Hi,

I'm looking to get a small C program compiled on 3B1 SysV Unix and get
the binary back? Can anyone with a working system help me out? I have
floppies with 3B1 development kit if anyone requires.

Thanks a lot
Atenox

On 2008-04-20, tenox <a...@tenoware.com> wrote:

        Well ... at the moment, I have a rack with a Sun Fire 280R, an
Exabyte 430 Mammoth-2 tape library, a Sun D-1000 JBOD full of SCA
drives, and a Criiterion/EMC Fibre Channel JBOD full of FC drives where
I use to have a table just big enough to support the 3B1 and the
external drive housing (modified from a 3B2 system).  Things are really
too crowded to do much here.  The system did run a bit less than a year
ago, before being moved into storage again.

        I hope that you get someone else who can do it for you.

        Is there any chance that this program will require libs which
don't exist in the 3B1?  One example of a simple program which can't be
compiled on the 3B1 is the "ping" net utility.  It requires "utime()"
which is not supported in the 3B1 kernelj, so the ethernet utilities for
the 3B1 lack that -- and I wondered why and tried to compile from net
sources.

        And looking up your domain I get quite confused.  ".tc" as the e-mail
domain for the administrator, but the supposed physical location being
in Vancouver BC Canada.  And an IP block in Ireland?  Very strange.  And
more so the deeper I dig. :-)

        What is this program supposed to do?  If it has anything to do
with crypto I would have to pass even if my machine were sitting there
ready to turn on.

        Good Luck,
                DoN.

--
 Email:   <dnich...@d-and-d.com>   | Voice (all times): (703) 938-4564
        (too) near Washington D.C. | http://www.d-and-d.com/dnichols/DoN.html
           --- Black Holes are where God is dividing by zero ---

I have a 3b1 available for doing compiles.  If you are interested, please
email me privately.  The 3B2 that was 'SDF' is accessible via TELNET and
you can 'cu' from there to the 3B1.

That's nice. I have few 220/420R and 250/450 boxes lying around in
various places. They are quite nice. What I don't like though is where
SUN is going with the OS. But perhaps it's the only way to fight the
plague.

I hope as well. I actually have two 7300 systems but they both DOA
with blown power supply. One day maybe someone will fix them up.

I don't think so. It did compile on quite few obscure systems around
already so I suspect there shouldn't be anything special required.

You're still missing few countries, but I wouldn't scratch it too
much ;)

No I leave cryptography to those who are in need of it... just this:
http://www.tenox.tc/out/#aclock

;)

Regards
A

On 2008-04-22, tenox <a...@tenoware.com> wrote:

        Well ... there are also several other OS's depending on your
tastes.  I run OpenBSD on some of the systems, and the number of
possible systems has just increased with 4.3, because they now support
more than one SPARC CPU.  I've got several machines which I will
probably try that on, with the Ultra 60 being the most likely first one.
The SB-1000s and the Sun Fire 280R are too busy actually *doing* things
under Solaris 10 to rip that out at present.

        Hmm ... you know that you could splice in a PC XT or later power
supply if you are willing to run it opened up.  The PS produces +5V,
+12V, and -12V -- all of which you can get from a PC power supply.  You
could even run a cable out the power cord connector/fuse holder hole in
the case.

        Hmm ... what happened to the power supplies?  One failure which
I have seen in the past is overheating on one or two of the power pins,
causing the solder on the PC board to melt and either drip off or turn
into a cold solder joint.  Clean the contacts and re-flow the solder
(perhaps with a bridge of solid copper wire on the underside of the
board) and you might bring the system back to life with an internal PS.
Unfortunately, the schematics for the PS are not available, making it
more difficult to fix a more serious problem.

        O.K.  And I just compiled it on Solaris 10 using gcc.  (Of
course, you can't compile it as a fully static program, since they don't
supply libc.a as a static lib -- only as a shared lib. :-(

        And it complies nicely with the standard C compiler from their
fancy development set as well.

        A pity I don't have a place to set up the 3B1.

        I stopped tracing after a while -- I was confused enough so
there was no point finding out anything else. :-)

        A neat program.

        Good Luck,
                DoN.

--
 Email:   <dnich...@d-and-d.com>   | Voice (all times): (703) 938-4564
        (too) near Washington D.C. | http://www.d-and-d.com/dnichols/DoN.html
           --- Black Holes are where God is dividing by zero ---

On Apr 22, 11:17 pm, "DoN. Nichols" <dnich...@d-and-d.com> wrote:

Sorry, to ask such a primitive question, but I've never owned a
desktop machine other than the 3B1. Is the 18-pin connector to the
system board ribbon cable a standard type used on other power
supplies?

Cool. I've been running OpenBSD on several Netras t1 I have. The only
problem I was having so far is their disk mirroring. As far as I can
remember there were three different ways of mirroring and none of them
were even close to something like Disk Suite on Solaris.

Didn't know about that. Interesting.

Just blown on power on. First one, I saw something flashed on the
screen and then kaboom! The second one actually started booting from
the disks and then, same story. I've gave it to an old computer/tv
repair shop and the guys looked at this and say FUBAR. There are many
parts blown up on the mainboard. Very sad story.

Surely there has to be a way of getting libc from opensolaris.org and
compiling it statically. Also I always wanted to build it as a stand-
alone binary so I can boot it from open prom. Shouldn't be too complex
having sources.

Yep.

A pity my two 3B1s blown up. :(

Thanks ;) I run it on few consoles around. Also I've made a version
for Novell NetWare so it can run on this begging for help never used
server console. If there are any NetWare servers still around...

Regards,
a

On 2008-04-23, Dennis Lefebvre <uni...@gmail.com> wrote:

        Nope -- it is unique to the 3B1/7300 -- at least as a power
supply connector.  PC power supplies used two shorter connectors side by
side on flat bladed pins, not the round or square ones which the 3B1 used.

        However, the pins are standard Molex style pins, and can be
acquired stand alone or in rows in a nylon carrier.  (That carrier will
probably melt if you solder wires directly to the pins unless you have a
heat sink of some sort clipped on the pins.

        As for the power connector layout --

        +12V    12, 14, 16
        +5V     2, 4, 6, 8, 10
        -12V    18
        GND     1, 3, 5, 7, 9, 11, 13, 15, 17  (All odd pins)

        The Technical Reference Manual does not show which end the pins
start at, so you'll have to measure resistance to the chassis mounting
screws, which are ground.  Check the underside of the power supply board
if it is removed from the chassis, because only one of the corner screws
may actually connect to ground.

        Alternatively, measure resistance from the first to the next to
last pin starting from both ends.  The ones which give a zero (or
extremely low) resistance will be the ground pins (the odd pins), and
you know that 1 is the one of those which is at the end of the array.

        As for the pinout of the XT style power supply, this can be
found by web searches, and I don't know it without doing such a search.
IIRC, the last of one connector and the first of the second are ground
(black wire), and red are +5V, yellow is +12V, orange might be -12V and
any other colors should be ignored. :-)

        I believe that the connector used on the stiff ribbon cable was
made by TRW.  I was able to find some at a hamfest, and have so far
replaced one badly burned connector on one system.

        Hold the connector in a drill press vise or something similar
(smooth jaws), and use a screwdriver to drive each wire (one at a time)
into the wire grip at the top of each pin.

        Oh yes -- if it is a 3B1 instead of a 7300, you will also
probably need a short extension for a drive power connector, as the hard
drive gets its power directly from the power supply, while on the 7300,
the power for the hard drive is routed through the system board.

        Good Luck,
                DoN.

--
 Email:   <dnich...@d-and-d.com>   | Voice (all times): (703) 938-4564
        (too) near Washington D.C. | http://www.d-and-d.com/dnichols/DoN.html
           --- Black Holes are where God is dividing by zero ---

On 2008-04-23, tenox <a...@tenoware.com> wrote:

        I've tried the RAID system on OpenBSD, and abandoned it.  The
same with Online Disk Suite on Solaris -- but on recent versions of
Solaris 10, there is a new system called zfs which is really easy to set
up and use.  I've got two zfs pools (clusters of disk drives treated as
a single resource) with a number of filesystems on each.  one is five 18
GB SCA drives with one hot spare serving to hold the multiple home
directories (mine, my wife's, and some for local guest accounts), and
the other containing lots of export filesystems inlcuding three
collections of photo images).  This second pool is on five active and
two hot spare 36 GB FC drives in an EMC FC JBOD.

        Assuming that the system board itself survived.  I fear that it
did not given the following description.

        What is the AC voltage where you are located?  (I've lost track
of your physical location).  Most 3B1s come with a single input
voltage power supply for 110-120 VAC input. If your local voltage is
instead 220-240 VAC, then yes, I would expect it to die quickly, and to
possibly take out componenets on the system board as well.

        If you get another one, and have 110-120 VAC available for it,
one recommendation is to (while the system is open) use a vacuum cleaner
on the ridges above the power supply.  This is actually ventilation for
the poweer supply, and if it is clogged with dust, the power supply will
overheat during long running.

        Yes -- you can build it -- if you build the whole system as part
of the task.  And I think that you may have to modify the Makefile to
keep it from deleting libc.a when it has done what it needs to do. :-)

        O.K.

        Enjoy,
                DoN.

--
 Email:   <dnich...@d-and-d.com>   | Voice (all times): (703) 938-4564
        (too) near Washington D.C. | http://www.d-and-d.com/dnichols/DoN.html
           --- Black Holes are where God is dividing by zero ---

In article <slrng0vppb.3eb.dnich...@katana.d-and-d.com>,
        "DoN. Nichols" <dnich...@d-and-d.com> writes:

Does anyone know the current requirements?  The fact they use so many pins
for each would make on think the current requirements are higher than what
a cheapie PC PS might provide.

If industry practice was followed the solder pad for pin 1 should be
square while the others are round.  I would still measure resistance
to a known ground point, but a square pad would be a good place to start.

bill

--
Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolves
billg...@cs.scranton.edu |  and a sheep voting on what's for dinner.
University of Scranton   |
Scranton, Pennsylvania   |         #include <std.disclaimer.h>