Re:Use an NP-hard problem
on
Gates on Spam
·
· Score: 2, Informative
Just for your edification, O(NP-hard) = O(NP-complete).
That is not correct.
That is, all NP-complete problems are NP-hard, but not visa versa.
That is correct.
So it's misleading to say "NP-Complete or at least NP-hard".
Yes, I agree with that.
To be a litle more precise NP-complete is defined to be the intersection of NP and NP-hard. And P is a subset of NP which is very easilly shown from the definition. If there exist a problem which is in both P and NP-complete, then it will follow that P=NP (and for all practical purposes NP-complete will be the same set), however even if P=NP there will be NP-hard problems that are outside P.
How fast can you change focus? If you can change faster than the time it takes to take a picture, you could actually use different focus for the same picture. Each pixel could look on the values of a few neighbour pixels to find out when the picture is sharpest in this region and save only the pixel value from that time. Information about the actual focus used could be saved in the alpha channel. Imagine a picture of an object 10cm from you where both that object and the background is sharp.
That have been disscussed before, but actually I got 15281677 results when searching for Linux right now. Anyway the reason some terms sometimes give fewer results apears to be that the first results are sponsored results. And until you have gone through all sponsored results it shows only the number of sponsored results. Of course sponsored results might very well include pages explaining why Windows is better than Linux.
XFree86+XFree86
no warning either.
it is very strange
I tried a few searches. It is more strange than you would imagine. XFree86 is filtered, typing it twice bypass the filter. Porn is filtered, typing it twice cause msn to remove the space between the two instances of the word and instead search for pornporn which is allowed. But you can search for porn by typing it three times. Sex is filtered, typing it two or three times doesn't help. If you really want to search for sex, you must type it at least four times.
I read the article, and I still haven't got any clue to what the problem is. I wouldn't be surprised, if it turned out, the problem really is the 99% of people still thinking Windows is the only option. It is not like Linux doesn't have the applications needed for the average office, but maybe accessing your data stored in proprietary formats is a problem. That means switching is a good idea, the sooner the better. Because the price of switching doesn't go down if you wait. And the higher the price of switching away from Windows becomes, the higher Microsoft can turn the price of staying. Deploying Linux in an organization doesn't mean you need everybody to have Linux expertise. But you do need a few people with Linux expertise.
I just tried with ogle, it works just fine. No need to use dd, just use cp/dev/dvd filename.iso. I even tried starting to play the DVD over NFS while I was still copying from DVD to file. That also worked, though performance wasn't good. And BTW a DVD can be larger than 3-5GB. I have a DVD with 7.6GB. I don't know if there could be any problems with CSS encrypted DVDs. I just picked one random DVD, and it worked. I don't even know which of my DVDs are encrypted.
If a program expects to have all memory writeable, and tries to function under those conditions, anything that stops it, regardless of how small the patch necessary is, will break these programs.
Programs can still write to the exact same pages as they could before. The difference is, that now they cannot execute code from nonexecutable pages.
Don't forget that 24-bit colour is woefully inadequate for film work. Cine film has got so much more dynamic range than video
This has often got me thinking that maybe it would make more sense to use a logarithmic scale rather than a linear scale for digital sound and pictures.
Although Patriot has plans to move forward with its 32-bit processors and application-specific integrated circuits, Wallin said that product revenues were currently "negligible."
"Our main focus is the IP [intellectual property] business now," he said.
we once had two deliberate fires in one month, and a couple of attempts that didn't suceed. Shortly after we got a few surveillance cameras. Since then we have not had any fires around the building where I live, but then it started happen in different places not very far from here.
my ISP's mail servers are slow as hell because of this crap.
Our department have recently upgraded our mailserver again. That have happened every one or two years for as long as I have been there. Started out with a Sun server switchted to a SGI server. Then it was a Linux server for a while until we recently got this shiny new one which BTW also runs Linux. Am I the only one who thinks an x345 is one hell of a server just to filter away those 90% of the incomming email because it is spam or vira? And still two thirds of the email I get is spam. Imagine what this Dual 3GHz Xeon with 8GB ram could do if it didn't have to filter spam all day long.
Re:It's like Netscape v. Microsoft in that...
on
Google v. Microsoft
·
· Score: 1
The slick thing would be for competing search engines to 'crawl' google itself.
Everything that would be interesting to crawl is listed in http://google.com/robots.txt. And there are good technical reasons why google have done that. It is not to make life difficult for competitors. The robots.txt on Google lists only what such a file is supposed to list, it is generally a help to any search engine comming across the site, including Googlebot. But the robots.txt does forbid others to freeload on Google's data. That said robots.txt is only advisory in the sence, that respecting it have to be implemented on the client side. It wouldn't be difficult to actually crawl the net by downloading pages from Google's cache.
The article says Linux (pronounced "LINN-ix"). That is incorrect. In the samples you can download from kernel.org you can hear
Linus pronouncing Linux.
Re:It's like Netscape v. Microsoft in that...
on
Google v. Microsoft
·
· Score: 2, Insightful
Would Microsoft be so low as to "embellish" the robots.txt file hosted on IIS sites so as to include a line forbidding the GoogleBot?
I think we would have another antitrust case if Microsoft did so. Sure it would take years in court, but I think Google might decide to ignore the robots.txt file if they really believe it was illegal and a threat to Google. People picking side and creating robots.txt file probably isn't illegal, at least we are not facing an antitrust case there. Well, since Microsofts crawler haven't really found anything of interest on my site, and Google have already crawled most of it, I don't think there is yet any point in trying to give Google and advantage. They already have the advantage they need. Anyway how would things turn out if people starting placing this robots.txt on various webservers:
User-agent: * Disallow:/default.ida
User-agent: msnbot Disallow:/
Re:It's like Netscape v. Microsoft in that...
on
Google v. Microsoft
·
· Score: 1
Google has indexed the internet. That data is ultimately more
powerful
Presumably Microsoft have the resources to do that as well. Of course it is not as trivial as it may sound. Google has another advantage as well. I have seen robots.txt files allowing only Google to crawl the site and forbiding everybody else access. What are Microsoft going to do about those? Violate the directions and crawl sites without permissions? Even if they might be able to crawl sites without people noticing that some requests originate from a crawler, what will people say when they see Microsofts search site showing up as referer? Maybe it is time to boycot Microsoft's crawler and install robots.txt files explicitly forbiding them access. Considering Microsoft's methods I wouldn't feel bad about doing so myself.
Think of back-scatter: when a virus hits a mail server, and the server bounces the message (to the forged sender address)
For that exact reason I advice people to have the send an error message at once rather than first accept the email and then produce a bounce message. If your mailserver send an error message at the end of DATA, the error message is send directly to the worm. So no mail server will ever produce a bounce in that case.
I use Linux not because I rebel against anyone, it's just that I got tired of the blue-screen-of-death...
I use Linux because I needed something that would work nicely together with Unix. Being open source I figured I might as well start reading the source, and later I even started modifying it on my own. I became so happy with having access to the source, that now I don't want to give up on that. I was never a Microsoft user. Before I bought a computer for running Linux I used my Amiga 1200. Had Microsoft Office been available for AmigaOS ten years ago I might have used it. But not today. Latex just works better with CVS.
I hate Microsoft, not because of Windows crashing all the time (since I don't use it), but rather because of all the problems they are responsible for. Whenever I see hardware or websites not working with Linux, I know Microsoft Windows is part of the reason.
If I remember correctly 2.2 didn't have DMA support
I actually have some 2.2 kernels running with DMA on IDE drives. It didn't get enabled automatically though. I had to change some configuration files so hdparm would be used to enable DMA during boot. In one of those machines we just installed a new harddisk a few days ago. I couldn't get DMA to work on the new harddisk eventhough it worked on the old disk on the same controler. I tried booting a 2.4 kernel, then I could do DMA on both disks. Enabling DMA on the new disk improved throughput from 2MB/s to 26MB/s.
Slashdot Mirror
Just for your edification, O(NP-hard) = O(NP-complete).
That is not correct.
That is, all NP-complete problems are NP-hard, but not visa versa.
That is correct.
So it's misleading to say "NP-Complete or at least NP-hard".
Yes, I agree with that.
To be a litle more precise NP-complete is defined to be the intersection of NP and NP-hard. And P is a subset of NP which is very easilly shown from the definition. If there exist a problem which is in both P and NP-complete, then it will follow that P=NP (and for all practical purposes NP-complete will be the same set), however even if P=NP there will be NP-hard problems that are outside P.
How fast can you change focus? If you can change faster than the time it takes to take a picture, you could actually use different focus for the same picture. Each pixel could look on the values of a few neighbour pixels to find out when the picture is sharpest in this region and save only the pixel value from that time. Information about the actual focus used could be saved in the alpha channel. Imagine a picture of an object 10cm from you where both that object and the background is sharp.
Well thats funny a search for linux on msn brings up only 414 results, I WONDER why? http://search.msn.com/results.aspx?q=Linux&FORM=SM CRT
That have been disscussed before, but actually I got 15281677 results when searching for Linux right now. Anyway the reason some terms sometimes give fewer results apears to be that the first results are sponsored results. And until you have gone through all sponsored results it shows only the number of sponsored results. Of course sponsored results might very well include pages explaining why Windows is better than Linux.
XFree86+XFree86 no warning either. it is very strange
I tried a few searches. It is more strange than you would imagine. XFree86 is filtered, typing it twice bypass the filter. Porn is filtered, typing it twice cause msn to remove the space between the two instances of the word and instead search for pornporn which is allowed. But you can search for porn by typing it three times. Sex is filtered, typing it two or three times doesn't help. If you really want to search for sex, you must type it at least four times.
go check Netcraft....they have actually dabbled in BSD for a bit on www.sco.com
www.sco.com doesn't exist anymore. However sco.com does still use Linux.
Oh the irony. Imagine:
I read the article, and I still haven't got any clue to what the problem is. I wouldn't be surprised, if it turned out, the problem really is the 99% of people still thinking Windows is the only option. It is not like Linux doesn't have the applications needed for the average office, but maybe accessing your data stored in proprietary formats is a problem. That means switching is a good idea, the sooner the better. Because the price of switching doesn't go down if you wait. And the higher the price of switching away from Windows becomes, the higher Microsoft can turn the price of staying. Deploying Linux in an organization doesn't mean you need everybody to have Linux expertise. But you do need a few people with Linux expertise.
I just tried with ogle, it works just fine. No need to use dd, just use cp /dev/dvd filename.iso. I even tried starting to play the DVD over NFS while I was still copying from DVD to file. That also worked, though performance wasn't good. And BTW a DVD can be larger than 3-5GB. I have a DVD with 7.6GB. I don't know if there could be any problems with CSS encrypted DVDs. I just picked one random DVD, and it worked. I don't even know which of my DVDs are encrypted.
If a program expects to have all memory writeable, and tries to function under those conditions, anything that stops it, regardless of how small the patch necessary is, will break these programs.
Programs can still write to the exact same pages as they could before. The difference is, that now they cannot execute code from nonexecutable pages.
All of the CGI effects in the movie were done on big iron Silicon Graphics machines at ILM
Actually I have been wondering if those were the same computers as those seen in the movie.
Don't forget that 24-bit colour is woefully inadequate for film work. Cine film has got so much more dynamic range than video
This has often got me thinking that maybe it would make more sense to use a logarithmic scale rather than a linear scale for digital sound and pictures.
Am I the only person who accidentially read this as Mr Bean?
we once had two deliberate fires in one month, and a couple of attempts that didn't suceed. Shortly after we got a few surveillance cameras. Since then we have not had any fires around the building where I live, but then it started happen in different places not very far from here.
and it's more like 99.9
I have often wondered how some people have come up with the usual 50% estimate for the amount worldwide.
my ISP's mail servers are slow as hell because of this crap.
Our department have recently upgraded our mailserver again. That have happened every one or two years for as long as I have been there. Started out with a Sun server switchted to a SGI server. Then it was a Linux server for a while until we recently got this shiny new one which BTW also runs Linux. Am I the only one who thinks an x345 is one hell of a server just to filter away those 90% of the incomming email because it is spam or vira? And still two thirds of the email I get is spam. Imagine what this Dual 3GHz Xeon with 8GB ram could do if it didn't have to filter spam all day long.
The slick thing would be for competing search engines to 'crawl' google itself.
Everything that would be interesting to crawl is listed in http://google.com/robots.txt. And there are good technical reasons why google have done that. It is not to make life difficult for competitors. The robots.txt on Google lists only what such a file is supposed to list, it is generally a help to any search engine comming across the site, including Googlebot. But the robots.txt does forbid others to freeload on Google's data. That said robots.txt is only advisory in the sence, that respecting it have to be implemented on the client side. It wouldn't be difficult to actually crawl the net by downloading pages from Google's cache.
The article says Linux (pronounced "LINN-ix"). That is incorrect. In the samples you can download from kernel.org you can hear Linus pronouncing Linux.
I think we would have another antitrust case if Microsoft did so. Sure it would take years in court, but I think Google might decide to ignore the robots.txt file if they really believe it was illegal and a threat to Google. People picking side and creating robots.txt file probably isn't illegal, at least we are not facing an antitrust case there. Well, since Microsofts crawler haven't really found anything of interest on my site, and Google have already crawled most of it, I don't think there is yet any point in trying to give Google and advantage. They already have the advantage they need. Anyway how would things turn out if people starting placing this robots.txt on various webservers:
Google has indexed the internet. That data is ultimately more powerful
Presumably Microsoft have the resources to do that as well. Of course it is not as trivial as it may sound. Google has another advantage as well. I have seen robots.txt files allowing only Google to crawl the site and forbiding everybody else access. What are Microsoft going to do about those? Violate the directions and crawl sites without permissions? Even if they might be able to crawl sites without people noticing that some requests originate from a crawler, what will people say when they see Microsofts search site showing up as referer? Maybe it is time to boycot Microsoft's crawler and install robots.txt files explicitly forbiding them access. Considering Microsoft's methods I wouldn't feel bad about doing so myself.
Think of back-scatter: when a virus hits a mail server, and the server bounces the message (to the forged sender address)
For that exact reason I advice people to have the send an error message at once rather than first accept the email and then produce a bounce message. If your mailserver send an error message at the end of DATA, the error message is send directly to the worm. So no mail server will ever produce a bounce in that case.
I use Linux not because I rebel against anyone, it's just that I got tired of the blue-screen-of-death...
I use Linux because I needed something that would work nicely together with Unix. Being open source I figured I might as well start reading the source, and later I even started modifying it on my own. I became so happy with having access to the source, that now I don't want to give up on that. I was never a Microsoft user. Before I bought a computer for running Linux I used my Amiga 1200. Had Microsoft Office been available for AmigaOS ten years ago I might have used it. But not today. Latex just works better with CVS.
I hate Microsoft, not because of Windows crashing all the time (since I don't use it), but rather because of all the problems they are responsible for. Whenever I see hardware or websites not working with Linux, I know Microsoft Windows is part of the reason.
cuz you've set your google preference to only search for english pages.
When I search it says only about 493. And I haven't changed any preferences.
If I remember correctly 2.2 didn't have DMA support
I actually have some 2.2 kernels running with DMA on IDE drives. It didn't get enabled automatically though. I had to change some configuration files so hdparm would be used to enable DMA during boot. In one of those machines we just installed a new harddisk a few days ago. I couldn't get DMA to work on the new harddisk eventhough it worked on the old disk on the same controler. I tried booting a 2.4 kernel, then I could do DMA on both disks. Enabling DMA on the new disk improved throughput from 2MB/s to 26MB/s.
I'd love to browse around sources without having to download a tarball using only my web browser.
That is possible. The best place I know is The Linux Cross-Reference.