click the "Patent Pending" Slashdot links, because they are almost always depressing. Nice to see a change of pace!
I really hope this wins. I get so sick and tired of people filing abusive patents to strangle the whole world just so they can make a quick buck. My company is currently fighting something similar, so it's a pet peeve of mine.
If W3C wins this, they should keep the momentum going and go after Jeff Bezos. He's a prime example of this kind of abuse.
Weaselmancer
Brilliant. Very well put.
Weaselmancer
I'm an AC, and I want answers.
Normally I wouldn't respond, but you don't sound like you're trolling so why not? AC, you shall have your answers.
So, Weaselmancer my friend, are you saying that it takes a certain psychological makeup to avoid becoming addicted to cocaine?
Nope. What I'm saying is that some kinds of people are more susceptible to addiction than others. Any twelve stepper would tell you the same. And this guy sounds like an addictive personality. The whole "keep at it until it's fun again" is textbook. People who OD usually are thinking something similar.
What about crack?
The author shouldn't do crack, either. :^)
Are you speaking from experience, or just trying to make a joke?
I was modded 70% funny, 30% insightful and that reflects what I was feeling when posted pretty accurately.
And I am not speaking from direct experience - I've never done the coke/crack thing. But a lot of my friends growing up did. Oddly enough, most of them now (you guessed it) play EQ. Compulsively. They think nothing of "camping" some imaginary monster for 2 weeks at a time just to get the Toenail of Yendor that gives you +1 to your knitting skill. It boggles me.
Weaselmancer
But you're a trooper. You stiffen that upper lip and press on, certain that if you can only hang in there the good times will arrive and the game will be FUN again.
You, sir, should avoid cocaine at all costs.
Weaselmancer
They've switched to saying that only fortune 1000 companies can get a license in order to stop people from trying to get one.
That's a really interesting idea.
So what you're saying is that the whole SCO fiasco boils down to stock inflation only. They know their claim is bogus - and selling a license would be a different offense (fraud), with a higher penalty. They're unwilling to risk real jail time over this, so they don't sell any licenses to anyone on purpose, to maintain the scam as long as possible while they cash out.
I wish I could split my mod points with you - good call.
Weaselmancer
Makes sense to me. SCO is playing the lottery here, and hoping one of the tickets is a winner.
Why go after Joe Consumer? SCO knows their odds of even finding private citizens using Linux are next to zero. Private citizens hardly ever get busted using a pirated copy of Windows, and Redmond has cash to burn to go looking for them. And even if they were to nail a few guys, so what? They're looking for a big payoff here, not nickel-and-dime end users.
But Fortune 1000 companies, ah! Big bucks. They're hoping to intimidate some huge company with the threat of audits and huge legal expenses vs. the relatively low cost of a site license.
And they know they only have a limited time to try their horseshit before some judge somewhere finally makes them show the "you can't see it yet" infringing code, and that'll be game over when it happens. So they're in a hurry - no time for small potatoes.
So please, don't bother SCO unless you have obscene piles of cash lying around and a panicky board of directors!
Weaselmancer