I see the reverse issue with my Radeon 9200 - brilliant display on the VGA connector, shocking on the DVI+adapter secondary output. I've resorted to not using that output at all, and sticking a PCI TNT2 back in to drive my second monitor. The TNT2 looks noticeably worse at times than the Radeon's primary output, but it's a lot better the secondary output. I may end up paying through the nose for a Matrox card next time I upgrade my graphics card - which probably won't be for a few years - because I have heard generally good things about their 2D output quality and multi-head ability.
Most newer CRTs will also report their physical dimensions, and modern XFree86 will, by default, set its resolution setting to something computed based on that. Sadly, on Debian at least, kdm/xdm override this setting and force the default DPI to 100dpi. Also, X's dpi setting breaks when Xinerama is used: I have one 115dpi monitor and another 96dpi monitor, and there's no way to set their resolution independently.:-(
Windows is rather broken when it comes to DPI setting in my experience - it is clumsy to set, and a lot of programs break badly because they assume the Windows default of 96dpi. If you have a 17" LCD (or a 17" CRT @1152x864) it'll be more-or-less correct by accident though...
PS. Just noticed who I was replying to. Hi Craig;-)
this morning apt got errors and suggested upgrades never annouced for after an update. I smelled something bad and did not upgrade, waiting for just this kind of news;-)
I suspect the real reason that it found packages to upgrade was that a new stable version had been released, not because the packages had been compromised. At least, that's how I read the announcement:-)
Xinerama works very well, but I sometimes want to drag a window between desktops, without paying for a commercial X.
In that case you haven't actually enabled Xinerama and are using X's traditional dual-head mode. Try sticking Option "xinerama" "true" in the serverlayout section of your XF86Config and see how it goes.:)
I don't know what programs you're running and through what kind of a pipe, but Mozilla used remotely over a 512/512 DSL line is simply useless.
I suspect that this is mainly due to the latency of the DSL line, rather than the raw throughput, although 512kbit isn't really a lot for X to play with. Something like TightVNC will feel quite snappy - it's even usable on a 64kbit connection. On the other hand, over 100Mbit ethernet you will likely never notice the difference between a remote X app and an app running on the local machine unless you're doing something silly like streaming video or playing quake.
I just tried test5-mm4, it won't even boot. Quite unfortunate as this system, although fairly new, is pretty bare-bones made of off the shelf components. (A7N8X nforce2 mb, Athlon XP CPU.)
Try disabling local APIC support. I too have an A7N8X mobo and it seems as though the board and the kernel's APIC code really don't like each other; I tried an earlier 2.6-test release and it just spat out an endless stream of errors during booting with APIC enabled. (Incidentally turning off local APIC made the system a lot more stable under 2.4.x too.)
NB. APIC/= ACPI. ACPI seems to work fine on this board.
No, don't! Lossless compression on audio and video is unlikely to make a significant dent in the file size, and you're looking at ~1GB/min for uncompressed video. Even if you have half a terabyte of storage you're not likely to be able to keep much lying around at that rate.
There are plenty of filesystems, even Linux ones, that don't like files bigger than 2GB.
Which in turn has nothing to do with the processor. Certainly ext2/ext3 on Linux and NTFS on Windows have no trouble handling multi-gigabyte files on a 32-bit CPU.
the link below is from a social scientist but I remember some physicists doing the same thing
Actually, as the URL you gave seems to imply, Alan Sokal is a physicist, but his article was published in Social Text. The other one you're probably thinking of was the Bogdanov brothers; unlike Alan Sokal, their paper was (apparently) not intentionally gibberish.
Not quite... DotNET is platform-independent. The Mono project is developing an independent implementation of.NET for Linux - with Microsoft's help, as I understand it.
Java is slow for desktop applications because its graphics library is rubbish, not because the VM is inherently slow.
Re:The only problem with Ogg
on
AAC vs. OGG vs. MP3
·
· Score: 2, Insightful
I'm guessing there's something wrong with your software set-up, as my poor 800MHz Duron can rip+encode at around 4x real-time.
As for not needing larger hard drives... well, I have a 60 gig which is about half full; it feels rather constraining at times - capturing video in full PAL resolution sucks up close to 1GB/min, and my CPU is nowhere near fast enough to encode that in real time.
It doesn't have a fan any more, just the heat sink. I ripped out the fans from both my graphics cards because they were rather noisy. (The Geforce4 has an enormous heat sink and barely gets warm; the TNT2 is rather hot to the touch, but seems to be okay.) I'd also imagine that a fan that wasn't spinning would be a lot worse, from a heat dissipation perspective, than no fan at all.
The TNT2 should only ever be doing 2D, because it's powering my secondary monitor, and neither Windows nor XFree support 3D on more than one head with two separate cards.
In my experience, it depends a lot on the card in question; also, some of the more recent Linux drivers are more stable than others.
I have a Geforce 4MX and a TNT2 on my machine (i.e. dual-head), and ran the Geforce4 with the NVIDIA closed-source binary drivers and the TNT2 with the XFree drivers. Why? Because the system locks hard every few hours with the NVIDIA drivers for the TNT2. Interestingly, it has exactly the same symptoms in Win2K - if I hadn't been using Linux on the same system, I would probably be blaming Microsoft for NVIDIA's buggy drivers.
On the other hand, getting the NVIDIA drivers to work at all is a pain. In fact, getting working drivers for any hardware that isn't supported by the stock kernel is a pain.
Slashdotters are not isomorphic.p> It is likely that the people who condone conning PPV movies and illegally trading music, software and movies are not the same people who grizzle about GPL violations and other corporate nastiness.
Heh, I read those instructions and was thinking how in Linux something like that would be the answer to a question like "How do I get my backspace key to work?"
Grr, the backspace key hasn't been a problem in Linux for at least the last seven years, but people still keep bringing it up. It's as if people were complaining about having to mess with config.sys and autoexec.bat to get Windows to work.
Slashdot Mirror
I see the reverse issue with my Radeon 9200 - brilliant display on the VGA connector, shocking on the DVI+adapter secondary output. I've resorted to not using that output at all, and sticking a PCI TNT2 back in to drive my second monitor. The TNT2 looks noticeably worse at times than the Radeon's primary output, but it's a lot better the secondary output. I may end up paying through the nose for a Matrox card next time I upgrade my graphics card - which probably won't be for a few years - because I have heard generally good things about their 2D output quality and multi-head ability.
Windows is rather broken when it comes to DPI setting in my experience - it is clumsy to set, and a lot of programs break badly because they assume the Windows default of 96dpi. If you have a 17" LCD (or a 17" CRT @1152x864) it'll be more-or-less correct by accident though...
PS. Just noticed who I was replying to. Hi Craig ;-)
I suspect the real reason that it found packages to upgrade was that a new stable version had been released, not because the packages had been compromised. At least, that's how I read the announcement :-)
That would be User Mode Linux.
In that case you haven't actually enabled Xinerama and are using X's traditional dual-head mode. Try sticking Option "xinerama" "true" in the serverlayout section of your XF86Config and see how it goes. :)
I suspect that this is mainly due to the latency of the DSL line, rather than the raw throughput, although 512kbit isn't really a lot for X to play with. Something like TightVNC will feel quite snappy - it's even usable on a 64kbit connection. On the other hand, over 100Mbit ethernet you will likely never notice the difference between a remote X app and an app running on the local machine unless you're doing something silly like streaming video or playing quake.
Try disabling local APIC support. I too have an A7N8X mobo and it seems as though the board and the kernel's APIC code really don't like each other; I tried an earlier 2.6-test release and it just spat out an endless stream of errors during booting with APIC enabled. (Incidentally turning off local APIC made the system a lot more stable under 2.4.x too.)
NB. APIC /= ACPI. ACPI seems to work fine on this board.
Hope that helps!
Cameron.
It may sound unlikely but that doesn't mean it hasn't happened before.
Shameless plug: I have a TI-83 Plus (as well as an HP48) and have written an RPN emulator for it.
No, don't! Lossless compression on audio and video is unlikely to make a significant dent in the file size, and you're looking at ~1GB/min for uncompressed video. Even if you have half a terabyte of storage you're not likely to be able to keep much lying around at that rate.
That shouldn't stop you from using vi, though! Head over to vim.org and grab a copy of Vim for Windows and you're all set...
Which in turn has nothing to do with the processor. Certainly ext2/ext3 on Linux and NTFS on Windows have no trouble handling multi-gigabyte files on a 32-bit CPU.
Actually, as the URL you gave seems to imply, Alan Sokal is a physicist, but his article was published in Social Text. The other one you're probably thinking of was the Bogdanov brothers; unlike Alan Sokal, their paper was (apparently) not intentionally gibberish.
Your's not to make reply, ... although possessive forms normally have apostrophes, yours (and its) is (are) exception(s).
Try Yours
No it won't. Much to my chagrin they "fixed" that bug in 4.7something.
Java is slow for desktop applications because its graphics library is rubbish, not because the VM is inherently slow.
As for not needing larger hard drives... well, I have a 60 gig which is about half full; it feels rather constraining at times - capturing video in full PAL resolution sucks up close to 1GB/min, and my CPU is nowhere near fast enough to encode that in real time.
I thought it was William of Occam...
The TNT2 should only ever be doing 2D, because it's powering my secondary monitor, and neither Windows nor XFree support 3D on more than one head with two separate cards.
I have a Geforce 4MX and a TNT2 on my machine (i.e. dual-head), and ran the Geforce4 with the NVIDIA closed-source binary drivers and the TNT2 with the XFree drivers. Why? Because the system locks hard every few hours with the NVIDIA drivers for the TNT2. Interestingly, it has exactly the same symptoms in Win2K - if I hadn't been using Linux on the same system, I would probably be blaming Microsoft for NVIDIA's buggy drivers.
On the other hand, getting the NVIDIA drivers to work at all is a pain. In fact, getting working drivers for any hardware that isn't supported by the stock kernel is a pain.
Erm, it does, actually.
Heck, you can even combine all the important letters and buzzwords into one project, like KxmlRPCd did.
Slashdotters are not isomorphic.p> It is likely that the people who condone conning PPV movies and illegally trading music, software and movies are not the same people who grizzle about GPL violations and other corporate nastiness.
It was definitely there in Netscape 3.0, and I suspect it was there all the way back to Netscape 1.1N.
Grr, the backspace key hasn't been a problem in Linux for at least the last seven years, but people still keep bringing it up. It's as if people were complaining about having to mess with config.sys and autoexec.bat to get Windows to work.