Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!usc!ucsd!nosc!humu!pegasus!richard
From: richard@pegasus.com (Richard Foulk)
Newsgroups: comp.unix.sysv386
Subject: Re: tape drive info wanted.
Message-ID: <1991May2.123012.17083@pegasus.com>
Date: 2 May 91 12:30:12 GMT
References: <1991Apr8.124255.26553@qut.edu.au> <1991Apr25.015612.1847@pegasus.com> <1991Apr25.174716.3500@ico.isc.com>
Organization: Pegasus,  Honolulu
Lines: 57

>
>(regarding 8mm tapes)
>> Don't buy certified data tapes -- they're a lot more money for the same
>> item.  Regular Sony video tapes perform just as well...
>
>They're not (quite) the same item.  The data-grade tapes come out of the
>same manufacturing process as the video tapes, so in that sense they're the
>same--but the data-grade tapes are selected.  A tape that fails the data-
>grade test becomes a video tape.  [...]

The tapes in question are (as far as I'm aware) made by one outfit
(Sony) for video use, and verified and sold as data tapes by another
outfit.  The data tape market is so minuscule compared to the video
tape market that this graded-product approach just doesn't seem likely
at the manufacturing level (if that's what was meant).

> [...]                        Still, the video tape is acceptable.
>You'll get more errors--normally all recoverable, but since a bad spot
>causes a skip on the tape, you get somewhat reduced capacity and slower
>overall data rate.

From some testing reported in Sun-spots, over a period of a couple of
years, the general consensus was that standard Sony video tapes
suffered no disadvantage to the "data grade" tapes.  (The Exabyte drive
counts the number of errors it corrects, given support in the device
driver it can be accessed by the user -- thus you can do your own tape
grading.)

>> (Error correction capabilities on the Exabyte drives are very
>> impressive and seldom used -- at least with Sony tapes.)
>
>Yes and no...they *are* impressive, but they do get used fairly often.  You
>don't normally see the error correction done with ECC, but you will see the
>bad-spot skipping.  The error correction has to be good because it does get
>used.

The testing I mentioned earlier showed that many Sony tapes were error-free.
And that most had less than a handful of errors.  (This is recollection from
a while back -- I'll go find the data and post it when I get a chance.)

I know of several large computing centers that buy video-grade Sony
tapes by the case, and consider the "data grade" tapes as nothing but a
ripoff directed at the uninformed.

If you must, get a device driver that returns the error count, and do
your own tape grading -- if the error count is higher than you'd like
on a backup then use a fresh tape and do it again.  Keep in mind that
using the worst video-grade tapes available it's unlikely that you'll
actually lose any data.

[I have no connections with Exabyte or Sony, never have.  I don't even
currently have an Exabyte drive, though I wish I did.  I've used them
in the past and had nothing but good results.]


-- 
Richard Foulk		richard@pegasus.com