Citiraj:
Originally posted by msvrtan
ma ja se odpocetka slazem s time sto nara kaze oko kvalitete kablova..nisam se slagao oko onoga s korumpiranim podatcima...
i ja bih rado kupio 4 kabla po 80kn (mislim da su cooler masterovi toliko) ali ako dodan na tu cifru jos par sto kuna eto ti novoga diska i sl...
|
sada malo i argumentiranja mojih tvrdnji, za referencu sam
koristio draft ata-5 standarda sa sluzbenih stranica tijela koje
se brine o predlaganju standarda za komunikaciju IDE uredjaja.
dokument mozete naci na slijedecoj adresi:
http://www.t13.org/project/d1321r3.pdf
a radi se o 350+ stranica detaljne specifikacije po kojoj je
donesen ansi standard (a koji se placa, pa zato navodim draft)
ukratko, nigdje unutar dokumenta nisam nasao ista sto bi me
uvjerilo da je data transfer nepogresiv, citiram (str 375):
Once the burst is terminated and the host sends the CRC data
to the device (the host always sends the CRC independent of
whether the burst was a data in or data out transfer), the device
compares this to the CRC it has calculated. While other CRC
validation implementations may be possible, a CRC input register
may be used on the device in combination with a digital
comparitor to verify that the CRC value in the input register
matches the value in its own CRC calculation register.
na strani 286 mozete procitati da se radi o 16-bit CRC provjeri.
CRC nije nepogresiv, za primjer navodim slijedece:
With Cyclical Redundancy Check (CRC) technique, CRC adds 8, 16,
24, or 32 bits to the message. A communication protocol using a
16 bits CRC calculates a 16 bit number that is a function of all the
data in the message. This 16 bits number is added to the end of
the message block. The receiver recalculates its own 16 bit CRC
as the block is received. If the number are the same, the
probability is very high that the data is error free. If they are
different, an error has occurred. The number of undetected errors
may be as low as 1 in 10 9 characters. To put this in more familiar
terms: On a 9600 bps circuit transmitting 8 bit characters 24
hours a day, one would expect no more than 1 undetected error
every 231 + hours, or 9.6 days.
sa adrese:
http://home.apu.edu/~jcox/projects/individual/Vu/
naravno, moguce je da sam u krivu, ali molim vas da takve
tvrdnje popratite odgovrajucim dokazima.