Subject: RE: Conner minicartridge tape drive - quirky drive
To: Lindgren, Jon <jlindgren@espus.com>
From: Hauke Fath <hauke@Espresso.Rhein-Neckar.DE>
List: current-users
Date: 07/25/1999 23:57:03
At 12:39 Uhr +0200 23.07.1999, Lindgren, Jon wrote:
>
>
>> -----Original Message-----
>> From: Hauke Fath
>> According to my humble opinion, the NetBSD tape driver is in
>> a state of bit rot ("maintenance mode" in Apple diction).
>
>In terms of functionality, code, kludge factor, or what?

Well - "bit rot" implies for me that there is noone left who really feels
responsible for the code. Wrt. the tape driver there are lots of "quirks"
but no description about the "quirks" mechanism (kern/7734), there is a
hopelessly outdated density table from a SCSI II document (the
corresponding SCSI III and QIC documents have been available for a while)
and there is a remark wrt Tandberg 4xxx drives and "missing documentation"
which is available online from Tandberg. People seem to hack on the code
when they feel like it...

I looked at the FreeBSD CAM tape driver code for comparison -- boy, what a
difference!

>It should be relatively easy to fix if there's a problem...

When you understand a) the problem and b) the SCSI framework. I found that
at least for me, there is a long way to go. For CAM, there is at least
_some_ documentation...

Sigh.

	hauke


--
"It's never straight up and down"     (DEVO)