I didn't take a second look at ATI because of their Linux (lack-of) support on my laptop's ATI Radeon Mobility card. And also, last time I looked, ATI was about as big as NVIDIA is and they make the same products- graphics cards and motherboard chipsets. Oh, and the competition between the two is fierce.
I ran VMware Server on my 2.2GHz Pentium 4-M laptop with 1GB RAM with SuSE 10.0 as the host OS and Windows XP as the guest. It installed and ran cleanly, but yes, my 4-year-old laptop wasn't that fast. The Windows guest ran about like it was on a PIII-800, but the entire computer was still usable, inside and outside the VM. Now, my 2.2GHz dual-core Athlon 64 desktop with 2GB RAM runs the same guest OS and host OS and it runs far faster and the computer is far more responsive inside and outside the VM. After looking at the CPU usage, RAM usage, and hard drive I/O, it appears that the responsiveness is mostly due to the HDD speed once you have >= 1GB RAM. A 10,000rpm HDD sure works a lot better for running 2 OS's requests off of versus a 5400rpm one.
Apparently that is not really true anymore with the 2.6.x kernel, whereas XFS was a little flaky with the 2.4.x kernels. I had several ex-SGI employees on the local LUG link to a ream of data showing that the flakiness has been fixed for the 2.6 kernel series. I had thus switched my ReiserFS 3.6 partitions to XFS and they benchmarked a little better and I've had zero problems with them even after a bunch of unclean umounts due to playing with laptop suspend/resume.
Or you could use something like SuSE, where a significant amount of the packages in the AMD64 version are actually i586. Sure, everything works just great as you run 32-bit OpenOffice, 32-bit Firefox with 32-bit Java and 32-bit Flash, and also 32-bit Xine to run the w32codecs. But, as you might have imagined, "she's built for comfort, not for speed." I have a lot of horsepower running OpenSuSE 10.1 (AMD64) and it's not that poky, but not as fast as you'd expect an AMD 2.2GHz dual-core CPU with 2GB RAM in dual channel running off of a 10,000rpm hard drive to be.
I have a friend that runs Gentoo, maybe I'll see if it's worthwhile. He was running it on an a dual PII-400 and also his pride and joy, an old Compaq ProLiant 1GHz Quad PIII Xeon 4U server, and the server was roughly as fast as my 2.2GHz Pentium 4-M laptop in single-threaded apps. Are there any "gotchas" to using Gentoo that I should know, other than the "stable" Gentoo branch is roughly like Debian stable- outdated and the masked ~ packages are the ones to use?
I have to say the same for my 2.2GHz Pentium 4-M Northwood laptop. It's simply the hottest CPU I have ever used- the sucker IDLES at 65 C in my laptop and it's only a 35W TDP chip- not the desktop 70W ones. Under full load, I've hit 80 C but usually it is around 78 C. And of course, the fan is a hair-dryer one that goes from merely loud to screaming loud when the chip hits 75 C. My new desktop's AMD X2 4200+ puts off 89W but runs full-bore at 45 C with 1700rpm fans, so it's near silent. At idle, it is maybe 2 C higher than the room temp.
So if the ISP backbone data costs a lot of money and the ISP customer costs little, why don't ISPs have local caches of large files, things like popular free TV show downloads, a Linux distro mirror and cached copies of Windows updates?
I'll tell you why: it costs money to host data servers, and it is simply easier and cheaper to try to get the customers and now the sites they visit to pay the ISP than for the ISP to host legal files on their own. However, if the Net neutrality bills all pass, I betcha you will see ISP FTP/HTTP download mirrors all stocked with this kind of thing.
Yeah, I saw a dead Socket A board about a week ago. The cheesy little chipset fan on the VIA KT266A northbridge died, rendering the board dead. The CPU was surely not overclocked as the owner is very vocal against overclocking. Maybe his master's in EE taught him something the middle-school gamers don't know...
From what I hear, most cable users are on DHCP. In fact, my cable provider doesn't even allow the option of a static IP whereas the DSL provider will (for an extra fee, of course.)
Re:That's _exactly_ what we need...
on
Explorer Destroyer
·
· Score: 2, Interesting
I'd like to get one of those banners saying "Switch to Linux." All the banners I ever see are "This browser [KDE Konqueror] is not supported, download Microsoft Internet Explorer 6.0 or higher." The WebCT online course system always tells me "You appear to be running Linux. Linux is not a supported operating system. Please select your operating system below: {Windows} {Macintosh.}" But the ironic thing is that Konqueror on Linux runs flawlessly on WebCT and IE 6.1 on Windows will often crash. Poetic justice, perhaps.
I would second your post. I have fought with a rather expensive specialized program (PIXCI's XCAP camera software) written in Java and it is absolutely what you describe- slow, buggy, and crappy on both Windows and Linux. However, Azureus runs like a gazelle compared to XCAP. Maybe Java just has a continuum of crappiness from "just slightly crappy" down to "Windows ME."
Azureus is a Java program and the Java Runtime Environment is a resource hog. But, it does allow Aelitis to write one version and offer it for a multitude of different OSes. Otherwise I bet it would be Windows-only and that would suck.
They;ll use something built into the TPM chips to do the logging and hashing so it cannot be user-tamperable and still valid. Mark my words- this is what will be done.
Most of the problem with the prices on textbooks in my school (Univ. of MO-Columbia) is that the same professor rarely teaches the exact same class year after year after year. Different professors like to use different books, so for the two years that Prof. Smith is teaching Widget Design, we use his book of choice, and then when he teaches Widget Encoding II and Prof. Jenkins teaches Widget Design, we use her favorite book.
It has mostly to do with the relative lack of structure in the department curriculum and the large freedom of the instructors to change things as they wish. It's not bad, per se, but it does lead to a lot of new/different books being required. Of course, there are the sinister kickback/back-scratching setups, but I see that a lot less than just the shakeup of professors and their personal preferences driving the book market.
Now the publishers are doing some ripping off, but as long as the professors don't have a clue and keep mandating the newest, slightly-altered editions, why should publishers NOT take advantage of it? Would you want to work for less that you're getting paid if you think that you really are overpaid? Hell no. Why should they be any different?
Exactly. When you see about any new electronics part here in the U.S., it has all sorts of certification stickers on it, and usually one of them is "RoHS compliant." Why? Because it's cheaper to make ONE version of a product than two in most cases. Another case in point is reformulated gasoline- the fewer the number of kinds, the cheaper it is to make each of them.
Slashdot Mirror
I didn't take a second look at ATI because of their Linux (lack-of) support on my laptop's ATI Radeon Mobility card. And also, last time I looked, ATI was about as big as NVIDIA is and they make the same products- graphics cards and motherboard chipsets. Oh, and the competition between the two is fierce.
I ran VMware Server on my 2.2GHz Pentium 4-M laptop with 1GB RAM with SuSE 10.0 as the host OS and Windows XP as the guest. It installed and ran cleanly, but yes, my 4-year-old laptop wasn't that fast. The Windows guest ran about like it was on a PIII-800, but the entire computer was still usable, inside and outside the VM. Now, my 2.2GHz dual-core Athlon 64 desktop with 2GB RAM runs the same guest OS and host OS and it runs far faster and the computer is far more responsive inside and outside the VM. After looking at the CPU usage, RAM usage, and hard drive I/O, it appears that the responsiveness is mostly due to the HDD speed once you have >= 1GB RAM. A 10,000rpm HDD sure works a lot better for running 2 OS's requests off of versus a 5400rpm one.
Apparently that is not really true anymore with the 2.6.x kernel, whereas XFS was a little flaky with the 2.4.x kernels. I had several ex-SGI employees on the local LUG link to a ream of data showing that the flakiness has been fixed for the 2.6 kernel series. I had thus switched my ReiserFS 3.6 partitions to XFS and they benchmarked a little better and I've had zero problems with them even after a bunch of unclean umounts due to playing with laptop suspend/resume.
Or you could use something like SuSE, where a significant amount of the packages in the AMD64 version are actually i586. Sure, everything works just great as you run 32-bit OpenOffice, 32-bit Firefox with 32-bit Java and 32-bit Flash, and also 32-bit Xine to run the w32codecs. But, as you might have imagined, "she's built for comfort, not for speed." I have a lot of horsepower running OpenSuSE 10.1 (AMD64) and it's not that poky, but not as fast as you'd expect an AMD 2.2GHz dual-core CPU with 2GB RAM in dual channel running off of a 10,000rpm hard drive to be.
I have a friend that runs Gentoo, maybe I'll see if it's worthwhile. He was running it on an a dual PII-400 and also his pride and joy, an old Compaq ProLiant 1GHz Quad PIII Xeon 4U server, and the server was roughly as fast as my 2.2GHz Pentium 4-M laptop in single-threaded apps. Are there any "gotchas" to using Gentoo that I should know, other than the "stable" Gentoo branch is roughly like Debian stable- outdated and the masked ~ packages are the ones to use?
I have to say the same for my 2.2GHz Pentium 4-M Northwood laptop. It's simply the hottest CPU I have ever used- the sucker IDLES at 65 C in my laptop and it's only a 35W TDP chip- not the desktop 70W ones. Under full load, I've hit 80 C but usually it is around 78 C. And of course, the fan is a hair-dryer one that goes from merely loud to screaming loud when the chip hits 75 C. My new desktop's AMD X2 4200+ puts off 89W but runs full-bore at 45 C with 1700rpm fans, so it's near silent. At idle, it is maybe 2 C higher than the room temp.
So if the ISP backbone data costs a lot of money and the ISP customer costs little, why don't ISPs have local caches of large files, things like popular free TV show downloads, a Linux distro mirror and cached copies of Windows updates?
I'll tell you why: it costs money to host data servers, and it is simply easier and cheaper to try to get the customers and now the sites they visit to pay the ISP than for the ISP to host legal files on their own. However, if the Net neutrality bills all pass, I betcha you will see ISP FTP/HTTP download mirrors all stocked with this kind of thing.
Yeah, I saw a dead Socket A board about a week ago. The cheesy little chipset fan on the VIA KT266A northbridge died, rendering the board dead. The CPU was surely not overclocked as the owner is very vocal against overclocking. Maybe his master's in EE taught him something the middle-school gamers don't know...
Then to be more true to its roots, should /. become ./ ?
So *THAT'S* where Xena went to...
I've heard people piss and moan about XP, but it's mostly when it gives them blue screens or bugs them about activation/WGA, etc.
So are x86 boxes, and if you build one yourself, it's FAR more upgradeable than any OEM box as you will only use standard parts.
That's called Internet Protocol Version 6, where some of the address is your MAC address.
From what I hear, most cable users are on DHCP. In fact, my cable provider doesn't even allow the option of a static IP whereas the DSL provider will (for an extra fee, of course.)
I'd like to get one of those banners saying "Switch to Linux." All the banners I ever see are "This browser [KDE Konqueror] is not supported, download Microsoft Internet Explorer 6.0 or higher." The WebCT online course system always tells me "You appear to be running Linux. Linux is not a supported operating system. Please select your operating system below: {Windows} {Macintosh.}" But the ironic thing is that Konqueror on Linux runs flawlessly on WebCT and IE 6.1 on Windows will often crash. Poetic justice, perhaps.
He threatens to display it at the next LinuxWorld conference unless a donation of $100,000 to the FSF is made.
I would second your post. I have fought with a rather expensive specialized program (PIXCI's XCAP camera software) written in Java and it is absolutely what you describe- slow, buggy, and crappy on both Windows and Linux. However, Azureus runs like a gazelle compared to XCAP. Maybe Java just has a continuum of crappiness from "just slightly crappy" down to "Windows ME."
Azureus is a Java program and the Java Runtime Environment is a resource hog. But, it does allow Aelitis to write one version and offer it for a multitude of different OSes. Otherwise I bet it would be Windows-only and that would suck.
They;ll use something built into the TPM chips to do the logging and hashing so it cannot be user-tamperable and still valid. Mark my words- this is what will be done.
Most of the problem with the prices on textbooks in my school (Univ. of MO-Columbia) is that the same professor rarely teaches the exact same class year after year after year. Different professors like to use different books, so for the two years that Prof. Smith is teaching Widget Design, we use his book of choice, and then when he teaches Widget Encoding II and Prof. Jenkins teaches Widget Design, we use her favorite book.
It has mostly to do with the relative lack of structure in the department curriculum and the large freedom of the instructors to change things as they wish. It's not bad, per se, but it does lead to a lot of new/different books being required. Of course, there are the sinister kickback/back-scratching setups, but I see that a lot less than just the shakeup of professors and their personal preferences driving the book market.
Now the publishers are doing some ripping off, but as long as the professors don't have a clue and keep mandating the newest, slightly-altered editions, why should publishers NOT take advantage of it? Would you want to work for less that you're getting paid if you think that you really are overpaid? Hell no. Why should they be any different?
That happens all the time for anything tech-related. The only thing that gets worse and more expensive with time is gasoline!
'..and our rootkit DOES have open-source software in it!" -Sony
They hate Broadcom because their Linux support sucks.
How about a doubly recursive name: (GNU) HURD.
Exactly. When you see about any new electronics part here in the U.S., it has all sorts of certification stickers on it, and usually one of them is "RoHS compliant." Why? Because it's cheaper to make ONE version of a product than two in most cases. Another case in point is reformulated gasoline- the fewer the number of kinds, the cheaper it is to make each of them.
Nah, he's like a dog- his size is 99% fur.