Michael B. Brutman <mbbrutman-cctalk at brutman.com> wrote ..
> Hi,
>
> I made some changes in my TCP/IP code to make it behave better when
> packets are lost and it has to retransmit. The old code used to wait a
> fixed amount of time before retrying, which was painfully slow on a
> local connection. The new code is adaptive and should work much better
> on a wide range of connections.
>
> If you'd like to help me test it I have my PCjr running the mTCP FTP
> server with some old computer pictures, some software, and some text
> files. (I'll add more during the day to keep it interesting.) You can
> connect by going to the following address:
>
> ftp://96.42.84.69:2021/
Ok, I did a SITE STATS after I disconnected and it showed the server was lightly loaded (only me), so I went ahead and restarted wget. Will let it run and note any more hangs.
later,
jbdigriz
Up for auction at a local site : an IBM 10SR harddisk.
Might be rare enough to mention here. What was it used in anyway ?
no affiliation, too expensive in my opinion, etc etc, but there might be people looking for exactly this.
Jos
>> DG used in-house part numbers, very much like IBM and HP did. But I
>> haven't come across an equivalence list, so if you don't have the
>> schematics, you're lost.
> I vaguely seem to recall I had a DG equivalence list in the 90's.
Aha! Found a binder labeled "Components Guide" which has an equivalence list for some discretes in the back, and enough plagiarism from the standard IC manuals in the chip descriptions to help find the SSI/MSI original chips.
Same thing on the web at http://bitsavers.trailing-edge.com/pdf/dg/015-000028-03_DG_Components_Guide…
I have an old G70 that's currently running HP-UX 11 (I had 11i v1 on it for a while - bit slow for that, and the patches would break it if you weren't careful). Anyway, I have a workstation for all my HP-UX 11 needs and was kicking around the idea of putting 10.20 or 9.04 on the beast to give it a bit more vintage flair, but all of the CDs I have and have seen have been 10.20 for the 700s. Anyone seen a 10.20 for 800 around?
I actually had an Aviion 66 and a ?33? available
at work. No one wanted it. It's probably gone
now. Not Motorolla based though. It was a box full of Pentium Pros, IIRC.
At 05:18 PM 2/17/2013, you wrote:
>I saw what I believe was a Motorola based Aviion
>tower in a junk pile in a scrap pile a few years
>ago. I remember asking the list if anyone wanted
>me to retrieved it. There were no replies.
71 . [Computing] Hit the philistines three times
over the head with the Elisp reference manual.
--petonic at hal.com (Michael A. Petonic)
a50mhzham at gmail.com ? N9QQB (amateur radio)
"HEY YOU" (loud shouting) ? Second Tops (Set Dancing) ? FIND ME ON FACEBOOK
43? 7' 17.2" N by 88? 6' 28.9" W ? Elevation 815' ? Grid Square EN53wc
LAN/Telecom Analyst ? Open-source Dude ? Musician
? Registered Linux User 385531
Hi Bruce,
thanks for taking your time to answer my mail! Sorry to leave out details,
i know how frustrating it is if people forget the important bits of
their problems ;)
I have (at least sighted) all documentation from bitsavers:
http://bitsavers.informatik.uni-stuttgart.de/pdf/dg/
and read a good amount of code and documentation of the simh project
(because i would like to build a media emulator for this machine
later....maybe)
My main problem is that most of the stuff from bitsavers seems to be for
newer machines
that have a builtin ROM? My machine does not have the program load
option (at least theres no switch for it)
and i was under the impression that all documentation i found somehow
just loaded a stored
program for the check procedures. Am i wrong with this?
So, in short i did not find any checkout procedure that looked adequate
for my machine so i just
hacked in some small asm programs that at least ran on the simulator.
Actually, the machine runs simple nop jumps at least
1: jmp 2
2: jmp 3
3: jmp 1
If i remember right when running this program it did not even hang
itself up (aka you could "stop" the machine). But i prepared the popular
"Hello World"
asm example to run without RDOS, and it obviously failed at least
because of the stuck ACs. I mean, how can you write
anything meaningful without using an AC? ;)
By "Reset" i mean pushing the reset switch up when pressing down for
"stop" does nothing. Again i may be subject to
interpret the meaning of STOP and RESET wrong. Clearly this machine is
confusing me ;)
We will examine the power supply for sure! It bugs me that we did not
think of that, since for now we only get
to the machine once per week. We may also have the option to take it
home though....
Also a friend of mine (which should be registering here soon, too - and
is more experienced than me when it comes to electronics)
said that the machine was acting strange when he first powered it up
after it sat for a week (aka cold). Maybe that points
to another aging/heat related/power supply problem.
Thanks for clarifying the internal structure of the machine. I was not
aware of that although i looked at alot bus pinouts,
transfer timings, signal descriptions etc. See above, confusing!
The third party board is unidentified for now. We tried to run without
it but saw the same error. We also already noticed that the
machine runs only with 2 CPU boards and nothing else, at least the PC
counts and the ACs still deposit and load.
I promise to take quality photos of anything you like once we get back
to the machine. I apologize for the bad photos my phone makes.
best regards
Alex
On 02/15/13 18:03, Bruce Ray wrote:
> G'day Alex -
>
> Your e.mail begs many question.
> What documentation do you have?
> Are you following the checkout procedures in the DG hardware manual?
> What "simple programs" can be run?
> What does "reset the machine" mean? Does it mean use the Reset
> switch, power off the machine, or...?
>
> The original Nova is an interesting whose big weakness was its power
> supply (mainly the higher-voltage memory Vinh and Vmem supply).
> Note that console accumulator functions exam and deposit may be
> functionally tested without having a memory board in the computer(!).
>
> As Al said, the Nova uses data paths of 4-bit "nibbles" for most
> internal transfers - even to and from main core memory(!). Your
> symptom appears a common single-bit failure mode.
>
> What is the 3rd party board?
>
>
> Bruce Ray
> Wild Hare Computer Systems, Inc.
> bkr at WildHareComputers.com
>
> www.WildHareComputers.com
>
>
> www.NovasAreForever.org
>
Pulled it out, reseated the cables, checked the connections, plugged it into a big enough power supply,
and gave it another go.
Spun up, head load or lock solenoid clicked, waited about 4 seconds (no sound of seeking), spin down.
Sounds like the problem might be somewhere in the head positioning or voice coil/drive.
> DG used in-house part numbers, very much like IBM and HP did. But I
> haven't come across an equivalence list, so if you don't have the
> schematics, you're lost.
I vaguely seem to recall I had a DG equivalence list in the 90's. Or maybe confusing it with something else and my memory is just a red herring. Maybe Carl Friend has it?
Poly Paks always had lots of house-numbered TTL, I'm guessing floor sweepings from DG :-)
When I had to read DG schematics, if the part didn't have a standard number on it, it was usually pretty obvious from pinout or other parameters, what the standard part was.
By the time the Eclipse came along, those were just chock full with bipolar PROM's.
Tim
