Xenix vs. Unix?

[Karl Denninger]

In article <26 at fleet.UUCP> mel at .UUCP () writes:
>In article <34 at nstar.UUCP> larry at nstar.UUCP (Larry Snyder) writes:
>>> Xenix also works with nearly everyone's tape drives.  386/ix works with
>>> Wangtek only.  I could go on; the hardware support on 386/ix is "somewhat
>>> lacking" IMHO.
>
>SCO Xenix isn't all that great at tape drive support either.  Ever since I 
>switched to 2.3.1 Xenix it takes TWO attempts at starting the tape drive.
>(I'm using an Archive 150 meg/QIC02 internal on my HP Vectra 386/RS20) 
>(I also tried an Emerald controller as well as the original Archive unit)

We don't seen anything like this.  We have both 60 and 150 MB archive drives
in the field, with both long and short controller boards.  No problems 
reported.

Let's not even mention the Irwin 145 series, which SCO also supports.  And
that even works (which, given the brain-dead nature of the Irwin drives, is
amazing in itself).  We used one of those for several months before we got
a QIC/02 unit here.

>I recently called SCO support after going trough numerous checks of addresses.
>interrupts, etc.  Their answer was "O'yeah, we do have a few problems with
>some tape drivers.  They're on our 'list', but I'm not sure when (or if)
>they're going to fix them".
>
>Great!!  I can't wait to tell my customers about this.  Just start up your
>backups twice because the first time there's an allergy between the tape
>unit and the CPU so it requires a smooth-out-the-wrinkles false start.  
>BTW 'tape reset' hangs the tape drive so completely that the CPU has to be
>rebooted.  Would you trust any backups that require this type of maneuvering
>to make it go? 

'Tape reset' hangs the drive?  Then you definately have problems or your
driver/board isn't configured right.  I use "tape reset" all the time, and 
it has never hung a drive or controller, or a machine (with the exception of 
when our 60MB QIC/02 controller really did die, then it hung the entire 
system!  Can't blame SCO for that one, however, the controller was smoked.)

>Archive wasn't the only one with problems either.  The rep was hard pressed
>to give me an answer on just one "brand" that worked correctly.

Well, Wangtek, Archive VP series, and the older SC499-based units (long
Archive controller boards) all work without any "false starts" in our
systems, with both the Televideo (which we used to sell) and our current line 
of clones (Legend).   Believe me, we'd hear it if this kind of problem occurred
with our customers; they're "office people" and when there is an error
message, or a (horror) hung process we get a phone call!

We're running 2.3.2 and 2.2.3 (one instance), as are our customers.  
Perhaps you need an update.

2.3.1 (one) did have problems, but not that kind (that I remember anyway).
It would, if memory was fragemented too much, return a "system too busy"
message (this was one that really pissed me off).  2.3.2 seems to have fixed
that one.

>Perhaps by letting "you" know what's going on some action will be taken on this.
>It seems like having reliable backups should require problems of this nature
>to be given the highest priority.

Our backups are highly reliable.  In fact, with the ecc tape devices
("erct0") we haven't had a single bad tape yet on read-back -- and we quite
often do need to restore the things (a result of playing with partition
tables, driver development :-), etc.

386/ix, on the other hand, recently was unable to restore a tape that was
made a few hours prior on 2.0.1; this was with a 60MB Wangtek.  Our customer
was not amused.  There is no "ecc" tape device on 386/ix.

--
Karl Denninger (karl at ddsw1.MCS.COM, <well-connected>!ddsw1!karl)
Public Access Data Line: [+1 312 566-8911], Voice: [+1 312 566-8910]
Macro Computer Solutions, Inc.		"Quality Solutions at a Fair Price"