4 Dec
2009
4 Dec
'09
7:31 p.m.
On Fri, Dec 4, 2009 at 1:21 PM, Chuck Guzis <cclist at sydex.com> wrote:
My gripe in general with most Travan and QIC-40/80
consumer-level
devices. A separate verify pass is necessary, which takes about as
much time as the writing pass, so most users simply never bothered.
And if an error was discovered on the tape, the entire tape had to be
rewritten, a real disincentive to verifying if there ever was one.
An awful lot of data was written in drifting sand as a result.
Back in the early 90s, I had a job installing computers in dentist offices.
It was a dos based application with an early version of novell. So there'd
be half a dozen ISA based dos machines with 10base2 network boards. The
novell servers would have QIC drives in them (iirc). The machines had an
optical sensor with a light. When the tape neared the end of the reel,
holes in the tape would pass between the light and detector telling the
drive that the end was near. These things would fill up with dust and
obscure the light or sensor. So it wouldn't detect the end of the reel and
wind the tape all the way off the other spool. The tape wasn't attached at
all, just wound around the reel with enough extra to hold it in place. I
don't know how many times I've opened those QIC cartridges and crazy-glued
the tape back onto the reel after blowing the dust out of the drive. They
didn't want their data back. They were just too cheap to buy new
cartridges. :-/
brian