If you run defrag on a fairly empty NTFS partition it's noticeable that some data will get shoved to the end of the partition and probably won't get moved back to the beginning.
If I were to be unkind, I would suggest that this is deliberate behaviour to prevent third party partition resizing applications reclaiming enough space to make a partition for a competing operating system install.
Accesses on a hard drive are theoretically faster closer to the edge of a platter.
there is no way to tell whether the parity, or the original data is at fault
Incorrect.
Take a four-disk RAID-5 array, with disks A, B, C, and P. There are four "copies" of the data: {A,B,C}, {B^C^P,B,C}, {A,A^C^P,C}, and {A,B,A^B^P}.
Randomly flip a bit on disk C. You've just changed the first three copies, but not the fourth one. In this situation, it is possible to say with reasonable certainty that the three copies that agree are wrong and the fourth one is correct.
define "still accessible" for this webserving nOOb? i see no qualifiers so i ask.
I'll bite.
Until this post, nothing on the Internet or my webserver linked here. Now this post links to it, but even without this link someone could've typed that URL into a web browser and gotten to that important confidential document.
However, even though I am linking here, you can't get to it. Rather than security through obscurity, that's real access control. (No such file exists; I have a mostly-empty/private directory that 403s almost everyone and 401s the rest.)
That's just not true. I have two 320GB hard drives in a ZFS mirror, with no less than 64 filesystems, and "only" 1GB of RAM. I had a slightly smaller non-mirrored array for a long time on a weaker machine (32-bit, 512MB RAM) with no problems also.
The point is that, in raytracing, you can assign each of your 800 "stream processors" different pixels. Done. You're parallel. When one finishes, give it another pixel to work on, and repeat until you've rendered the whole thing.
Each core still has to iterate over all (well, some, I'm oversimplifying) of the triangles, but it can do so COMPLETELY INDEPENDENTLY of the other cores and still come up with a good result. Your performance gains are almost linearly proportional to the number of cores.
You can even have a relatively high-latency connection (Gigabit Ethernet, for instance) between the various cores, broadcast the scene data over this connection, and then receive individual "chunks" of rendered pixels back. I defy you to do that with rasterization.
Rasterization is not embarrassingly parallel in the same way that raytracing is. Distributing tasks among those 800 "stream processors" is exceedingly complicated, because the underlying "task" involves iterating over every pixel that intersects with a given triangle rather than (as in raytracing) iterating over every triangle that could intersect with a given pixel.
I wrote a whole long, insightful comment that went through all the technical issues on both sides, but Discussion 2.0 ate it and now I'm ticked off, so I'll just write the conclusion.
In raytracing, you deal with pixels, so you can get linear performance gains just by adding nodes (even over a slowish network like gigabit Ethernet) and divvying out pixels to work on. In rasterization, you deal with triangles, so you can't.
Konqueror lets you scroll up and down with the cursor keys (shift+cursor key has some neat effects, too) and make all links on-screen keyboard-accessible by tapping "control."
Slashdot Mirror
Accesses on a hard drive are theoretically faster closer to the edge of a platter.
-:sigma.SB
Incorrect.
Take a four-disk RAID-5 array, with disks A, B, C, and P. There are four "copies" of the data: {A,B,C}, {B^C^P,B,C}, {A,A^C^P,C}, and {A,B,A^B^P}.
Randomly flip a bit on disk C. You've just changed the first three copies, but not the fourth one. In this situation, it is possible to say with reasonable certainty that the three copies that agree are wrong and the fourth one is correct.
-:sigma.SB
I'm still waiting for the Solaris port.
-:sigma.SB
They do, however, have a 6mb pipe to your house.
With wireless, everyone's sharing the same "last mile." A LOT of everyone. Not like with cable where it's just everyone on your block.
-:sigma.SB
You mean UTF-16 characters.
-:sigma.SB
(someone who hates the fact that a Java 'char' can only hold a single UTF-16 word, which can be half a surrogate pair and mess things up.)
Personally, I find 480p quite sufficient, and I have a screen almost as big diagonally as I am tall.
Then again, I play Unreal at 320x240 on a system with a Radeon 9800, on purpose. So, maybe I'm just insane.
-:sigma.SB
I'll bite.
Until this post, nothing on the Internet or my webserver linked here. Now this post links to it, but even without this link someone could've typed that URL into a web browser and gotten to that important confidential document.
However, even though I am linking here, you can't get to it. Rather than security through obscurity, that's real access control. (No such file exists; I have a mostly-empty /private directory that 403s almost everyone and 401s the rest.)
-:sigma.SB
http://www.linuxdevices.com/articles/AT8574944925.html
Unfortunately, it seems to have died before I could get my hands on one. :(
-:sigma.SB
From the inside, certainly.
-:sigma.SB
Just what I've always wanted. A never-ending loading screen with frames and ads.
No, I won't switch to Firefox. Konqueror suits me just fine.
-:sigma.SB
That's just not true. I have two 320GB hard drives in a ZFS mirror, with no less than 64 filesystems, and "only" 1GB of RAM. I had a slightly smaller non-mirrored array for a long time on a weaker machine (32-bit, 512MB RAM) with no problems also.
This is under FreeBSD.
-:sigma.SB
Until very recently, my Linux laptop was 266MHz and had 256MB of RAM. It ran FINE, thanks. (Even KDE did, when I tried it out.)
And I'm a performance freak.
-:sigma.SB
The point is that, in raytracing, you can assign each of your 800 "stream processors" different pixels. Done. You're parallel. When one finishes, give it another pixel to work on, and repeat until you've rendered the whole thing.
Each core still has to iterate over all (well, some, I'm oversimplifying) of the triangles, but it can do so COMPLETELY INDEPENDENTLY of the other cores and still come up with a good result. Your performance gains are almost linearly proportional to the number of cores.
You can even have a relatively high-latency connection (Gigabit Ethernet, for instance) between the various cores, broadcast the scene data over this connection, and then receive individual "chunks" of rendered pixels back. I defy you to do that with rasterization.
-:sigma.SB
Actually, no.
Rasterization is not embarrassingly parallel in the same way that raytracing is. Distributing tasks among those 800 "stream processors" is exceedingly complicated, because the underlying "task" involves iterating over every pixel that intersects with a given triangle rather than (as in raytracing) iterating over every triangle that could intersect with a given pixel.
-:sigma.SB
I wrote a whole long, insightful comment that went through all the technical issues on both sides, but Discussion 2.0 ate it and now I'm ticked off, so I'll just write the conclusion.
In raytracing, you deal with pixels, so you can get linear performance gains just by adding nodes (even over a slowish network like gigabit Ethernet) and divvying out pixels to work on. In rasterization, you deal with triangles, so you can't.
-:sigma.SB
Of course The End isn't in Metal Gear Solid 4. He died in 3, and that was a massive prequel. (Plus, he was already pretty ancient by then.)
-:sigma.SB
twm's window behavior is different from what most apps expect, enough so that sometimes they explode.
-:sigma.SB
(twm user who is forced to use sawfish to get good workspace support)
FUSE is quite stable. It's ZFS FUSE that's unstable.
-:sigma.SB
Your entire post is flawed. I know you're trolling, so I'm going to limit myself to one correction.
Ever heard of ECC? It's all the rage in storage media...
-:sigma.SB
So it'll eventually collapse into a singularity and suck up Earth? Wonderful.
-:sigma.SB
Konqueror lets you scroll up and down with the cursor keys (shift+cursor key has some neat effects, too) and make all links on-screen keyboard-accessible by tapping "control."
-:sigma.SB
Because that would invalidate the patent, silly.
-:sigma.SB
I must be tired, since I read "pollinators" as "politicians."
-:sigma.SB
We need a -1 Joke Ruined By Spelling Error mod.
-:sigma.SB
I don't use GIMP on the Mac for two reasons.
That said, I use and am satisfied with GIMP on Linux, simply because it's not that bad and there's no real alternative.
-:sigma.SB