I always noticed when I dl an episode of an anime that it always has something like
[ANBU-AonE]_Naruto_84_[47D87090].avi
What does the 47D87090 part mean exactly?
I always noticed when I dl an episode of an anime that it always has something like
[ANBU-AonE]_Naruto_84_[47D87090].avi
What does the 47D87090 part mean exactly?
Yeah i wanna know as well!Originally posted by: gaaralovessand
I always noticed when I dl an episode of an anime that it always has something like
[ANBU-AonE]_Naruto_84_[47D87090].avi
What does the 47D87090 part mean exactly?
But i'm guessing that its a reference number when a torrent is created.
CRC stands for Cyclic Redundancy Check - it's a method of creating a unique reference number for a file so you can verify that you have complete and accurate copy of it downloaded.
You can find more information using google, but here's a starting point....
http://www2.rad.com/networks/1994/err_con/crc.htm
http://www.users.cloud9.net/~stark/ctappc.pdf
CRC, breifly:
Take all the bits in a file and stack them ontop of each other. Use funny math to add them all up. Use funny math to invert the number, tack new number onto file. The new number is the CRC.
When a file is downloaded, the process is repeated, bits are stacked, funny math is used. If the number doesn't match what the file says, then there was a problem in the download somewhere.
K.
P.S. You'll also sometimes hear it referred to as a Hash check (I guess cause people are used to the phrase because of the bad Hash Check/redownload errors).
while crc is a form of hashing, I'll have to correct a technicality here (I'll leave the other not-quite-accurate points out for now).
Generally "Hash check/redownload" errors are in bittorrent, which doesn't use CRC. It uses SHA1, which is a significantly different hashing algorithm.
CRC is really simplistic. Because of the address space it uses, there are only slightly more than 4 billion possible CRC values. This means that if you CRC'ed 4.3 billion files, you'd have a guaranteed overlap on at least a couple of them. Seems like a lot, but address collisions (particularly engineered address collisions) can happen at much lower numbers (for example, HQA's encoder tweaked a lot of Da Capo episodes to have "meaningful" crc values like "DCDCDC18"). SHA1 has a much bigger address space, so is less likely o have address space colliions.
So CRC and MD5 are ways to check if a file's what it's supposed to be? How can I perform this check? Is there a way to just see it, or do I have to download some program to do it?
Presuming you're in windows:
using mirc: in a status window "//echo $crc($sfile(c:\))" and browse to the file.
for a dedicated checker, google "crc checker" or "sfv checker". Pick some freeware. I've used QuickSFV (which has some nice shell integration features) -- if you're using an sfv checker, you'd probably create a .sfv file (by running the checker on it) and comparing the crc listed in the file to the crc in the filename.
The actual algorithm to calculate a crc is very well-known and pretty trivial to code up, so there's a lot of implementations of it. Just look around and you'll find them.
(Incidentally, the same with md5sums, there's a lot of reasonably good implementations, just have to look).
In linux ... search your distro of choice's package manager for "sfv" or "checksum" for best results.
"Hash Calc" is a decent program if you only want to check one file quickly.
Thankyou very much. I went googling for some programs, and have found some ones greatly recommended by users and editors, including the two you have mentioned. Thankyou very much.
Posting Permissions
Reply With Quote