Don't you mean inter-ISP and international? I don't mean to be nit-picky, but intra means the opposite of inter and as such dramatically changes the meaning of your sentence.
Computing on demand is destined for failure. The only way it will fly is for it to be forced down our throats, which is pretty hard to fathom considering the cheapness of off-the-shelf computing resources.
Today, you can pick up a cheap server for $500. You want to double your capacity, spend another $500. As long as those machines meet your needs, there is no need to put out more money. Say they last 3 years -- $1000/5 = $200 per year. I haven't seen prices for computing on demand, but it pretty much _has_ to be more expensive due to the "on demand" system overhead and complexity. Not to mention that the vendor providing the computing resources has a huge overhead expense just keeping the latest hardware (and server space) available for use at a moment's notice. How does that company increase available bandwidth or processor resources if several companies need their services at once?
The only thing I can see on-demand being a good thing is if your computing needs are very high for a short time and then go back to normal levels. However, companies already have complete packages like this available for purchase (eg. distributed download sites, P2P computation clients).
Between P2P and Spam, I'm suprised we have any bandwidth left!
I find it hard to believe P2P is using "as much as" that much of the total bandwidth. What is the average, not the peak usage? "As much as" implies a maximum throughput which is not sustained.
My first question is what use is acceptible for the number one spot in bandwidth usage? What about CD image downloads (650MB each), porn (ISPs are probably too embarassed to mention what % this is), forwarded emails with attachments, search engine spiders, and so on?
Next, if the P2P bandwidth carried 95% legal content, would there be an issue here if a peak of 60% bandwidth was used? Is this really about the bandwidth?
Are those who share their files via P2P really bandwidth hogs, or are those who download the files the bandwidth hogs? Merely providing the files for download would produce zero bandwidth (aside from protocol overhead) otherwise.
Near the bottom of the article, they say that intra-ISP and intra-country bandwidth is the most expensive and is what must be kept under control. So what brings us all together should be regulated? Maybe they don't like how free the Internet is, unless portraying freedom and unlimited access helps them sell more services through their commercials.
Standards are great, but they have to be implemented to be useful. Implemented standards can make or break a project. You can take all of these commodity protocols, data formats, etc.. and use them without royalty in your project to make it a success. This saves a lot of reinventing the wheel with less thought-out designs and implementations.
Every Matrix fan will go and see the movie and/or wait for it on DVD. The pirated copy basically allows someone to preview it to be sure it's not a piece of crap (which I'm sure it isn't), so I wouldn't get my knickers in a knot over this.
Welcome to the USA, where common sense is absolutely irrelevant. Got a sensational case? There's a lawyer and a judge out there somewhere who'll see to it that you win. Disgusting.
The old myth that floating point operations are slower than integer is wrong too. We tested that one 5 years ago and a floating point operation of the same values often beat the integer operation.
Very interesting. I would think that since floating point storage is greater than integer storage (at least on 32-bit PCs), the speed would automatically be better for integers. Can you give more information?
Actually, exceptions can impact performance, if you are throwing the exception the majority of the time. If it is more the exception than the rule that the exception is thrown, your performance will not be affected.
A lot of them are things that actually used to be good advice, but for some reason or another (changes in hardware, compilers, etc.) aren't anymore.
I remember a case with Wolfenstein 3D, IIRC. There was a note in the source or readme saying that they did something a certain (nonstandard) way because of how the cache worked on 486-class machines. In Pentium machines, it would actually be counterproductive. I might be wrong on the details, but it was an interesting note.
This is why so many intranet and internet applications seem slow. People put-off worrying about performance until the last step (just like they are told to). And then it might be too late.
Good point. At the end when everything is working fine, you manager might say something like "everything is working just fine. Why do you want to spend the next two weeks on performance issues? I have more higher priority things for you to do".
Get rid of the duplicates, or at least keep no more than two of a kind (hey, it worked for Noah). The magazines and crap can either be recycled or take the choice ones only (and scan them in).
Once you have the collection down to a more manageable size, then ask for help. Storing loads of junk at someone else's expense is a little much to ask.
Or, have a yard sale and give the shit away. At least _someone_ might enjoy it. A Beowulf cluster of junk collectors, if you will. Cost: zero.
Would someone in an NEV have to purchase insurance? What if someone in a huge Chevy Tahoe or Ford Explorer mowed them down because they didn't see them or couldn't stop in time? Sounds really messy. Major liability because you have basically no protection AND you move at such slow speed. No thanks!
Here's a niche for P2P software... obtaining data from a website in a distributed way so as not to stick out in the website logs the way a one or a few download clients would. The collected data could be processed remotely or uploaded to a central server.
Finally, if the stolen code is so bloody obvious then why not show even one example of where there is direct copying. This wouldn't affect their legal strategy one bit (despite claims to the otherwise) and would grant them so much more credibility.
Yes, SCO's case has FUD written all over it. Why should they not try and settle out of court? Remove the offending code and have someone clean-room it, and all is well again. Just like you said - why not even show any code but rather threaten half the world with law suits? It makes absolutely NO sense.
My guess is they either want a buyout or they have something to gain by fear, uncertanty and doubt about Linux.
Let's say SCO is right and some code was copied. That leaves 99.99% of Linux in the clear. That's why I'm not concerned about this lawsuit at all. As soon as we find out what code is covered, we rewrite it. And SCO is left behind with millions of wasted legal fees and a destroyed reputation.
I think before I spent a significant amount of cash on a project like this I would like to be assured that the government wouldn't shut it down for some reason. Cool things often are, it seems. And if not the government, then hopefully not a stupid company and its lawyers who would come running and say (short of breath) "hey! that's miiine! I THOUGHT OF IT FIRST GODDAMM *cough cough* IIITTT!"
Why would I want a $2 coin? It would just increase the inflation rate of everything sold in a vending machines (anything less than $2 will go for $2 and stuff that was $2.5 would go for $4). No thanks.
You're right, actually. And you forgot bus fair.:)
But I have to laugh at this story, because I've heard a number of Americans laugh at Canada's "monopoly money" over the years and now coloured money is coming to the US! Though I have to admit the new Canadian $10 bills are a little funky.
What kind of a world do we live in when the top priority is stopping spam from hitting our inboxes? What about social issues facing us every day? Cleaner air, water, nutritious food, mental health, affordable living, I could go on and on.
Where are these issues in our government's agenda? Spam will go away on its own - we have the tools to fight it, so legislation is not needed. Laws should not be a first resort.
Notice on just about every chart they cut it off before Microsoft Windows crapped out. You can plainly see that Red Hat is steadily providing good throughput, moving a little lower over time, while Microsoft's throughput is moving relatively quickly lower as the # of clients is increased.
Who could believe any Microsoft-sanctioned benchmark report given their track record?
I find it interesting how M$ used 64k clusters and left Red Hat at the default, when M$ default is 4K clusters. You'll note that you can't defragment a Windows volume with cluster size != 4KB out of the box.
I'd also be interested to know the CPU usage on each configuration while running these tests. Windows XP uses over 80% CPU to keep the pipes full, while Red Hat (and even Win2k) use ~ 10%.
Of course, you still can't argue about the price. How much does 160 client access licenses cost for Windows? "You can't beat free."
Everyone's all about freedom, but nobody seems to want to pay the price for it.
That's quite a blanket statement. Open source advocates pulled Mandrake's ass out of the gutter once when they needed it. Red Hat is doing just fine with their subscriptions... they don't even _need_ the OS revenue IMO.
Open source is open source. It's meant to be free as in speech AND free as in beer. Those trying to sell it realize that and take the associated risks.
I've given money on several occasions to open source causes, but I'm not going to shell out for every upgrade since that would just defeat much of the incentive to use open source.
What would you do if they decided to switch over to Microsoft? You'd all be screwed then.
No, "we" wouldn't be screwed. Red Hat != Open Source. Red Hat is a definite asset to the community, but it is not THE community.
I think turning to the law is the wrong place to look for a solution. Plenty of ideas have been thought of to limit spam, and more will come. Let the market decide. We have way too many laws as it is. In fact, shouldn't there come a point when we don't need NEW laws?
That's a big YEP. Eventually once all IT jobs are outsourced to those c*cks*ckers in India, we'll all be relics of a byegone era.
I think you're directing your anger at the wrong people. The people in India are just doing their jobs. The c*cks*ckers are the ones in America hiring them.
What I find funny is that outside workers being hired in the US is being clamped down on, while companies are still free to send work offshore. Huh?
Slashdot Mirror
You may find this story interesting.
Don't you mean inter-ISP and international? I don't mean to be nit-picky, but intra means the opposite of inter and as such dramatically changes the meaning of your sentence.
:)
You're right, that's what I meant. Thanks.
Computing on demand is destined for failure. The only way it will fly is for it to be forced down our throats, which is pretty hard to fathom considering the cheapness of off-the-shelf computing resources.
Today, you can pick up a cheap server for $500. You want to double your capacity, spend another $500. As long as those machines meet your needs, there is no need to put out more money. Say they last 3 years -- $1000/5 = $200 per year. I haven't seen prices for computing on demand, but it pretty much _has_ to be more expensive due to the "on demand" system overhead and complexity. Not to mention that the vendor providing the computing resources has a huge overhead expense just keeping the latest hardware (and server space) available for use at a moment's notice. How does that company increase available bandwidth or processor resources if several companies need their services at once?
The only thing I can see on-demand being a good thing is if your computing needs are very high for a short time and then go back to normal levels. However, companies already have complete packages like this available for purchase (eg. distributed download sites, P2P computation clients).
On an Athlon XP2100+ box, I can encode Vorbis files at 9x realtime speed. You could do realtime encoding on 1/9th this amount of 'box'.
And thus serve nine customers? They have millions of customers. Should they invest in a million boxes?
Between P2P and Spam, I'm suprised we have any bandwidth left!
I find it hard to believe P2P is using "as much as" that much of the total bandwidth. What is the average, not the peak usage? "As much as" implies a maximum throughput which is not sustained.
My first question is what use is acceptible for the number one spot in bandwidth usage? What about CD image downloads (650MB each), porn (ISPs are probably too embarassed to mention what % this is), forwarded emails with attachments, search engine spiders, and so on?
Next, if the P2P bandwidth carried 95% legal content, would there be an issue here if a peak of 60% bandwidth was used? Is this really about the bandwidth?
Are those who share their files via P2P really bandwidth hogs, or are those who download the files the bandwidth hogs? Merely providing the files for download would produce zero bandwidth (aside from protocol overhead) otherwise.
Near the bottom of the article, they say that intra-ISP and intra-country bandwidth is the most expensive and is what must be kept under control. So what brings us all together should be regulated? Maybe they don't like how free the Internet is, unless portraying freedom and unlimited access helps them sell more services through their commercials.
Standards are great, but they have to be implemented to be useful. Implemented standards can make or break a project. You can take all of these commodity protocols, data formats, etc.. and use them without royalty in your project to make it a success. This saves a lot of reinventing the wheel with less thought-out designs and implementations.
Every Matrix fan will go and see the movie and/or wait for it on DVD. The pirated copy basically allows someone to preview it to be sure it's not a piece of crap (which I'm sure it isn't), so I wouldn't get my knickers in a knot over this.
I wouldn't be surprised if this was a result of corporate America pressuring IEEE, namely big ISPs and content providers.
Welcome to the USA, where common sense is absolutely irrelevant. Got a sensational case? There's a lawyer and a judge out there somewhere who'll see to it that you win.
Disgusting.
The old myth that floating point operations are slower than integer is wrong too. We tested that one 5 years ago and a floating point operation of the same values often beat the integer operation.
Very interesting. I would think that since floating point storage is greater than integer storage (at least on 32-bit PCs), the speed would automatically be better for integers. Can you give more information?
Exceptions cripple performance.
Actually, exceptions can impact performance, if you are throwing the exception the majority of the time. If it is more the exception than the rule that the exception is thrown, your performance will not be affected.
A lot of them are things that actually used to be good advice, but for some reason or another (changes in hardware, compilers, etc.) aren't anymore.
I remember a case with Wolfenstein 3D, IIRC. There was a note in the source or readme saying that they did something a certain (nonstandard) way because of how the cache worked on 486-class machines. In Pentium machines, it would actually be counterproductive. I might be wrong on the details, but it was an interesting note.
This is why so many intranet and internet applications seem slow. People put-off worrying about performance until the last step (just like they are told to). And then it might be too late.
Good point. At the end when everything is working fine, you manager might say something like "everything is working just fine. Why do you want to spend the next two weeks on performance issues? I have more higher priority things for you to do".
Get rid of the duplicates, or at least keep no more than two of a kind (hey, it worked for Noah). The magazines and crap can either be recycled or take the choice ones only (and scan them in).
Once you have the collection down to a more manageable size, then ask for help. Storing loads of junk at someone else's expense is a little much to ask.
Or, have a yard sale and give the shit away. At least _someone_ might enjoy it. A Beowulf cluster of junk collectors, if you will. Cost: zero.
Would someone in an NEV have to purchase insurance? What if someone in a huge Chevy Tahoe or Ford Explorer mowed them down because they didn't see them or couldn't stop in time? Sounds really messy. Major liability because you have basically no protection AND you move at such slow speed. No thanks!
Here's a niche for P2P software... obtaining data from a website in a distributed way so as not to stick out in the website logs the way a one or a few download clients would. The collected data could be processed remotely or uploaded to a central server.
Finally, if the stolen code is so bloody obvious then why not show even one example of where there is direct copying. This wouldn't affect their legal strategy one bit (despite claims to the otherwise) and would grant them so much more credibility.
Yes, SCO's case has FUD written all over it. Why should they not try and settle out of court? Remove the offending code and have someone clean-room it, and all is well again. Just like you said - why not even show any code but rather threaten half the world with law suits? It makes absolutely NO sense.
My guess is they either want a buyout or they have something to gain by fear, uncertanty and doubt about Linux.
Let's say SCO is right and some code was copied. That leaves 99.99% of Linux in the clear. That's why I'm not concerned about this lawsuit at all. As soon as we find out what code is covered, we rewrite it. And SCO is left behind with millions of wasted legal fees and a destroyed reputation.
I think before I spent a significant amount of cash on a project like this I would like to be assured that the government wouldn't shut it down for some reason. Cool things often are, it seems. And if not the government, then hopefully not a stupid company and its lawyers who would come running and say (short of breath) "hey! that's miiine! I THOUGHT OF IT FIRST GODDAMM *cough cough* IIITTT!"
Why would I want a $2 coin? It would just increase the inflation rate of everything sold in a vending machines (anything less than $2 will go for $2 and stuff that was $2.5 would go for $4). No thanks.
:)
You're right, actually. And you forgot bus fair.
But I have to laugh at this story, because I've heard a number of Americans laugh at Canada's "monopoly money" over the years and now coloured money is coming to the US! Though I have to admit the new Canadian $10 bills are a little funky.
What kind of a world do we live in when the top priority is stopping spam from hitting our inboxes? What about social issues facing us every day? Cleaner air, water, nutritious food, mental health, affordable living, I could go on and on.
Where are these issues in our government's agenda? Spam will go away on its own - we have the tools to fight it, so legislation is not needed. Laws should not be a first resort.
Notice on just about every chart they cut it off before Microsoft Windows crapped out. You can plainly see that Red Hat is steadily providing good throughput, moving a little lower over time, while Microsoft's throughput is moving relatively quickly lower as the # of clients is increased.
Who could believe any Microsoft-sanctioned benchmark report given their track record?
I find it interesting how M$ used 64k clusters and left Red Hat at the default, when M$ default is 4K clusters. You'll note that you can't defragment a Windows volume with cluster size != 4KB out of the box.
I'd also be interested to know the CPU usage on each configuration while running these tests. Windows XP uses over 80% CPU to keep the pipes full, while Red Hat (and even Win2k) use ~ 10%.
Of course, you still can't argue about the price. How much does 160 client access licenses cost for Windows? "You can't beat free."
Everyone's all about freedom, but nobody seems to want to pay the price for it.
That's quite a blanket statement. Open source advocates pulled Mandrake's ass out of the gutter once when they needed it. Red Hat is doing just fine with their subscriptions... they don't even _need_ the OS revenue IMO.
Open source is open source. It's meant to be free as in speech AND free as in beer. Those trying to sell it realize that and take the associated risks.
I've given money on several occasions to open source causes, but I'm not going to shell out for every upgrade since that would just defeat much of the incentive to use open source.
What would you do if they decided to switch over to Microsoft? You'd all be screwed then.
No, "we" wouldn't be screwed. Red Hat != Open Source. Red Hat is a definite asset to the community, but it is not THE community.
I think turning to the law is the wrong place to look for a solution. Plenty of ideas have been thought of to limit spam, and more will come. Let the market decide. We have way too many laws as it is. In fact, shouldn't there come a point when we don't need NEW laws?
Society is but a veneer over our true nature.
That's a big YEP. Eventually once all IT jobs are outsourced to those c*cks*ckers in India, we'll all be relics of a byegone era.
I think you're directing your anger at the wrong people. The people in India are just doing their jobs. The c*cks*ckers are the ones in America hiring them.
What I find funny is that outside workers being hired in the US is being clamped down on, while companies are still free to send work offshore. Huh?