Need help to reanimate an AT&T AT&T 3B1/7300 UNIX-PC
Juergen Sievers
Hallo,
Peter da Silva (he publishes this (http://unixpc.taronga.com/)
wonderfully web suites) informed me about this group. So I post the
whole request to Peter again here because he does not own an UNIX-PC
and so he could not help me.
--- snip ---
|
|I no longer have a UNIX PC. I would suggest you try the UNIX-PC
Usenet |group comp.sys.3b1.
|
|
|On 2008-03-25, at 14:25, Juergen Sievers wrote:
|| Hi,
||
|| I have found your pages on the internet and so I want to ask you
for
|| some help to repair an AT&T 3B1/7300 computer.
||
|| I bought such system on eBay and it was shipped from Canada to
|| Germany. May be on this trip the system's HD had got a head-crash.
|| Anywhere after it has arrived it does not boot. Some times, after
|| 20-30 retries it comes up to the console. But I have no plan to
create
|| a boot disk eg. Fsck always tries to repair something. But after
|| repair the system still does not boot proper.
|| After some sectors have been loaded I get a Disk Read Error
message.
||
||
||
|| After opening the pizza box I could see the battery is missing and
one
|| of the electronic parts are burn down completely. Only the both
|| connection wire was remaining. See attached picture (my own hand
made
|| art).
||
||
||
|| So I need every kind of information to get the system back to
running
|| state.
|| What kind of electronic part was burned and how can I get a boot
and
|| installation disk for that system.
||
||
||
|| Could you help me?
||
||
||
|| Regards
||
|| Juergen
||
|| ---
|| http://www.nadhh.hanse.de.com/10subm_Computer/
|| <DSC01196.JPG||<AT&TMissigPart.PNG||
|
--snap -
Juergen Sievers
DoN. Nichols
> Oh, sorry, the pictures are here now :)
>
> http://www.nadhh.hanse.de.com/eBay/AT&T.UNIX-PC/
Hmm ... that is an unfortunate location in that there were
several versions of the CPU board, and on some there were empty pin
layouts which have been used for such things as modifying the system to
accept two hard disk drives of up to 160 MB formatted (190 MB
unformatted) -- drives which were made by both Maxtor and Piram (I have
one of the latter, and have worked with a couple of the former.)
Obviously, only one of the drives could be internal. The drives were
all MFM drives, and to handle drives that large, it was necessary to
replace the hard disk controller IC with a later model. I would have to
rip a machine back apart to tell precisely what was there -- and it is
likely to be different on different versions of the system board.
1) 512 KB of RAM on board --7300 only.
2) 1 MB of RAM on board -- 7300 and 3B1 I believe.
3) 2 MB of RAM on board -- 3B1 only I believe.
4) Export (Not sure whether there were three versions of this, but
the modem and telephone connectors were disabled in this, either
because of differing modem standards, or just because some of the
countries did not want their citizens to have access to modems.
Which you got will depend on where you got it, among other
things. If you got it from an eBay auction in the USA, it probably has
the modem jacks active.
Now -- the position of the component and the fact that it is
only two pins suggests that it is likely to have been a bypass
capacitor -- many were scattered around the system board to keep logic
chips from injecting noise into the power supply lines. You can verify
this by measuring (with a low-current ohmmeter) whether one side is
connected to the ground pins on the disk drive connectors (black, of
course), and whether the other side is connected to the +5V (red of
course). It was likely to be something like a 47 pF 50 V capacitor or
something similar. Hmm ... maybe 100 pF. But it might be one of the
relatively rare electrolytic capacitors distributed around as well.
The missing battery should be a 3V coin cell. The original had
solder tabs, but I tended to replace them with sockets for the 2032
cells. The major problem without that is that you will have to reset
the time every time you boot, and will wind up with a few files with
very strange dates before the boot reaches the point to set the time.
Now -- from the rather oversized image of the system, what you
have seems to be a 3B1, as it has the raised area under the monitor stem
to clear the height of a full-height disk drive. The 7300 had only a
half-height drive. (Also, the file name for the missing part image was
a pain, with an '&' embedded, making it a pain to handle on a unix
command line.
As for your problem with booting -- the typical failure status
is for the inodes pointing to directories, or sectors making up part of
directories to be lost. resulting in the loss of quite a few files.
Many of them will be found and put in the lost+found directory by fsck,
but they will have names starting with a '#' and followed by the inode
number for the file, and you'll have to do a lot of detective work to
dig it out.
So -- what you need is to boot from the first floppy of the
installation set, and progress from there, including booting first from
the diagnostics set to cleanly format the disk first. Note that the
probable cause of the bad sectors on the disk is oxidation on the power
supply connector. Look for areas on the connector which have turned
somewhat brown -- or at worst have blackened enough so that when I
installed a replacement connector, the old one fell into two pieces.
You'll need a *good* contact cleaning preparation. Back when we were
working with these, the preferred one was "Cramolin" from Cano labs, but
that has been replaced since and there may be better ones by now. Spray
the pins and into the cable connector's sockets, slide it onto and off
of the row of pins sticking up from the power supply several times, wipe
the pins dry and re-spray them. Also, it does not hurt to look at the
underside of the power supply board, because sometimes the heat from the
oxidized pin will result in the solder melting and crystalizing, so
removal of the solder and replacement with fresh lead-tin solder (not
the lead-free stuff made these days) before you re-format the drive.
Now -- the installation set puts in most of what you need,
except a good editor and no compiler. And there was one floppy pair
which was not to be included with exports -- it had the crypto software
and the programs which used it (like an encryption capable vi editor) --
so it could be left out of export systems.
But you'll also want a set of floppys containing the
"Development Set", which supplies the compiler, the header files, and an
encryption-free version of vi, among other things. Note that if you
*do* have the encryption floppies, you want to install the development
set first, then the encryption set, because it looks for the presence of
the plain vi and only installs the encrypted version if the plain is
already installed. It is possible to fool it into installing the
encrypted version if you put a file of the right name in the right
place, but I forget whether it was in /bin or /usr/bin .
If you find a place to download the floppies, beware that the
Unix-PC used two non-standard floppy formats. The typical MS-DOS
machine of the period put nine sectors on a track. The boot floppy and
some others had eight sectors per track, and the rest had ten sectors
per track -- making it difficult to build the floppies on anything other
than a 3B1. Some linux systems of the period had optional floppy
drivers which could generate and work with the odd sectored floppys, but
I never worked with that so this is only hearsay.
The "3.51" on the screen shows that it had the latest version of
the OS installed -- except the patched one. It also suggests that this
one was modified to handle two disk drives -- so look for ways to get
the extra cables out to an external drive. Everybody did this
differently. I managed to work up a way to put both drives in an
external housing, and route both data cables through one ribbon cable.
If you have extra cards in the base you may need extra drivers --
for things like the floppy tape, or the ethernet card, and for the combo
card, which could hold up to 1.5 MB of RAM plus supply two extra serial
ports. You can have only one fully-loaded combo card -- and no
combination of cards which add up to more than 2 MB or you will confuse
the system.
Oh yes -- if you do have the modifications to allow a second
hard drive, you'll need the updated diagnostics disk to format the
second drive.
But given your museum focus, I suspect that you will want it as
close to factory supplied as possible, rather than as commonly modified
by the user community.
Best of luck,
DoN.
--
Email: <dnic...@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 ---
Dennis Lefebvre
> You'll need a *good* contact cleaning preparation. Back when we were
> working with these, the preferred one was "Cramolin" from Cano labs, but
> that has been replaced since and there may be better ones by now. Spray
Cramolin still seems to be available in Germany...
Juergen Sievers
> On 2008-03-25, Juergen Sievers <JSiev...@nadhh.hanse.de.com> wrote:
Hallo DoN.
Thank you very much for your big replay.
To give you more information about the system I added some more high
resolution pictures to the following location.
http://www.nadhh.hanse.de.com/eBay/AT&T.UNIX-PC/
>
> Hmm ... that is an unfortunate location in that there were
> several versions of the CPU board, and on some there were empty pin
> layouts which have been used for such things as modifying the system to
Oh, you are right; there are some unused areas on this location. So I
have taken a detailed picture from it. But I believe too, there must
have been a capacitor on this location.
> accept two hard disk drives of up to 160 MB formatted (190 MB
> unformatted) -- drives which were made by both Maxtor and Piram (I have
It is a ST51 (64MByte MFM Drive) and the controller chip is an
WD1010A.
> one of the latter, and have worked with a couple of the former.)
> Obviously, only one of the drives could be internal. The drives were
> all MFM drives, and to handle drives that large, it was necessary to
> replace the hard disk controller IC with a later model. I would have to
> rip a machine back apart to tell precisely what was there -- and it is
> likely to be different on different versions of the system board.
>
> 1) 512 KB of RAM on board --7300 only.
>
> 2) 1 MB of RAM on board -- 7300 and 3B1 I believe.
>
> 3) 2 MB of RAM on board -- 3B1 only I believe.
>
On the main board are eight rows a (8+1) 256er RAM chips installed.
> 4) Export (Not sure whether there were three versions of this, but
> the modem and telephone connectors were disabled in this, either
> because of differing modem standards, or just because some of the
> countries did not want their citizens to have access to modems.
>
> Which you got will depend on where you got it, among other
> things. If you got it from an eBay auction in the USA, it probably has
> the modem jacks active.
>
Yes, the system was used in USA and I’m sure the modem line is active.
As you can see on the pictures I own a expansion box too, on which are
one more modem line is available. But I never tried this box yet.
> Now -- the position of the component and the fact that it is
> only two pins suggests that it is likely to have been a bypass
> capacitor -- many were scattered around the system board to keep logic
OK, I decide it must be a capacitor. One like all the others in front
of each DIL-IC.
> The missing battery should be a 3V coin cell. The original had
> solder tabs, but I tended to replace them with sockets for the 2032
> cells. The major problem without that is that you will have to reset
> the time every time you boot, and will wind up with a few files with
> very strange dates before the boot reaches the point to set the time.
OK, I saw this message already and I’m glad to hear that there is no
setup or configuration SRAM which I have to set up proper first.
> Now -- from the rather oversized image of the system, what you
> have seems to be a 3B1, as it has the raised area under the monitor stem
> to clear the height of a full-height disk drive. The 7300 had only a
> half-height drive. (Also, the file name for the missing part image was
> a pain, with an '&' embedded, making it a pain to handle on a unix
> command line.
I saw a file named “unix” on the root as also a link to it with a
uppercase name like “UNIX351a” or so. See pictures of screens.
>
> As for your problem with booting -- the typical failure status
> is for the inodes pointing to directories, or sectors making up part of
> directories to be lost. resulting in the loss of quite a few files.
> Many of them will be found and put in the lost+found directory by fsck,
> but they will have names starting with a '#' and followed by the inode
> number for the file, and you'll have to do a lot of detective work to
> dig it out.
I’m willing to do all the work if the system would be get back to a
working state:)
But I can’t boot because I have no boot and installation disk now.
Some times after many retries the system has been booted from the HD.
But on that case many errors are appearing and I can not switch to the
graphical mode e.g.
> So -- what you need is to boot from the first floppy of the
> installation set, and progress from there, including booting first from
Oh yes, I would do that immediately if I would have a bootable floppy
disk
I have some other old (MFM) systems which are able to write many
exotics disks formats on a 51/4” disk. But for that I need the images
of the AT&T’s boot and installation disks. Do you or somebody else can
send a dd image to me?
> Now -- the installation set puts in most of what you need,
That would be wonderfully if some installation set would do that :)
But I haven’t any installation set available yet
> except a good editor and no compiler. And there was one floppy pair
> which was not to be included with exports -- it had the crypto software
> and the programs which used it (like an encryption capable vi editor) --
> so it could be left out of export systems.
OK, I will not use the system for real work anymore. It should only
become a very nice piece of my collection of running systems.
>
> But you'll also want a set of floppys containing the
> "Development Set", which supplies the compiler, the header files, and an
Yes, I really needing such floppies
> encryption-free version of vi, among other things. Note that if you
> *do* have the encryption floppies, you want to install the development
> set first, then the encryption set, because it looks for the presence of
> the plain vi and only installs the encrypted version if the plain is
> already installed. It is possible to fool it into installing the
> encrypted version if you put a file of the right name in the right
> place, but I forget whether it was in /bin or /usr/bin .
Good hint, thanks a lot.
> If you find a place to download the floppies, beware that the
> Unix-PC used two non-standard floppy formats. The typical MS-DOS
I’m full of hope you are such a place for me, isn’t ?
> machine of the period put nine sectors on a track. The boot floppy and
> some others had eight sectors per track, and the rest had ten sectors
You are sure? I remember I was reading something about ten sectors???
> per track -- making it difficult to build the floppies on anything other
> than a 3B1. Some linux systems of the period had optional floppy
> drivers which could generate and work with the odd sectored floppys, but
> I never worked with that so this is only hearsay.
I use a old DOS machine with some tools from the good, old CP/ M era.
These tools can read and write many old disk formats. Hope it works
for my UNIX-PC too.
> But given your museum focus, I suspect that you will want it as
> close to factory supplied as possible, rather than as commonly modified
> by the user community.
Yes, you are right. I prefer systems in their original state as like
as they were delivered.
>
> Best of luck,
> DoN.
Thanks and same return to you
Juergen
PS: Sorry about my bad English. My brain (if there is any) may be to
small to learn it much better. Abe rich kann etwas Deutsch, wenn das
helfen kann?
DoN. Nichols
> On 26 Mrz., 03:50, "DoN. Nichols" <dnich...@d-and-d.com> wrote:
>> On 2008-03-25, Juergen Sievers <JSiev...@nadhh.hanse.de.com> wrote:
>
> Hallo DoN.
>
> Thank you very much for your big replay.
> To give you more information about the system I added some more high
> resolution pictures to the following location.
>
> http://www.nadhh.hanse.de.com/eBay/AT&T.UNIX-PC/
Again -- an unfortunate directory name this time. Please use
"-and-" in place of the "&" symbol since unix shells interpret the '&'
on a command line as the end of one command to be run in the background,
and the (possible) start of another, so when I cut and pasted that URL
to a command line to start up my browser pointed to that web page. it
tried to find one named:
http://www.nadhh.hanse.de.com/eBay/AT
and then tried to execute:
T.UNIX-PC/
as a command, which of course failed.
>>
>> Hmm ... that is an unfortunate location in that there were
>> several versions of the CPU board, and on some there were empty pin
>> layouts which have been used for such things as modifying the system to
> Oh, you are right; there are some unused areas on this location. So I
> have taken a detailed picture from it. But I believe too, there must
> have been a capacitor on this location.
>
>> accept two hard disk drives of up to 160 MB formatted (190 MB
>> unformatted) -- drives which were made by both Maxtor and Piram (I have
>
> It is a ST51 (64MByte MFM Drive) and the controller chip is an
> WD1010A.
O.K. That sounds pretty close to the standard for the 3B1
-- 67MB formatted for the system, 80MB unformatted.
[ ... ]
> On the main board are eight rows a (8+1) 256er RAM chips installed.
Looking at the screen shots, plus the shots of the back of the
system, I would have to say that you have the full 2MB on the system
board, and 1.5 MB in a COMBO board (the one with the two DB-25
connectors on it). The other board is the ethernet board. You'll need
a transceiver to plug into the AUI connector visible and to accept
either 10BaseT (twisted pair with a RJ-45 connector) or 10Base2 thinnet
(BNC connector) depending on which network you happen to have.
>> 4) Export (Not sure whether there were three versions of this, but
>> the modem and telephone connectors were disabled in this, either
>> because of differing modem standards, or just because some of the
>> countries did not want their citizens to have access to modems.
>>
>> Which you got will depend on where you got it, among other
>> things. If you got it from an eBay auction in the USA, it probably has
>> the modem jacks active.
>>
> Yes, the system was used in USA and I??m sure the modem line is active.
> As you can see on the pictures I own a expansion box too, on which are
> one more modem line is available. But I never tried this box yet.
Actually -- what is likely to be in the expansion box is a Voice
Power card -- which was used to make answering machines (sometimes
multi-line -- one per card) from the computer -- and the right
software.)
>> Now -- the position of the component and the fact that it is
>> only two pins suggests that it is likely to have been a bypass
>> capacitor -- many were scattered around the system board to keep logic
> OK, I decide it must be a capacitor. One like all the others in front
> of each DIL-IC.
That is the most likely. The photo was too dark and too
out-of-focus in the back corner for me to get anything useful from it.
Will your Sony camera focus closer?
However -- looking at the CPU on a daughterboard with a PAL
beside it suggests that this has been modified to allow a user-level
program to do graphics access to the system to allow an alternative
windowing system to work. I forget the name of it, but AT&T was making
it available in source code format, and some people played with it.
>> The missing battery should be a 3V coin cell. The original had
>> solder tabs, but I tended to replace them with sockets for the 2032
>> cells. The major problem without that is that you will have to reset
>> the time every time you boot, and will wind up with a few files with
>> very strange dates before the boot reaches the point to set the time.
>
> OK, I saw this message already and I??m glad to hear that there is no
> setup or configuration SRAM which I have to set up proper first.
I see that there have been some red and black wires soldered to
where the cell goes, but it is not clear what else was done.
It looks as though the power connectors (both the one on the
stiff ribbon cable from the system board, and the row of pins on the
power supply) look to be in pretty good condition -- but I would still
suggest using the contact cleaner on it just to be sure that the power
is clean to the system or you will be constantly re-formatting the disk
drive.
>> Now -- from the rather oversized image of the system, what you
>> have seems to be a 3B1, as it has the raised area under the monitor stem
>> to clear the height of a full-height disk drive. The 7300 had only a
>> half-height drive. (Also, the file name for the missing part image was
>> a pain, with an '&' embedded, making it a pain to handle on a unix
>> command line.
>
> I saw a file named ??unix?? on the root as also a link to it with a
> uppercase name like ??UNIX351a?? or so. See pictures of screens.
Yes -- that is the kernel and is what is normally booted. If
you have the enhanced diagnostics floppy (and that one I am sure was
available on the net for download) it can install an alternate bootblock
which will boot from one of several diagnostics programs, the primary
one of which is in a file called "s4test". There were alternate
diagnostics on the installation floppys for things like the ethernet
board and the COMBO card among other things. I tended to keep them in a
subdirectory which I called /kernels so I could boot from one of them
easily -- instead of having to dig out the proper floppy.
Note that the s4test diagnostic program is needed to format your
disk -- there is no program on the installed OS to do it. And the
install floppy had a copy of s4test in it for formatting the disk during
installation -- but not the one which could handle formatting a second
disk.
>>
>> As for your problem with booting -- the typical failure status
>> is for the inodes pointing to directories, or sectors making up part of
>> directories to be lost. resulting in the loss of quite a few files.
>> Many of them will be found and put in the lost+found directory by fsck,
>> but they will have names starting with a '#' and followed by the inode
>> number for the file, and you'll have to do a lot of detective work to
>> dig it out.
>
> I??m willing to do all the work if the system would be get back to a
> working state:)
> But I can??t boot because I have no boot and installation disk now.
> Some times after many retries the system has been booted from the HD.
> But on that case many errors are appearing and I can not switch to the
> graphical mode e.g.
I noticed that -- including one boot attempt trying to find a
program ".!" There was one in the system called ".!." which popped up a
window and scrolled the names of everybody who worked on the design
project through the screen.
>> So -- what you need is to boot from the first floppy of the
>> installation set, and progress from there, including booting first from
>
> Oh yes, I would do that immediately if I would have a bootable floppy
> disk??
> I have some other old (MFM) systems which are able to write many
> exotics disks formats on a 51/4?? disk. But for that I need the images
> of the AT&T??s boot and installation disks. Do you or somebody else can
> send a dd image to me?
Well ... about a year ago, I was building a directory of images
of the floppys -- but that seems to have been lost in a disk crash, and
I currently don't have room to set the system up again to repeat the
process. I had been planning to burn a CD-ROM or DVD-ROM for that, but
the crash came before that got done.
I guess that if all else failed, I can move the two Sun
Ultra-60s to clear sufficient space. The system is one which I had
modified to have two hard drives -- a 161 MB and a 67 MB (formatted).
Note, BTW, that you will also have problems with the Y2K situation. The
program which sets the unix time can't handle dates which start with
"20" in the year. I wrote a pair of programs -- one to run on the Suns
and to spit out the time as the number of seconds since the start of
1970 (as unix keeps its time), and the other to run on the 3B1 and to
re-install that time to set the date on that system. I was working on
other things for the time when I got the Sun Fire 280R which needed
the space where the 3B1 had been sitting when running.
But -- I remember someone else having a web site with downloads
for many of the floppys -- and I even supplied him with one or two which
he was missing. You may find it by going through the past two or three
years of the archives of this newsgroup. I'm pretty sure that it was
mentioned here and within that period.
>> Now -- the installation set puts in most of what you need,
> That would be wonderfully if some installation set would do that :)
> But I haven??t any installation set available yet ??
>
>> except a good editor and no compiler. And there was one floppy pair
>> which was not to be included with exports -- it had the crypto software
>> and the programs which used it (like an encryption capable vi editor) --
>> so it could be left out of export systems.
> OK, I will not use the system for real work anymore. It should only
> become a very nice piece of my collection of running systems.
O.K. And for that, you would not need the encryption set,
which I think that people coudl still get into big trouble for sending
out of the USA.
[ ... ]
>> If you find a place to download the floppies, beware that the
>> Unix-PC used two non-standard floppy formats. The typical MS-DOS
> I??m full of hope you are such a place for me, isn??t ?
BTW what is your system using in place of an apostrophe "'"? Whatever
it is comes out as two '?'s together in my e-mail client, and in my
editor of choice. I see it twice in your line above.
>> machine of the period put nine sectors on a track. The boot floppy and
>> some others had eight sectors per track, and the rest had ten sectors
> You are sure? I remember I was reading something about ten sectors???
Yes -- the system initially was distributed with only 8-sector
per track floppys, and then sometime near the 3.51 version or a bit
earlier, they found out that they could make it format the
10-sector-per-track floppys and work well with them. But the boot still
wants 8 sectors-per-track, so all diagnostic disks are in that format,
along with the first one or two install floppys.
>> per track -- making it difficult to build the floppies on anything other
>> than a 3B1. Some linux systems of the period had optional floppy
>> drivers which could generate and work with the odd sectored floppys, but
>> I never worked with that so this is only hearsay.
> I use a old DOS machine with some tools from the good, old CP/ M era.
> These tools can read and write many old disk formats. Hope it works
> for my UNIX-PC too.
O.K. Is there an equivalent to DD on it too? I'm not sure
whether the images were DDs of the whole disk format, or of the CPIO
file which was normally how the install disks were implemented.
>> But given your museum focus, I suspect that you will want it as
>> close to factory supplied as possible, rather than as commonly modified
>> by the user community.
> Yes, you are right. I prefer systems in their original state as like
> as they were delivered.
[ ... ]
> PS: Sorry about my bad English. My brain (if there is any) may be to
> small to learn it much better. Abe rich kann etwas Deutsch, wenn das
> helfen kann?
Your English is certainly much better than the little Deutsch
which I learned back around 1963 and never had a chance to use
afterwards. Since I had already learned and used Spanish in visits to
Latin-American countries, I was accused of being the only one in the
class who was speaking German with a Spanish accent. :-)
It looks as though your IP address range is not yet blocked for
spam (I've been fairly gentle on German sites, along with a few other
countries). So -- hopefully you can reach me by e-mail.
But perhaps there is another article in the newsgroup with a
pointer to the floppys, so we'll see.
Good Luck,
Bill Gunshannon
"DoN. Nichols" <dnic...@d-and-d.com> writes:
> On 2008-03-26, Juergen Sievers <JSie...@nadhh.hanse.de.com> wrote:
>> On 26 Mrz., 03:50, "DoN. Nichols" <dnich...@d-and-d.com> wrote:
>>> On 2008-03-25, Juergen Sievers <JSiev...@nadhh.hanse.de.com> wrote:
>>
>> Hallo DoN.
>>
>> Thank you very much for your big replay.
>> To give you more information about the system I added some more high
>> resolution pictures to the following location.
>>
>> http://www.nadhh.hanse.de.com/eBay/AT&T.UNIX-PC/
>
> Again -- an unfortunate directory name this time. Please use
> "-and-" in place of the "&" symbol since unix shells interpret the '&'
> on a command line as the end of one command to be run in the background,
> and the (possible) start of another, so when I cut and pasted that URL
> to a command line to start up my browser pointed to that web page. it
> tried to find one named:
>
> http://www.nadhh.hanse.de.com/eBay/AT
>
> and then tried to execute:
>
> T.UNIX-PC/
>
> as a command, which of course failed.
That's a URL:, not a Unix command and the ampersand is perfectly
valid (and fairly common) in a URL:. Of course, it also perfectly
valid anywhere in Unix as putting a "\" in front of it takes away
the special meaning.
bill
--
Bill Gunshannon | de-moc-ra-cy (di mok' ra see) n. Three wolves
bill...@cs.scranton.edu | and a sheep voting on what's for dinner.
University of Scranton |
Scranton, Pennsylvania | #include <std.disclaimer.h>
Juergen Sievers
> Don,
> That's a URL:, not a Unix command and the ampersand is perfectly
> valid (and fairly common) in a URL:. Of course, it also perfectly
> valid anywhere in Unix as putting a "\" in front of it takes away
> the special meaning.
>
> bill
Hi bill.
Never mind. You are right but it would be a little bit easier to use
the link in a "past & copy" environment. So I changed the name to
http://www.nadhh.hanse.de.com/eBay/ATandT-UNIX-PC/
and added some more and hopfully better pictures of the burned area
and the CPU's daughter board.
mfg
j.
DoN. Nichols
> In article <slrnfum4gp....@katana.d-and-d.com>,
> "DoN. Nichols" <dnic...@d-and-d.com> writes:
>> On 2008-03-26, Juergen Sievers <JSie...@nadhh.hanse.de.com> wrote:
>>> On 26 Mrz., 03:50, "DoN. Nichols" <dnich...@d-and-d.com> wrote:
>>>> On 2008-03-25, Juergen Sievers <JSiev...@nadhh.hanse.de.com> wrote:
>>>
>>> Hallo DoN.
>>>
>>> Thank you very much for your big replay.
>>> To give you more information about the system I added some more high
>>> resolution pictures to the following location.
>>>
>>> http://www.nadhh.hanse.de.com/eBay/AT&T.UNIX-PC/
>>
>> Again -- an unfortunate directory name this time. Please use
>> "-and-" in place of the "&" symbol since unix shells interpret the '&'
[ ... ]
>> as a command, which of course failed.
>
> Don,
> That's a URL:, not a Unix command and the ampersand is perfectly
> valid (and fairly common) in a URL:.
Perhaps common in URLs hosted on non-unix systems. But
considering that I use a newsreader (by choice) which does not honor
URLs, I have to cut and paste to get it to the browser. I have a shell
script named "op" which fires up opera after performing a cd to the
directory where I want most things to be saved. So I do a cut-and-paste
to generate the following command line:
op http://www.nadhh.hanse.de.com/eBay/AT&T.UNIX-PC/
and I am doing this late enough at night so I don't notice the '&' in
the line. And of course, the browser tries to open:
http://www.nadhh.hanse.de.com/eBay/AT
and then tries to treat:
T.UNIX-PC/
as a separate command.
> Of course, it also perfectly
> valid anywhere in Unix as putting a "\" in front of it takes away
> the special meaning.
Yes -- or I could have put the whole thing inside double quotes.
But it is still a nuisance to have to watch for that -- and to have to
rename saved images with embedded '&' in the name. (I tend to download
images so I can play with them using some convenient tool such as "xv"
(for most images) or "display" for the ".PNG" images which xv does not
handle.
And it is *very* bad practice to leave files with embedded '&'
or with newlines in the name sitting around the system, as it can cause
shell scripts (which may be doing cleanup of the system in the middle of
the night) to do undesirable things, depending on what is after the '&'
or the newline.
Since this was posted to a newsgroup which specializes in a
particular (and old) unix system, it is reasonable to consider that
those who are likely to be reading it and to be answering are likely to
still be running some flavor of unix by choice. I'm currently using
Solaris 10, FWIW.
So -- why not make things easier for such people, instead of
putting out roadblocks.
Note that an '&' in a URL, even if it is not in the filename but
rather in a directory name still can be inconvenient on a unix system
hosting the web page -- especially during the running of cleanup
scripts. A poor choice can cause major damage, even though it is not
intended.
Of course -- if you *really* want to see awkward characters in
filenames, hook a pre OS-X Macintosh to a Sun (older SunOs 4.1.? or
later Solaris) via NFS. Let the Mac user treat the imported files just
like his internal disk space. After a while, he will create some files
with '/' in the name. Somehow, NFS manages to do this, even though unix
considers the '/' to be one of the two totally impossible characters in
filenames. The other is the NULL character.
This happened on a system which I administered, and we needed to
move the user's files to a different server. Everything failed until I
unmounted the filesystem, and then used the "diredit" function in emacs
to change the characters from '/' to '~' just so things could be copied
and then deleted from the original disk partition. Had I known what I
know now, I would have changed that character to ':' so the Mac user
would have had similar problems. :-) Of course, now the Mac uses a unix
living under the GUI, so it has to play by unix rules -- even if
sensitivity to embedded space, '&' and various other characters is
hidden behind the GUI unless the user bothers to run the "Terminal"
program to gain access to the command line.
Enjoy,
DoN. Nichols
> On 27 Mrz., 12:42, billg...@cs.uofs.edu (Bill Gunshannon) wrote:
>
>> Don,
>> That's a URL:, not a Unix command and the ampersand is perfectly
>> valid (and fairly common) in a URL:. Of course, it also perfectly
>> valid anywhere in Unix as putting a "\" in front of it takes away
>> the special meaning.
>>
>> bill
>
> Hi bill.
> Never mind. You are right but it would be a little bit easier to use
> the link in a "past & copy" environment. So I changed the name to
>
> http://www.nadhh.hanse.de.com/eBay/ATandT-UNIX-PC/
Thank you!
> and added some more and hopfully better pictures of the burned area
> and the CPU's daughter board.
Hmm ... interesting -- it looks as though this never had the
stenciled white box, so it was not supposed to have anything there.
Perhaps someone put a wrong component there and it burned up? Can you
tell whether the leads were clipped or the component blown up?
Now -- for the daughter board -- I never had one of those, and I
suspect that it was the modification which allowed a program to directly
access the display memory thus making an alternative windowing system
run noticeably faster than it did with the loadable driver which was the
alternative. A pity that I don't remember the name of the windowing
system from AT&T.
The expansion box has three cards that I see. The top one in
the photo was the "floppy tape" controller card. (It treated a QIC
format tape drive as though it were a floppy, requiring formatting the
tape and resulting in the tape rewinding and re-reading each sector
which it wrote almost as though it were doing a shoeshine on the heads.
It took a long time to write the tape, and IIRC it got 27 MB on a tape
normally specified at 60 MB. Did you get the tape drive with it?
Oh yes -- you mentioned the hard disk controller in a previous
article as being the WD1010. That reminds me that the one needed for
accessing larger drives (more than 1024 cylinders, and/or more than eight
heads) is the WD2010. Of course, both are no longer made, so you need
to find an old hard disk controller card for an IBM-PC which has that
controller. (That is where I got mine, back when used PC hard disk
controller cards were common at hamfests (electronic fleamarkets held by
radio amateurs (HAMS)).
The second card I don't recognize at all.
The third card is a Voice Power card, and could record audio
from the phone line to the hard disk, play it the other way (your
outgoing message), dial by playing the TouchTone pad tones to the phone
line, and accept numeric input by recognizing the TouchTone tones.
The floppy tape and the voice power cards needed special drivers
installed into the OS (automatic loading at boot time). I don't
remember whether a driver was needed for the expansion box, but I don't
think so.
Enjoy,
blst...@bellsouth.net
> Now -- for the daughter board -- I never had one of those, and I
>suspect that it was the modification which allowed a program to directly
>access the display memory thus making an alternative windowing system
>run noticeably faster than it did with the loadable driver which was the
>alternative. A pity that I don't remember the name of the windowing
>system from AT&T.
Wasn't that MGR?
BLS
DoN. Nichols
That was it, indeed. Thanks!
You should be able to find it on one of the sites still hosting
programs ported to the Unix-PC. (Search for UnixPC, 3B1, and 7300, as
they are likely to be under any of those names. ;-)
And now we have had more traffic in this newsgroup in the past
week than in the previous several months. :-) Glad to see the activity.
Enjoy,
do...@75.usenet.us.com
> This happened on a system which I administered, and we needed to
> move the user's files to a different server. Everything failed until I
ls -bli would have revealed the inode number of the awkward file.
find . -inum NNN -exec mv {} sensiblename \;
should have worked.
If NFS put it there, then NFS would have been the place to undo the harm.
For older systems, NFS is the only decent way to deal with CDROM. The 3B1
and older HPUX systems come to mind. The same CDROM unusable when mounted
on an HPUX system could be read if mounted elsewhere and exported via NFS.
And of course, a CDROM jammed into the 3B1 floppy slot doesn't mount
easily. ;-)
--
Clarence A Dold - Hidden Valley Lake, CA, USA GPS: 38.8,-122.5
DoN. Nichols
> DoN. Nichols <dnic...@d-and-d.com> wrote:
>> This happened on a system which I administered, and we needed to
>> move the user's files to a different server. Everything failed until I
>
> ls -bli would have revealed the inode number of the awkward file.
> find . -inum NNN -exec mv {} sensiblename \;
> should have worked.
It might have, but I don't think so. you would still be at the
mercy of the system's interpretation of the path. Remember that find(1)
will just discover the name as it was, and mv(1) does a link(2) followed
by an unlink(2) to accomplish the task -- and the unlink(2) failed in a
hard-coded path compiled into a program.
I didn't think to try that one. at the time. I was sufficiently
mad at the user for creating filenames like that on a unix file server.
(Of course, he didn't know that it should not be done, and the NFS
didn't bother telling him. :-) I think that the main blame should have
been on the NFS implementations.
The problem is that the unix implementation simply refuses to
pass a filename with '/' in it without attempting to interpret whatever
precedes it as part of a path (directories), and what follows it as
another part of a path (more directories, ending in a possible
filename).
> If NFS put it there, then NFS would have been the place to undo the harm.
Yes -- when called from a Mac -- which did not consider '/' to
be a subdirectory separator. But the libraries on Solaris, at least,
would not allow it. Let's say the full path was:
/home-6/username/subdirecoryname/file/name
with "file/name" being the actual file. The libraries on Solaris, and
any other unix which I have used, would have attempted to locate a
directory named "file" at the end of the rest of the leading path, and
then attempted to remove a file simply named "name" -- *if* there were
also a directory named "file" at that level. In all cases it would
fail, unless there were also a file named "name" in the directory named
"file".
For that matter -- remember the old MS-DOS versions -- pre
Windows -- would not accept a space in a filename. I found a file with
such a name, and could not get any command-line approach to get rid of
it. Then I looked at the program which created it, and discovered that
it was written in BASIC. So -- I wrote a small BASIC program to remove
the file and it worked. Apparently the BASIC hooked into the filesystem
at a lower level that the rest of the OS -- which since Microsoft's
first product was a BASIC interpreter for the Altair 8800 (and later for
the Altair 680b which I had), and somewhat later there was a "Disk
BASIC" which booted from the floppy and all files were either program
files ".BAS", or data file -- all created by BASIC talking to a very low
level of the disk controller's drivers -- somewhere in the lower levels
of the BIOS in the MS-DOS versions. So -- It is not surprising that the
BASIC could do things which the OS prohibited.
> For older systems, NFS is the only decent way to deal with CDROM. The 3B1
> and older HPUX systems come to mind. The same CDROM unusable when mounted
> on an HPUX system could be read if mounted elsewhere and exported via NFS.
Interesting. So it would be mounted on something like SunOS
4.1.? I guess?
> And of course, a CDROM jammed into the 3B1 floppy slot doesn't mount
> easily. ;-)
And -- the 3B1 never supported NFS. I *think* that it supported
RFS, but I'm not sure. But -- the CD-ROM of the images would be
something to use for rebuilding damaged floppys. And it would be
possible to set up uucp on both some other system and the 3B1 and put
the ".cpio" files there, and let the install scripts work from there.
IIRC -- that was how downloaded files from "The Store" worked.
But, of course, you needed a system sufficiently built so you could
configure uucp before you could do that, which left out the possibility
of installing the "Foundation Set" that way at a minimum.
And there were certain nasties in the system backup scripts.
They checked the date/time on a file (/etc/installdate, I think), and
refused to back up anything older than that. So -- if you got programs
in tar form and installed them (e.g. an improved version of mkdir(1)
which included the "-p" option for building any missing subdirectories
in the path), it would not be saved in the backups. The principle was
to make sure that anything from the Foundation set was never overwritten
during a restore, but it should have done a compare of the output of
sum(1) to be sure that it was an unchanged version. :-)
do...@75.usenet.us.com
> On 2008-03-28, do...@75.usenet.us.com <do...@75.usenet.us.com> wrote:
> > find . -inum NNN -exec mv {} sensiblename \;
> It might have, but I don't think so. you would still be at the
fsdb is the tool. It should have been on the 3B1. I don't remember how to
use it now, but we used to do an undelete of a file in class. You had to
restore the two-byte inode number at the beginning of a file entry.
You could have changed the slash to a dash.
> And -- the 3B1 never supported NFS. I *think* that it supported
> RFS, but I'm not sure. But -- the CD-ROM of the images would be
Darn. I forgot about RFS. It seemed so much better to me, but then Sun
won the populist battle. I used to do something odd, like NFS-mount a
directory and RFS-export it so that nearby users could mount a directory
they otherwise couldn't get. One way worked, the other didn't.
I can't remember which systems had which. I may have been doing the
import-export on an RS6000.
I just found my two boxes of diskettes for the base install of the S50 when
closing up an office. I still can't get myself to throw them away. I think
I archived them as dd images to a CDROM somewhere, but I don't know what I
did with that. They were in a box with a 5-1/4" drive that I used to make
copies in a PC for someone once. It's been 23 years since I had a PC7300.
The diskettes are probably brittle. It's been 14 years since I powered up
the MiniFrame that I still have.
DoN. Nichols
> DoN. Nichols <dnic...@d-and-d.com> wrote:
>> On 2008-03-28, do...@75.usenet.us.com <do...@75.usenet.us.com> wrote:
>> > find . -inum NNN -exec mv {} sensiblename \;
>
>> It might have, but I don't think so. you would still be at the
>
> fsdb is the tool.
Hmm ... yes that is in Solaris 2.6, Solaris 10 and OpenBSD 4.2.
But I would have had to learn to use it -- and it, like diredit on emacs
would require an unmounted filesystem. Diredit was pretty easy to learn
to use, at least.
> It should have been on the 3B1.
Possibly -- but that wasn't the system where I encountered that
problem. We weren't using 3B1s as file servers to Macs, PCs and other
unix boxen (Suns, SGIs, and miscellaneous other machines).
> I don't remember how to
> use it now, but we used to do an undelete of a file in class. You had to
> restore the two-byte inode number at the beginning of a file entry.
> You could have changed the slash to a dash.
O.K.
>> And -- the 3B1 never supported NFS. I *think* that it supported
>> RFS, but I'm not sure. But -- the CD-ROM of the images would be
>
> Darn. I forgot about RFS. It seemed so much better to me, but then Sun
> won the populist battle.
I never used RFS. I learned to use NFS and that was sufficient.
(even though my first Sun -- a 2/120 -- also would export via RFS.
> I used to do something odd, like NFS-mount a
> directory and RFS-export it so that nearby users could mount a directory
> they otherwise couldn't get. One way worked, the other didn't.
:-)
> I can't remember which systems had which. I may have been doing the
> import-export on an RS6000.
O.K. Did you have any Suns available? I never got a chance to
play with an AIX system, though a neighbor has a couple running. (Yes,
two unix users on a single residential block. :-)
> I just found my two boxes of diskettes for the base install of the S50 when
> closing up an office. I still can't get myself to throw them away. I think
> I archived them as dd images to a CDROM somewhere, but I don't know what I
> did with that. They were in a box with a 5-1/4" drive that I used to make
> copies in a PC for someone once. It's been 23 years since I had a PC7300.
> The diskettes are probably brittle.
Well ... I had to use the 3.51 distribution floppys to
re-install my modified 3B1 somewhat less than a year ago. The floppys
were still readable, but their labels were getting difficult to read,
thanks to the gum soaking through the labels. I guess that pretty soon
they'll start falling off. It was this which motivated me to make the
images on a more recent system. (Now the trick would be getting those
images onto 5.25" floppys. :-) I wonder if anyone ever made a SCSI 5.25"
floppy drive? And something which would let me control the number of
sectors per track. :-)
> It's been 14 years since I powered up
> the MiniFrame that I still have.
Ah yes -- I never got one of those, so I moved to:
Tektronix 6130 (BSD 4.2 variant)
Sun 2/120
Sun 3/140
Sun 3/60
Sun 3/160
Sun 3/180
Various Sun4 systems from SS1+ up through my current ops which are Sun
Blade 1000 and Sun Fire 280R (pretty much the same system, except that
the 280R is rackmount and made as a server, so it has dual hot
swappable power supplies, and dual hot swappable Fibre Channel drives,
all from the front panel. To change anything in the SB-1000 you have to
power it down because there is an interlock switch on the side cover.
Oh yes -- and now a token Mac Mini to run the income tax
software which is not supplied for any of my other unix boxen. :-)
Also to allow upgrading the GPS receiver, which can be done only from
Windows or Macs -- and I won't let a Windows machine touch the net from
my IP block. :-)
do...@75.usenet.us.com
> O.K. Did you have any Suns available? I never got a chance to
> play with an AIX system, though a neighbor has a couple running. (Yes,
> two unix users on a single residential block. :-)
There was only one Sun that I can recall. I remember being fascinated with
how slow it was. Spending so much time doing stupid cute graphics like
expanding windows, it was slower than the Convergent machine with a slower
CPU performing identical tasks.
> were still readable, but their labels were getting difficult to read,
> thanks to the gum soaking through the labels.
I have some of those.
> (Now the trick would be getting those
> images onto 5.25" floppys.
That I can do, or could last time I tried. I think that's why I have that
one floppy drive in the box with the diskettes.
> Oh yes -- and now a token Mac Mini to run the income tax
I have a box of some Mac laptops. I wand to say DuoBook, an old iMac, and
a PowerPC Mac of some sort, sitting in the garage, awaiting some love.
> I won't let a Windows machine touch the net from my IP block. :-)
I gave that up around the time that the one true Unix was sold to Novell.
DoN. Nichols
>> O.K. Did you have any Suns available? I never got a chance to
>> play with an AIX system, though a neighbor has a couple running. (Yes,
>> two unix users on a single residential block. :-)
>
> There was only one Sun that I can recall. I remember being fascinated with
> how slow it was. Spending so much time doing stupid cute graphics like
> expanding windows, it was slower than the Convergent machine with a slower
> CPU performing identical tasks.
Hmm ... was this the 2/120 by any chance? Same CPU as the 3B1,
and slower than the MiniFrame.
And perhaps it was running SunTools instead of X11? I never
used SunTools, but I think that it went in for more eye candy than X11
did at the time.
But about a year ago, I was really amazed at how slow the 3B1
was when I fired it up after using Ultra-2s (dual 400 MHz CPUs) and
Ultra-60s (dual 450 MHz CPUs) as my desktop machine. And now those seem
slow with the Sun Blade 1000 with dual 900 MHz CPUs -- and the
possibility of moving up to 1200 MHz CPUs -- all of these are
UltraSPARCs, which are 64-bit machines native -- with the ability to run
32-bit code still, so most things don't need to be re-compiled unless
you really need the faster 64-bit math on math-intensive programs.
>> were still readable, but their labels were getting difficult to read,
>> thanks to the gum soaking through the labels.
>
> I have some of those.
>
>> (Now the trick would be getting those
>> images onto 5.25" floppys.
>
> That I can do, or could last time I tried. I think that's why I have that
> one floppy drive in the box with the diskettes.
I have some 5.25" floppy drives around, but not as many machines
to run them (other than the 3B1s and the Tektronix 6130) as used to be
the case. Most of my machines will still talk to 3.5" floppies, but not
to 5.25" ones. this is why I was wondering whether there were any
SCSI-interfaced 5.25" floppy drives. :-) I do have a SCSI controller
which will talk to 8" floppys and the 10 MB 8" hard drives by Shugart.
That is part of my first unix box -- a Cosmos CMS/UNX running on an 8
MHz 68000 CPU -- if you *really* want slow. It was slower than it
should have been, too. Not because of GUI graphics (no such in a v7
unix), but rather because the people who ported it had the compiler
treat the 68000 as a PDP-11/LSI-11. If those did not have the
instruction, the compiler did not use it -- with the exception of the
link an unlink instructions to create/destroy stack frames on
entry/exit to/from a function. :-)
In particular, I discovered that when both machines were running
the same simple cracking program (trying to crack a single password --
the root password on the Tektronix 6130) by running crypt(3) on
everything from a combination of the /usr/dict/words file on all of my
different systems. It finally found it -- "gnome". But the 3B1 was
trying two for every one on the v7 COSMOS system -- and the CPU
difference was 10 MHz 68010 vs 8MHz 68000 -- and the differences in the
CPUS should not have made a difference for the kind of programs being
run.
Oh yes -- I also got my first experience in patching a running
kernel on that COSMOS system. It had the TZ offset compiled into the
kernel -- and no sources to re-compile the kernel, which was the
*official* way to change the timezone on a v7 unix. :-)
>> Oh yes -- and now a token Mac Mini to run the income tax
>
> I have a box of some Mac laptops. I wand to say DuoBook, an old iMac, and
> a PowerPC Mac of some sort, sitting in the garage, awaiting some love.
One is quite enough. The Mac Mini is a tinly little brick, and
shares a monitor with three other systems via a switch, and is using a
USB Sun keyboard instead of a genuine Mac keyboard -- with no problems.
>> I won't let a Windows machine touch the net from my IP block. :-)
>
> I gave that up around the time that the one true Unix was sold to Novell.
Which then got sold to SCO -- who started (unsuccessfully)
sueing everyone in sight. :-)
Right now, I'm mostly using most Solaris 10 (SysVr4 descendant)
and OpenBSD (BSD 4.4 descendant).
do...@75.usenet.us.com
> Hmm ... was this the 2/120 by any chance? Same CPU as the 3B1,
> and slower than the MiniFrame.
Now you're making my head hurt. Shouldn't the 3B1 have been about the same
as the MiniFrame, or am I remembering that wrong? They were both
68010-10MHz.
> But about a year ago, I was really amazed at how slow the 3B1
> was when I fired it up after using Ultra-2s (dual 400 MHz CPUs) and
You are ruining one of my vague thoughts. I have the MiniFrame, which was
notably faster than its predecessor. I left it powered down for several
years, and I've never compared its performance to other machines, but I've
often thought of turning it on just to see if my memory could be correct.
You are making me think that it wouldn't be a magical visit to a previous
time, just boringly slow.
> But the 3B1 was trying two for every one on the v7 COSMOS system -- and
> the CPU difference was 10 MHz 68010 vs 8MHz 68000 -- and the differences
The 68010 had some improved hardware related to paging. The Convergent
memory design took excellent advantage of that, far superior to the
competition of the day, including Sun and NCR. That shouldn't have
affected a small memory CPU-intensive program, but the Convergent designs
were always faster.
DoN. Nichols
> DoN. Nichols <dnic...@d-and-d.com> wrote:
>> Hmm ... was this the 2/120 by any chance? Same CPU as the 3B1,
>> and slower than the MiniFrame.
>
> Now you're making my head hurt. Shouldn't the 3B1 have been about the same
> as the MiniFrame, or am I remembering that wrong? They were both
> 68010-10MHz.
I thought that the MiniFrame used the 68020, not the 68010. But
since I never had (nor even saw) one, I could easily be wrong.
>> But about a year ago, I was really amazed at how slow the 3B1
>> was when I fired it up after using Ultra-2s (dual 400 MHz CPUs) and
>
> You are ruining one of my vague thoughts. I have the MiniFrame, which was
> notably faster than its predecessor. I left it powered down for several
> years, and I've never compared its performance to other machines, but I've
> often thought of turning it on just to see if my memory could be correct.
> You are making me think that it wouldn't be a magical visit to a previous
> time, just boringly slow.
Hmm ... as it turns out, I've kept a copy of the dhrystone
bookmark and updated it as I ran newer and newer systems. So -- going
to a copy of it and extracting some entries from the comments where the
results were kept:
======================================================================
* MACHINE MICROPROCESSOR OPERATING COMPILER DHRYSTONES/SEC.
* TYPE SYSTEM NO REG REGS
* -------------------------- ------------ ----------- ---------------
* Cosmos 68000-8Mhz UniSoft v7 cc 305 322 Report with this distribution
* Cosmos 68000-8Mhz UniSoft cc 365 394 - My machine (Plessy ram cards)
* PDP-11/34 - UNIX V7M cc 387 438
* Onyx C8002 Z8000-4Mhz IS/1 1.1 (V7) cc 476 511
* ATT PC6300 8086-8Mhz MSDOS 2.11 b16cc 2.0 632 684
* ATT PC6300 8086-8Mhz MSDOS 2.11 CI-C86 2.20M 769 769
* ATT 3B2/300 WE32000-?Mhz UNIX 5.0.2 cc 735 806
* ATT PC7300 68010-10Mhz UNIX 5.2 cc 1041 1111
* ATT PC7300 68010-10Mhz UNIX 5.2(3.51) cc 1066 1137
* Sun2/120 68010-10Mhz Sun 4.2BSD cc 1136 1219
* PDP 11/70 - UNIX 5.2 cc 1162 1250
* Sun2/120 68010-10Mhz Standalone cc 1219 1315
* VAX 11/780 - UNIX 5.2 cc 1515 1562
* Sun 3/180 68020-16.67Mhz Sun 4.2 cc 3333 3846
* SUN-3/160C 68020-16.67Mhz Sun3.0ALPHA1 Un*x 3333 4166
* Sun 3/280 4.1.1_U1 cc 7731 7751
* Sun 3/280 4.1.1_U1 gcc 2.7.2 8695 8746
* Amdahl 470 V/8 ? ? - 15015
* Amdahl 580 - UTS 5.0 Rel 1.2 cc Ver. 1.5 23076 23076
* Amdahl 5860 ? ? - 28355
* Opus (SS1+ Clone) 4.1.3 cc 25641 25641
* Sun 4/40 (IPC) 4.1.3_u1 cc 22058 21739
* Sun 4/330 4.1.3_U1 cc 25000 25210
* Sun 4/40 (IPC) 4.1.3_u1 gcc 2.7.2 30612 30927
* Solbourne S4000DX 4.1B (4.1.2) cc 31250 31914
* SS-LX Solaris 2.6 gcc 2.8.1 32608 32258
* Sun 4/330 4.1.3_U1 gcc 2.7.2 34090 32967
* Opus (SS1+ Clone) 4.1.3 gcc 2.7.2 36585 35714
* Solbourne S4000DX 4.1B (4.1.2) gcc 2.7.2 53571 53571
* SS-10 4 Ross 33MHz CPUS SunOs 4.1.4 gcc 2.7.2.3 54545 53571
* SS-10 Dual 50MHz SuperCache Solaris 2.6 gcc 2.8.1 88235 90909
* SS-10 Sing 50MHz SuperCache Solaris 2.7 gcc 2.8.1 93750 90909
* SS-20 dual 50MHz It,TMS390Z50 Solaris 8 gcc 2.95.3 93750 93750
* 110 MHz SPARC-5 Solaris 2.6 gcc 2.95.3 115384 111111
* SPARC Ultra-1 140MHz Solaris 8 gcc 2.95.3 214285 214285
* SS-5 170 MHz Solaris 2.7 gcc 2.8.1 250000 230769
* SPARC Ultra-2 dual 300 MHz Solaris 8(2.8) gcc 3.3.2 585937 585937
* SPARC Ultra-2 dual 400 MHz Solaris 9(2.9) gcc 3.3.2 976562 967117
* SunFire 280R 2x900MHz 64 Solaris 10-U3 gcc 4.2.0 1199040 1210653 (slow -- why?)
* SunFire 280R 2x900MHz 64 Solaris 10-U3 cc 5.9 2145922 2145922
* SunFire 280R 2x900MHz 32 Solaris 10-U3 gcc 4.2.0 2173913 2183406
* SunFire 280R 2x900MHz 32 Solaris 10-U3 cc 5.9 2314814 2304147
*
* SHUTTLE S51G 2.3 GHz Celeron OpenBSD 3.5 gcc 2.95.3 4142502 4022526
======================================================================
Over the years, I had to keep increasing the number of loops per run,
and eventually had to change ints to longs to long longs.
>> But the 3B1 was trying two for every one on the v7 COSMOS system -- and
>> the CPU difference was 10 MHz 68010 vs 8MHz 68000 -- and the differences
O.K. From the above list:
======================================================================
* Cosmos 68000-8Mhz UniSoft cc 365 394 - My machine (Plessy ram cards)
* ATT PC7300 68010-10Mhz UNIX 5.2(3.51) cc 1066 1137
* Sun2/120 68010-10Mhz Standalone cc 1219 1315
======================================================================
So -- the Sun 2/120 (also a 10 MHz 68010) was about 20% faster
than the 3B1, but the 3B1 was a factor o 2.88 times faster than the
Cosmos CMX-16/UNX with only a 20% faster CPU. the Cosmos had a really
poor compiler for the 68000.
> The 68010 had some improved hardware related to paging.
Right -- it made virtual memory *possible* -- which the 68000
did not. (It did not trap enough data in a memory exception to allow
re-running the failed instruction after mapping new RAM into the
necessary address space.
> The Convergent
> memory design took excellent advantage of that, far superior to the
> competition of the day, including Sun and NCR. That shouldn't have
> affected a small memory CPU-intensive program, but the Convergent designs
> were always faster.
Except in the case of the 2/120 running a small program like
dhrystone. The executable on a Sun Blade 1000 is 8840 bytes long. (Of
course, that is dynamic libraries, using the following shared libs:
======================================================================
libc.so.1 => /usr/lib/libc.so.1
libm.so.2 => /usr/lib/libm.so.2
/platform/SUNW,Sun-Blade-1000/lib/libc_psr.so.1
======================================================================
The source for dhrystone (dry.c) is much larger (now) thanks to all the
records of various systems in it. My copy is currently at 17497 bytes.
Note that the SunFire 280R should turn out the same speeds as the Sun
Blade 1000 with the same CPU (which is what I have at the moment).
The SHUTTTLE at the end of the speed list is a 2.3 GHz Intel Celeron
based system in a nice compact package which I keep down in the shop to
talk to the Emco-Maier Compact-5/CNC with OpenBSD installed. It was
apparently designed to be carried to gamer's conventions, and is well
over the speed of anything which I really need to do. I haven't
bothered trying dhrystone on the Mac Mini yet.
Note that I tossed a couple of Amdahl systems in there for
perspective. :-)
Dennis Lefebvre
We have two fully functional 3B1 machines, one is is daily use and the
other is the spare. Whenever there is a failure of the primary machine
I put the spare into service and work on the other. This weekend I am
installing replacement floppy drives. If you have not found another
soloution I can make a copy of the 3.51 Foundation Set diskettes and
the 3.51m update and send them to you through the post.
This will be sufficient to allow you to reinitialize the hard drive
and install operating system.
If you would like to do this please send an email with your postal
address.
dennis
On Mar 26, 1:31 pm, Juergen Sievers <JSiev...@nadhh.hanse.de.com>
wrote:
> But I can't boot because I have no boot and installation disk now.
...
> Oh yes, I would do that immediately if I would have a bootable floppy disk
...
> But I haven't any installation set available yet
...
> OK, I will not use the system for real work anymore. It should only become a very nice piece of my collection of running systems.
...
Juergen Sievers
> Juergen
>
> We have two fully functional 3B1 machines, one is is daily use and the
> other is the spare. Whenever there is a failure of the primary machine
> I put the spare into service and work on the other. This weekend I am
> installing replacement floppy drives. If you have not found another
> soloution I can make a copy of the 3.51 Foundation Set diskettes and
> the 3.51m update and send them to you through the post.
Hallo Dennis.
Hi DoN.
Hi Folks.
I have ready a set of disks now. Thank you very much again for our
kindly and helpfully support. I have tried them immediately but the
system continue couldn’t boot from these floppies as also from its
hard drive.
So I think the system has some other hardware problem too. Possible it
is the burned part or something depending on it. So I have to make
some measurements to figure out what’s happened. I have a HP 16500B
including the processor pack for 68000. So I’m hopefully I could find
the error soon. But today I’m moving and so all of my equipment is
packed.
One of my ideas is the wrong frequency. I use a transformer to bring
down the German power of 230V/50Hz to the power used at the USA.
Possible 110V/60Hz. But the transformer does not change the frequency.
If some timing would be derived by the power line frequency then that
may cause a timeout or could impacting the disk access electronics
eg???.
To have schematics would be great, but for such commercial system it
could be hard to get them. OK, will come back to this wonderful group
if any state changes.
Regards
Juergen
DoN. Nichols
> On 30 Mrz., 18:16, Dennis Lefebvre <uni...@gmail.com> wrote:
>> Juergen
>>
>> We have two fully functional 3B1 machines, one is is daily use and the
>> other is the spare. Whenever there is a failure of the primary machine
>> I put the spare into service and work on the other. This weekend I am
>> installing replacement floppy drives. If you have not found another
>> soloution I can make a copy of the 3.51 Foundation Set diskettes and
>> the 3.51m update and send them to you through the post.
>
> Hallo Dennis.
> Hi DoN.
> Hi Folks.
>
>
> I have ready a set of disks now. Thank you very much again for our
> kindly and helpfully support. I have tried them immediately but the
> hard drive.
Oops! One thing to consider is the possibility that dust has
been sucked into the floppy drive by the fans and is blocking the
optical sensors which detect the index hole and the write-protect tab.
Only the index hole is likely to keep you from being able to read the
floppy.
> So I think the system has some other hardware problem too. Possible it
> is the burned part or something depending on it. So I have to make
> some measurements to figure out what?s happened. I have a HP 16500B
> including the processor pack for 68000. So I?m hopefully I could find
> the error soon. But today I?m moving and so all of my equipment is
> packed.
O.K.
> One of my ideas is the wrong frequency. I use a transformer to bring
> down the German power of 230V/50Hz to the power used at the USA.
> Possible 110V/60Hz. But the transformer does not change the frequency.
> If some timing would be derived by the power line frequency then that
> may cause a timeout or could impacting the disk access electronics
> eg???.
Well ... the power comes through a switching mode power supply,
so it is not going to pass timing information through the power supply,
and there is no other path for that. Too low a frequency could result
in ripple current appearing on the output of the switching mode power
supply -- but that could be fixed by adding capacitance just after the
rectifier.
> To have schematics would be great, but for such commercial system it
> could be hard to get them. OK, will come back to this wonderful group
> if any state changes.
Interestingly enough -- there *was* a "Technical Reference
Manual" -- which included schematics for most of the system. There were
never schematics available for the power supply, however (or for the
monitor) -- both of which were third-party devices. But the supply
should only produce +12V, +5V, and -12V (IIRC) -- all of which can be
measured at the connector going from the power supply to the system
board. Note that power supplies for the 3B1 also have a separate cable
supplying +5V and +12V on the standard disk drive connector. This was
to reduce the current through the system board, since the 3B1 used a
full-height hard drive, while the 7300 used only the half-height, which
drew less current.
While I have a copy of the Technical Reference Manual, I also
have some systems which I still bring up from time to time, so I do not
want to let go of mine. Perhaps someone else here has one which they
are now willing to part with?
Good Luck,
Dennis Lefebvre
1. If you have not already, I would clean and deoxidize the contacts
as Don mentioned in a previous message. When I tested that set of
diskettes by initializing a hard drive it would not boot reliably
until after I cleaned all of the connections.
2. It is possible the diskettes were damaged in mailing from the USA
to Germany if they were exposed to magnetism, such as an unshielded
conveyor belt motor. Both diskette #1 and diskette #2 should be
bootable, you may wish to try both. I will be in London in May and I
could send you another set from there.
Looking at a 3b1 power supply I see it is marked "Input 90-130V /
245W / 58-63Hz" , but I don't know enough about electrical system to
say if the Hz mismatch is important. It seems that if the machine was
able to boot a few times from the hard drive the problem should be
elsewhere.
dennis