Date: Mon, 15 May 2006 00:00:41 EST-10EDT,10,-1,0,7200,3,-1,0,7200,3600
Subject: [UnixOS2_Archive] No. 701
**************************************************
Sunday 14 May 2006
Number 701
**************************************************
Subjects for today
1 gtak aspitape scsitape : Dan Drake"
2 Re: gtak aspitape scsitape : Steven Levine"
3 Re: gtak aspitape scsitape : Steven Levine"
4 Re: gtak aspitape scsitape : Dave Yeo"
5 Re: gtak aspitape scsitape : Dan Drake"
6 Re: link line lentgh? : Steven Levine"
7 Re: gtak aspitape scsitape : Steven Levine"
8 Re: gtak aspitape scsitape : Yuri Dario"
**= Email 1 ==========================**
Date: Sat, 13 May 2006 19:00:49 -0700 (PDT)
From: "Dan Drake"
Subject: gtak aspitape scsitape
Is GTAK or any of the things it uses alive in any sense at all?
I'm curious about why aspitape happily deals with an Exabyte 8505 but not
with a VXA-1 (it says the device type inquiry failed), though the specs show
them giving essentially identical responses to inquiry. Is there source,
expertise, or anything anywhere?
--
Dan Drake
dd at dandrake.com
http://www.dandrake.com/
porlockjr.blogspot.com
**= Email 2 ==========================**
Date: Sat, 13 May 2006 19:35:23 -0700
From: "Steven Levine"
Subject: Re: gtak aspitape scsitape
In <200605140200.k4E20nWi018898 at a.mail.sonic.net>, on 05/13/06
at 07:00 PM, "Dan Drake" said:
>Is GTAK or any of the things it uses alive in any sense at all?
Sure. It works fine with the stuff it works fine with.
>I'm curious about why aspitape happily deals with an Exabyte 8505 but not
> with a VXA-1 (it says the device type inquiry failed), though the specs
>show them giving essentially identical responses to inquiry.
Well, it could have something to do with that fact that the last update to
aspitape that I know of will be 10 years old in a couple of days.
What evidence do you have that the VXA-1 is actually working in your
setup?
It's impossible to say exactly why it fails in your case without something
more than the general info you have posted so far.
>Is there
>source, expertise, or anything anywhere?
Unfortunately, I don't think the source was ever openly available.
Are you sure you have the correct options on the device statement? I am a
bit surprised the a simple inq command fails. Do any of the info commands
work?
FWIW, with at bit of work you could use Paul Ratcliffe's apsiroute to talk
to the drive.
Regards,
Steven
--
----------------------------------------------------------------------
"Steven Levine" MR2/ICE 2.67 #10183 Warp/eCS/DIY/14.103a_W4
www.scoug.com irc.fyrelizard.com #scoug (Wed 7pm PST)
----------------------------------------------------------------------
**= Email 3 ==========================**
Date: Sat, 13 May 2006 19:54:48 -0700
From: "Steven Levine"
Subject: Re: gtak aspitape scsitape
In <200605140200.k4E20nWi018898 at a.mail.sonic.net>, on 05/13/06
at 07:00 PM, "Dan Drake" said:
>Is there
>source, expertise, or anything anywhere?
I hit enter a bit too soon. There are sources for GTAK 1.00 out there.
Google for gtak100s.zip.
Regards,
Steven
--
----------------------------------------------------------------------
"Steven Levine" MR2/ICE 2.67 #10183 Warp/eCS/DIY/14.103a_W4
www.scoug.com irc.fyrelizard.com #scoug (Wed 7pm PST)
----------------------------------------------------------------------
**= Email 4 ==========================**
Date: Sat, 13 May 2006 20:09:36 -0800
From: "Dave Yeo"
Subject: Re: gtak aspitape scsitape
On Sat, 13 May 2006 19:00:49 -0700 (PDT), Dan Drake wrote:
>Is GTAK or any of the things it uses alive in any sense at all?
>
>I'm curious about why aspitape happily deals with an Exabyte 8505 but not
>with a VXA-1 (it says the device type inquiry failed), though the specs show
>them giving essentially identical responses to inquiry. Is there source,
>expertise, or anything anywhere?
I have GTAK100S, GTAK212B and GTAK258. 100S and 212B both have some source though IIRC it does not compile as something is missing. Source is old (some as old as Jan, 1992) and some was compiled with the Zortech C which I have never heard of.
Anyways I don't remember where I found the older versions but could mail them to you if you want to have a look
Dave
**= Email 5 ==========================**
Date: Sat, 13 May 2006 21:20:51 -0700 (PDT)
From: "Dan Drake"
Subject: Re: gtak aspitape scsitape
On Sat, 13 May 2006 20:09:36 -0800, Dave Yeo wrote:
>On Sat, 13 May 2006 19:00:49 -0700 (PDT), Dan Drake wrote:
>
>>Is GTAK or any of the things it uses alive in any sense at all?
My thanks to you and to steve53 for the suggestions.
Actually, I should have asked whether they were alive in any
*maintenance* sense, since I know they're being used.
>>...
>
>I have GTAK100S, GTAK212B and GTAK258. 100S and 212B both have some
source though IIRC it does not compile as something is missing...
>Anyways I don't remember where I found the older versions but could mail
them to you if you want to have a look
Actually I have some old source, and I think that what's missing is *all* the
OS/2-specific part, which seems not to be available in source or object form.
This is unfortunate, but at least the problem has now been solved without
needing any code. It's worth sharing here because it's really dumb (on my
part), and after all, someone might be as dumb in the future.
When I got the drive, I left the unit number at its factory default, because
what the hell, it's as good as any other, and it certainly didn't conflict with
anything I had.
As I got into the car to hurry home for dinner after spending a bit too long
on this, I realized that unit 11 is indeed as good as any other, but a driver
from 1996 might not see it that way!
Yanked the unit-id jumper for bit 3, reducing the number to 3 effective bits;
changed the driver statement; rebooted; aspitape is serenely happy with the
drive. Works like a charm.
PS. I really do like SCSI. Not saying it's worth all the money these days, but
-- ten years and multiple technologies later, and an address-bit limitation is
the only incompatibility?
--
Dan Drake
dd at dandrake.com
http://www.dandrake.com/
porlockjr.blogspot.com
**= Email 6 ==========================**
Date: Sat, 13 May 2006 23:12:57 -0700
From: "Steven Levine"
Subject: Re: link line lentgh?
In <20060513043349.37CAEB8204 at generation.lgisp.net>, on 05/12/06
at 09:34 PM, "Dave Yeo" said:
>>I believe this this is shell dependent.
>Seems you are right
The brian-dead cmd.exe is the worst offender. 4os2 is better, but suffers
a bit from the choice to be upwardly compatible cmd.exe.
>05-12-2006 18:25:45 SYS3175 PID 027d TID 0001 Slot 00d8
>E:\OS2\CMD.EXE
As expected.
>I ended up going with MAXLINE 10240 as that is the default for most
>things and luckily jam supports SET JAMSHELL= so I set JAMSHELL=sh.exe
>and the compile finished.
There's another limit lurking in there. The kernel stuffs the command
line and the environment into a 64KB segment. For some reason things tend
to get wonky if the command line exceeds 32KB. This is probably a
holdover from 16-bit compatibility.
>At that it spat out a 70 MB file which shrunk slightly when running
>emxbind on it .
:-)
>ps I never have seen a sys3170 before and help says
>EXPLANATION: The process was ended because the program generated an
>unhandled fatal user (non-system) exception through DosRaiseException.
It's pretty rare. I suspect what happened is that the application
neglected to set up an exception handler. I'm not entirely sure, but I
think that 3170 means that the kernel expected the application exception
handler to handle the exception since the application issued the
DosRaiseException.
>Would this be an OS/2 error, a Innotek_libc error or a Watcom error?
Could be an application error or could be a libc error. Knut might know.
Regards,
Steven
--
----------------------------------------------------------------------
"Steven Levine" MR2/ICE 2.67 #10183 Warp/eCS/DIY/14.103a_W4
www.scoug.com irc.fyrelizard.com #scoug (Wed 7pm PST)
----------------------------------------------------------------------
**= Email 7 ==========================**
Date: Sat, 13 May 2006 23:53:41 -0700
From: "Steven Levine"
Subject: Re: gtak aspitape scsitape
In <20060514030914.F3D4FB8E9D at generation.lgisp.net>, on 05/13/06
at 08:09 PM, "Dave Yeo" said:
Hi Dave,
>I have GTAK100S, GTAK212B and GTAK258. 100S and 212B both have some
>source though IIRC it does not compile as something is missing.
I took a quick look at the 212b source. I all seems to be there. What
you might be missing is the DDK headers.
>Source is
>old (some as old as Jan, 1992)
It certainly is old. It was probably written for v1.x. The drivers are
written to the old bi-modal driver standard.
>and some was compiled with the Zortech C
>which I have never heard of.
You are young. It was one of the early high quality C compilers.
Certainly better than Lattice and the others of it's time.
>Anyways I don't remember where I found the
>older versions but could mail them to you if you want to have a look
Google finds them easy enough. I forgot what I had and went through the
exercise of finding them once again. :-) Turned out I already had 212b
and totally forgot that it included the source code.
Regards,
Steven
--
----------------------------------------------------------------------
"Steven Levine" MR2/ICE 2.67 #10183 Warp/eCS/DIY/14.103a_W4
www.scoug.com irc.fyrelizard.com #scoug (Wed 7pm PST)
----------------------------------------------------------------------
**= Email 8 ==========================**
Date: Sun, 14 May 2006 10:33:36 +0200 (CDT)
From: "Yuri Dario"
Subject: Re: gtak aspitape scsitape
Hi,
>It certainly is old. It was probably written for v1.x. The drivers are
>written to the old bi-modal driver standard.
I think it is not worth to recompile device drivers: we have a supported aspi router
(aspirout.sys), so probably it is better to modify tape support application code to support this
driver.
This is my current plan for star backup, but since I don't do tape backups it is in the (very) slow
queue. ah, that's reminds me the old Digital VAX times...
Bye,
Yuri Dario
/*
* member of TeamOS/2 - Italy
* http://www.os2power.com/yuri
* http://www.teamos2.it
*/