home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!mcsun!news.funet.fi!news.cs.tut.fi!mk59200
- From: mk59200@cs.tut.fi (Kolkka Markku Olavi)
- Newsgroups: comp.sys.amiga.hardware
- Subject: Parity checking (was: Re: Moving 500 to 1200, BAD NEWS)
- Date: 10 Nov 1992 12:07:12 GMT
- Organization: Tampere University of Technology
- Lines: 22
- Distribution: world
- Message-ID: <1do8lgINN6ti@cs.tut.fi>
- References: <OD.6badnetOA92-901-302p0_52d49fb1@piraya.bad.se> <1davn7INNfhf@cs.tut.fi> <H0mXr*h41@dynam.adsp.sub.org>
- NNTP-Posting-Host: mustarastas.cs.tut.fi
-
- In article <H0mXr*h41@dynam.adsp.sub.org> cbmehq!cbmswi!dynam!jmpor@cbmvax.commodore.com writes:
- >In article <1davn7INNfhf@cs.tut.fi>, Kolkka Markku Olavi writes:
- >> Also, parity checking doesn't _fix_ any memory errors, it only guarantees
- >> that the machine will _crash_ when a memory error happens. Apparently some
- >> people prefer this sort of behavior.
- >
- >with 10 bits you will be able to correct an 1 bit failure and detect a
- >2 bits failure.
-
- But that isn't parity, that's ECC. Besides, I believe that more than
- 2 check bits are needed for SECDED'ing a byte.
-
- I have used a PC with parity checking, and I didn't find the total
- system lock-up resulting from a parity check error very useful. Somebody
- said it prevents storing erroneus data to a file, but what if you crash
- during a disk write? You may get totally corrupted file system instead of
- a single bit error!
-
- --
- Markku Kolkka
- mk59200@cc.tut.fi
-
-