Date: Wed, 21 May 1997 11:44:00 +0800 (CST) From: Greg Lehey <grog@FreeBSD.ORG> To: thompson@squirrel.tgsoft.com (mark thompson) Cc: joerg_wunsch@uriah.heep.sax.de, freebsd-chat@freebsd.com Subject: Re: floppy flaky Message-ID: <199705210344.LAA00601@papillon.lemis.com> In-Reply-To: <19970520132403.10028.qmail@squirrel.tgsoft.com> from mark thompson at "May 20, 97 01:24:03 pm"
next in thread | previous in thread | raw e-mail | index | archive | help
mark thompson writes: > Date: Tue, 20 May 1997 08:50:19 +0200 > From: j@uriah.heep.sax.de (J Wunsch) > > As mark thompson wrote: > >> Pretty strange, that no errors are reported, huh? Floppy drive, cable, >> controller, software? > > Writing a floppy can only detect catastrophic errors (sector marks not > found). Data-field only errors remain undetected. > > Reformat your floppy. > > True. But the read should detect a bad checksum, not just return bad > bits, nicht wahr? And besides, i DID reformat the floppies, it still > failed pretty much the same. > > I'm flummoxed. I seem to recall things like this happening in the "good old days". We frequently had two different 5 1/4" floppy drives (360 kb and 1.2 MB) on our machines to get around the problem, which was related to different track widths for the two formats. Here's a scenario: somehow (not difficult), your floppy drive has got out of alignment. As a result, you're accessing it slightly off-track. Formatting it writes new tracks, slightly off the old tracks, but reading still "sees" the old tracks as well. So why don't you get CRC errors? Good question. My guess is that something is ignoring them. Greg
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199705210344.LAA00601>