RP01 and RP02 Disk Drives
Tom94022
I'm working with the Computer History Museum, Mountain View CA, on a
project to identify significant disk drives - the RP01 and RP02 have
been so identified. Would anyone on one of these mail lists have any
knowledge of any such drives still in existence, operational or not?
Tom Gardner
Los Altos CA
bob....@gmail.com
The Computer History Museum site shows the
Memorex 660, aka DEC RP01/02, as the
"First OEM disk drive shipment to Digital":
http://www.computerhistory.org/corphist/view.php?s=stories&id=159
630-1 (RP01) Disk Drive (link)
Isn't that wrong or a mis-statement ?
I thought the Diablo/DEC RK01/02 came
before the Memorex 660.
It might be easier finding a 660, since it
was oem'd by several different manufacturers.
Guy Sotomayor
If I'm not mistaken, the Memorex 660 was the drive used in the RP06. I
think the Memorex 630(?) was what was used for the RP01/2.
--
TTFN - Guy
Tom94022
The RP02 is the 660-1; DEC was by far the largest customer, including
many of Memorex's end user versions, the 660-0, which were converted to
660-1's and sold to DEC to become RP02's.
Tom94022
The Memorex 660-1 became the RP02
The Memorex 667 became the RP06
ISS did the others.
John Santos
I seem to remember seeing Ampex on the RP04. (But I also remember seeing
ISS on *something*. Or was ISS and Ampex the same company? I distinctly
recall seeing the name "Ampex" on some disk drive and thinking, "Oh, those
are the guys who build the big studio audio and video tape decks."
Also, the RP05 was pack compatible with the RP04, but looked just like an
RP06, so I think they were also Memorex. (And I do remember seeing the
name Memorex on RP06's.) I don't think I ever saw a real RP01/02/03, just
clones.
No idea which was first, the RK series or RP series.
--
John Santos
Evans Griffiths & Hart, Inc.
781-861-0670 ext 539
jmfb...@aol.com
bob....@gmail.com wrote:
>Tom94022 wrote:
>> Hello:
>>
>> I'm working with the Computer History Museum, Mountain View CA, on a
>> project to identify significant disk drives - the RP01 and RP02 have
>> been so identified. Would anyone on one of these mail lists have any
>> knowledge of any such drives still in existence, operational or not?
>>
>> Tom Gardner
>> Los Altos CA
>
>The Computer History Museum site shows the
>Memorex 660, aka DEC RP01/02, as the
>"First OEM disk drive shipment to Digital":
>http://www.computerhistory.org/corphist/view.php?s=stories&id=159
>630-1 (RP01) Disk Drive (link)
>
>Isn't that wrong or a mis-statement ?
>I thought the Diablo/DEC RK01/02 came
>before the Memorex 660.
I don't remember seeing RK02s before RP01/2s.
<snip>
/BAH
Bob Koehler
> The Memorex 630-1 became the RP01
> The Memorex 660-1 became the RP02
> The Memorex 667 became the RP06
> ISS did the others.
RP07 wasn't a Memorex? I'd always assumed it was due to the RP
designation. (Maybe it just worked with the same controller?)
Sarr J. Blumson
:
: Isn't that wrong or a mis-statement ?
: I thought the Diablo/DEC RK01/02 came
: before the Memorex 660.
I can't help with the "which Memorex model was it" question, but we
(Cyphernetics/ADP) had RP-01s by the end of 1969. I don't think there
were PDP-11s by then, much less RK01s.
--
--------
Sarr Blumson sarr.b...@alum.dartmouth.org
http://www-personal.umich.edu/~sarr/
Al _Kossow
> Isn't that wrong or a mis-statement ?
> I thought the Diablo/DEC RK01/02 came
> before the Memorex 660.
RK01 was from IOMEC, used on the RK8 (not RK8E)
RK02/03 are 1100/2200tpi Diablo drives.
Drawings for the RK8 are from late '69 early '70
The printed documentation for the RP10 1st edition is
Nov 1969. The drawings in the manual don't have a date
I could ask the RP10 designer the exact dates the next
time I see him (he is a CHM volunteer).
RP03 is an ISS drive. I don't recall any Ampex drives
being sold to DEC directly, though many OEM systems used
them, including the AED 8000 subsystem (I worked for AED
in the early 80's).
--
Posted via a free Usenet account from http://www.teranews.com
bob....@gmail.com
Yes thats where I first saw the RK02/03 on PDP-8's, Diablo
drives and the was company later bought by Zerox IIRC.
Late 60's early 70's maybe.
As a field guy the RP03/04 ISS (Sunnyvale I think) drives
were better drives more reliable and easier to work on.
I used to work for Ampex and repaired their DEC
compatible drives (SMD) but I don't think DEC ever
oem'd them.
If memory serves me right the 660 (RP02) was the
only drive you could "tune" by ear. In the field the
quick way to do a radial/azimuth alignment check
(with no scope) was to do a repeated full track seek
and listen to the noisy mechanical detent drop as each
track was reached. You could then tweek the drive
or pots till you got the right "sound". Customer used to
love it. Non factory approved method though.
bob
Al _Kossow
> And there was also the DF32....
question was first removable pack.
William Pechter
>Tom94022 wrote:
>> Guy Sotomayor wrote:
>>
>>>bob....@gmail.com wrote:
>>>
>>>>Tom94022 wrote:
>>>>
>>>>>Hello:
>>>>>
>>>>>I'm working with the Computer History Museum, Mountain View CA, on a
>>>>>project to identify significant disk drives - the RP01 and RP02 have
>>>>>been so identified. Would anyone on one of these mail lists have any
>>>>>knowledge of any such drives still in existence, operational or not?
>>>>>
>>>>>Tom Gardner
>>>>>Los Altos CA
>>>>
>>>>The Computer History Museum site shows the
>>>>Memorex 660, aka DEC RP01/02, as the
>>>>"First OEM disk drive shipment to Digital":
>>>>http://www.computerhistory.org/corphist/view.php?s=stories&id=159
>>>>630-1 (RP01) Disk Drive (link)
>>>>
>>>>Isn't that wrong or a mis-statement ?
>>>>I thought the Diablo/DEC RK01/02 came
>>>>before the Memorex 660.
>>>>
>>>>It might be easier finding a 660, since it
>>>>was oem'd by several different manufacturers.
>>>
>>>If I'm not mistaken, the Memorex 660 was the drive used in the RP06. I
>>>think the Memorex 630(?) was what was used for the RP01/2.
>>>
The RP05 was originally called the RP04-II (a half density RP06 designed
because the RP04's were not the most reliable things DEC ever had in the
field... But they did beat the TU45's by a bit.
The RP04's liked to blow driver transistors and such so the DEC folks
were looking at the RP06 and another Sperry ISS drive and chose the RP06
and came up with the wider head reduced capacity RP04-II (named RP05
after the marketing folks realized RP04-II would be confusing).
Anyway, the RP06 class had one of the remaining RP04-II prototypes for
the field service training class demo drive...
>>>
>>>--
>>>
>>>TTFN - Guy
>>
>> The Memorex 630-1 became the RP01
>> The Memorex 660-1 became the RP02
Was the RPR02 an ISS drive? I thought so. I thought it was their first
sale to DEC (although it could've been the RP03...)
DEC went back to Memorex after the RP04... and used CDC's for the
RM02/3/5 drives. The RP07 was also an ISS/Sperry product.
>> The Memorex 667 became the RP06
>> ISS did the others.
>>
>
>I seem to remember seeing Ampex on the RP04. (But I also remember seeing
>ISS on *something*. Or was ISS and Ampex the same company? I distinctly
>recall seeing the name "Ampex" on some disk drive and thinking, "Oh, those
>are the guys who build the big studio audio and video tape decks."
>
>Also, the RP05 was pack compatible with the RP04, but looked just like an
>RP06, so I think they were also Memorex. (And I do remember seeing the
>name Memorex on RP06's.) I don't think I ever saw a real RP01/02/03, just
>clones.
>
>No idea which was first, the RK series or RP series.
>
>--
>John Santos
>Evans Griffiths & Hart, Inc.
>781-861-0670 ext 539
Been a long time since RM and RP06 classes.
Bill
--
--
"When I think back on all the crap I learned in Vax school
It's a wonder I fixed anything at all." (to the tune of Kodachrome)
pechter-at-ureach.com
bob....@gmail.com
The Digital RP05/06 Field Handbook
(company confidential) says:
"...RP05/06 consists of a 677-51 (RP05) or
677-01 (RP06) disk my Memorex with a
DCL by DEC...."
The RP04 IIRC was definitely an ISS drive.
Tom94022
So here is what we have:
RP01 Memorex 630-1; equivalent to IBM 2311 (7 MB); shipped late 1968 (I
know, I shipped it);
RP02 Memorex 660-2; equivalent to IBM 2314 (29 MB); shipped early 1969
(ditto);
RP03 ISS equivalent to double density 2314 (58 MB)
RP04 ISS equivalent to IBM 3330 (100 MB)
RP05 Memorex 677-51, equivalent to IBM 3300 (100 MB)
RP06 Memorex 677-01, equivalent to IBM 3330-11 (200 MB)
RP07 ISS equivalent to IBM3350?? or maybe just a 50% increase over RP06
(300 MB)
[MB capacities are in million bytes in the IBM full track record
format]
FWIW, the one salesperson who sold most these products to DEC was
Arnold Cooley, working variously for Memorex and ISS. Also there are
two operational RP06's at the PDP-10 project - probably the oldest
operating disk drives in the world.
But, has ANYONE SEEN AN RP01 or RP02 recently?
Tom Gardner
bob....@gmail.com
NOPE, not me...good summary though....
still unresolved, what was the
DEC;s first oem'd disk, Memorex, Diablo
or something else....
jmfb...@aol.com
"Tom94022" <t.ga...@computer.org> wrote:
<snip>
>So here is what we have:
>
>RP01 Memorex 630-1; equivalent to IBM 2311 (7 MB); shipped late 1968 (I
>know, I shipped it);
>RP02 Memorex 660-2; equivalent to IBM 2314 (29 MB); shipped early 1969
>(ditto);
>RP03 ISS equivalent to double density 2314 (58 MB)
>RP04 ISS equivalent to IBM 3330 (100 MB)
>RP05 Memorex 677-51, equivalent to IBM 3300 (100 MB)
>RP06 Memorex 677-01, equivalent to IBM 3330-11 (200 MB)
>RP07 ISS equivalent to IBM3350?? or maybe just a 50% increase over RP06
>(300 MB)
<snip>
You could probably get some dates for the others by looking
at TW's (Tony Wachs) sources in TOPS-10 monitor listings.
/BAH
William Pechter
The RP07 was a late 80's Sperry product in the 450+ mb range used to
fill the gap before the RA81 came out in numbers.
>FWIW, the one salesperson who sold most these products to DEC was
>Arnold Cooley, working variously for Memorex and ISS. Also there are
>two operational RP06's at the PDP-10 project - probably the oldest
>operating disk drives in the world.
>
>But, has ANYONE SEEN AN RP01 or RP02 recently?
>
>Tom Gardner
>
bob....@gmail.com
Not to belabor a point, but I seemed to
Remember some "first time products", in the
11 and 8 world, being shipped out of
CSS before becoming a production item.
I worked out of the DC area office and we
Installed a lot of CSS hardware at the various
Military and spy agencies. Later they would
Become production products.
I remember the DT03, DX11-B and several
Others as originally CSS products than later,
I think, they became Part of some product line.
Additionally the DEC compatible companies would
Beat DEC to the punch with new devices in the 11 and
8 world. I remember a company out of Texas that had
a hard drive for the PDP-8 but can't recall their name.
Maybe a CSS or plug compatible device can lay
Claim to the first disk drive, but not the first oem'd
drive ? Maybe not.
Rich Alderson
> The RP07 was a late 80's Sperry product in the 450+ mb range used to
> fill the gap before the RA81 came out in numbers.
The RP07 was an *early* 80s (or even late 70s) Sperry product in the 450-500MB
range used to increase the available disk space/square foot of computer room.
We had them at Chicago by 1981, and there was an RP07 login structure on each
LOTS system in 1984.
--
Rich Alderson | /"\ ASCII ribbon |
ne...@alderson.users.panix.com | \ / campaign against |
"You get what anybody gets. You get a lifetime." | x HTML mail and |
--Death, of the Endless | / \ postings |
Rich Alderson
> RP07 ISS equivalent to IBM3350?? or maybe just a 50% increase over RP06
> (300 MB)
> [MB capacities are in million bytes in the IBM full track record format]
In that format, the RP07 would come out to about 500MB. I don't think it was
a 3350 equivalent--was the 3350 track size 19069 bytes? (3330-II was 13030,
right?)
> FWIW, the one salesperson who sold most these products to DEC was
> Arnold Cooley, working variously for Memorex and ISS. Also there are
> two operational RP06's at the PDP-10 project - probably the oldest
> operating disk drives in the world.
There are three operational RP06s at PDPplanet, of which two have been spinning
for the last 18 months or longer. How long have yours been up? What are they
attached to? How are they being used?
Mark Crispin
> The RP07 was an *early* 80s (or even late 70s) Sperry product in the 450-500MB
> range used to increase the available disk space/square foot of computer room.
> We had them at Chicago by 1981, and there was an RP07 login structure on each
> LOTS system in 1984.
I created the first 4-RP07 PS on the new Score in 1983, much to the
chagrin of Digital field service who complained that nobody could ever
need a 2GB filesystem and that it couldn't be backed up.
Actually, the real problem was that the early RP07 HDAs were very
unreliable and generally would fail after a month or two of use. Digital
knew this, but sold them anyway.
Restoring a 2GB filesystem from 9-track tape required much more time than
restoring a 512MB filesystem, and the fact that it was PS: meant that this
restore time was downtime rather than a mere mountable structure.
Fortunately, my boss backed me up, and make it clear that Digital that we
expected reliable hardware and not workarounds for unreliable hardware.
IIRC, they eventually ended up replacing all the HDAs with a second
generation that didn't fail.
One thing that people forget about the old DEC mainframe hardware is how
unreliable it was. I put in a great deal of effort in TOPS-20 to banish
software crashes (and largely succeeded with the exception of the TCP/IP
code), but hardware was a different issue. Each KL had its own
idiosyncracies. The old Score regularly crashed with microcode CRAM
parity errors before it was finally tracked down to a clock cable on the
backplane (said cable being cut and pasted to the door of the processor as
a trophy). That was nothing compared to the crashes suffered by the
Stanford business school machine in 1979 (and memorialized in a T-shirt
reading "I survived the crash of 1979"). KS systems were generally
better, although people came to dread the failure of the Mighty Mite power
supply.
It wasn't just customers. DEC's TOPS-20 developers were saddled with
systems (such as 2102) which were infamously unreliable.
Later production KLs were quite a bit more reliable, but by then it was
too late.
DEC hardware is something that I do NOT miss...
-- Mark --
http://staff.washington.edu/mrc
Science does not emerge from voting, party politics, or public debate.
Si vis pacem, para bellum.
R.A.Omond
> [...snip...]
> In that format, the RP07 would come out to about 500MB. I don't think it was
> a 3350 equivalent--was the 3350 track size 19069 bytes? (3330-II was 13030,
> right?)
Good grief :-) I thought I was the only pervert on the planet
to have these figures (13,030 and 19,069) permanently implanted
in my brain.
jmfb...@aol.com
Rich Alderson <ne...@alderson.users.panix.com> wrote:
>pec...@pechter.dyndns.org (William Pechter) writes:
>
>> The RP07 was a late 80's Sperry product in the 450+ mb range used to
>> fill the gap before the RA81 came out in numbers.
>
>The RP07 was an *early* 80s (or even late 70s)
TW's RP07 work was after 7.01, IIRC. Check the date on his
listings.
> Sperry product in the 450-500MB
>range used to increase the available disk space/square foot of computer room.
>We had them at Chicago by 1981, and there was an RP07 login structure on each
>LOTS system in 1984.
They were not mountable and couldn't be used as the boot device. In
our shop it was pretty useless except for scratch storage.
/BAH
pr...@k9.prep.synonet.com
> RP07 ISS equivalent to IBM3350?? or maybe just a 50% increase over RP06
> (300 MB)
510MB or 512MB.
> [MB capacities are in million bytes in the IBM full track record
> format]
> FWIW, the one salesperson who sold most these products to DEC was
> Arnold Cooley, working variously for Memorex and ISS. Also there
> are two operational RP06's at the PDP-10 project - probably the
> oldest operating disk drives in the world.
> But, has ANYONE SEEN AN RP01 or RP02 recently?
No, but I have 2 RP06s, a 667 and 4 RP07s...
pr...@k9.prep.synonet.com
> Actually, the real problem was that the early RP07 HDAs were very
> unreliable and generally would fail after a month or two of use.
> Digital knew this, but sold them anyway.
They also had a glass fibre based absolute filter, and with age it
would get fragile. Any extra vibration was likley to send a cloud and
happy to meet you glass fibres into the HDA... ISS did an ECO with a
Goretex filter for that one.
jmfb...@aol.com
That usually happened on the -10 side, too. But it was
software and not hardware that were "first time" products.
It had to be software and usually not at the monitor level.
CSS had been disbanded on the PDP-10 side and all those projects
were handled by the regular guys.
<snip>
/BAH
William Pechter
<pr...@k9.prep.synonet.com> wrote:
>Mark Crispin <M...@CAC.Washington.EDU> writes:
>
>> Actually, the real problem was that the early RP07 HDAs were very
>> unreliable and generally would fail after a month or two of use.
>> Digital knew this, but sold them anyway.
>
I'm not so sure we knew it in advance. The RP07 seemed to be a lot
worse engineered than the standards we had.
The head retract on power loss was abysmal and battery packs were added
and IIRC they were adjusted with levels to keep a slight tilt back so
the heads wouldn't slip back out onto the "semi-sealed" disk HDA.
The RA81's looked a hell of a lot better until the breather filter glue
disaster made them look like a pile of dreck.
>They also had a glass fibre based absolute filter, and with age it
>would get fragile. Any extra vibration was likley to send a cloud and
>happy to meet you glass fibres into the HDA... ISS did an ECO with a
>Goretex filter for that one.
Did that ECO ever get to DEC?
I only worked on one RP07 in my field service days... the one at E.R.
Squibb in East Brunswick, NJ had HDA failure and it took Regional or
District Support to come down to approve the HDA replacement.
I remember scoping the drive and sleeping on my down coat behind the
11/780 while waiting for the support guys to show up to authorize the
swap. The branch support guy was already out there with me...
They were trying really hard to get the reliability up and hold the
DEC Sperry engineers to the fire to get these fixed and reliable.
I think they flew an HDA in to Newark for this one.
I realized this service call was around '84 or 85 -- so I guess my
date estimate for the RP07 was really off a few.
Bill
Mark Crispin
> >The RP07 was an *early* 80s (or even late 70s)
> TW's RP07 work was after 7.01, IIRC. Check the date on his
> listings.
The new Score's RP07s were in January 1983, so RP07s were available by
then. I doubt that they were available in the 1970s.
-- Mark --
http://panda.com/mrc
Democracy is two wolves and a sheep deciding what to eat for lunch.
Liberty is a well-armed sheep contesting the vote.
Mark Crispin
> >> Actually, the real problem was that the early RP07 HDAs were very
> >> unreliable and generally would fail after a month or two of use.
> >> Digital knew this, but sold them anyway.
> I'm not so sure we knew it in advance.
My understanding was that the problems with the RP07 were well known by
the engineers in Marlboro. Perhaps the problem was that management and
marketing wasn't listening -- wouldn't have been the first time.
> I only worked on one RP07 in my field service days... the one at E.R.
> Squibb in East Brunswick, NJ had HDA failure and it took Regional or
> District Support to come down to approve the HDA replacement.
That was a sore spot with customers too. We have a situation where an
RP07 was clearly non-functional, it was on service contract, it hadn't
been abused, and yet we still had to wait for the paper-pushers to approve
the HDA replacement.
This was, mind you, the same year that Digital cancelled 36-bits (although
that didn't happen until May). So customers were not particularly
inclined to be patient when 512MB of their storage was down (or, in the
case of Score, 2GB).
jmfb...@aol.com
Mark Crispin <m...@CAC.Washington.EDU> wrote:
>On Sat, 20 Jan 2007, jmfb...@aol.com wrote:
>> >The RP07 was an *early* 80s (or even late 70s)
>> TW's RP07 work was after 7.01, IIRC. Check the date on his
>> listings.
>
>The new Score's RP07s were in January 1983, so RP07s were available by
>then. I doubt that they were available in the 1970s.
I know they weren't. TW's RP07 project was after SMP. Look
in the listing for the date.
/BAH
Paul Repacholi
> Did that ECO ever get to DEC?
Not sure. The 4 I have are from a KL10E. They all have off-white
filters, and the Goretex ones are a charming baby blue. They
came from a unisucks site from memory.
Paul Repacholi
One jumper removed, $7000. They did also run the formatter for
you, and MAY have restored your data. For a 780, you needed to
add a second memory controller and interleave them to avoid
data overruns on the SBI. 10s and 11/70s had no need of such.
Mark Crispin
> >The new Score's RP07s were in January 1983, so RP07s were available by
> >then. I doubt that they were available in the 1970s.
> I know they weren't. TW's RP07 project was after SMP. Look
> in the listing for the date.
With all due respect, Barb; Score was a TOPS-20 system. So TW's RP07
project for TOPS-10 is somewhat orthogonal.
However, I am sure that you are right: RP07s were not available in the
1970s.
I'm a bit surprised that TW had to have a product to do RP07s in TOPS-20.
The RP07 was programmed more or less the same as the RP04 and RP06. Once
you had the RP04/6 code written, an RP07 was simply a device with
different geometry. The RP20 was a different matter altogether (did
TOPS-10 ever support the RP20, a.k.a. DEC's answer to the SA10?).
Now, implementing RH20 would definitely be a project! From "Tony in RH20
Land" we can see how difficult the RH20 was for someone who works on
software that expects the DF10 ways of doing things. Seemingly, the RH20
was not designed with TOPS-10 considerations in mind.
On the other hand, they did choose the TOPS-10 style 128 word record size.
This was an annoyance to TOPS-20 people who used 512 word pages; the disk
address was based upon a record number in the structure which limited the
maximum number of records in a structure. As distributed by DEC, the
TOPS-20 disk address space could only support a single RP07 in a single
structure. You can actually support as many as 16 RP07s (8GB) in a
structure if you go to whole page record sizes (gain of 2 bits) and adjust
the disk address bits (gain of two more bits).
The Panda monitor does something more modest; it adjusts the disk address
bits in a way that gains one bit (thus allowing a two-RP07 structure) but
can still mount DEC filesystems. The code to get the other three bits
would make it impossible to mount a DEC filesystem.
jmfb...@aol.com
Mark Crispin <M...@CAC.Washington.EDU> wrote:
>On Sun, 21 Jan 2007, jmfb...@aol.com wrote:
>> >The new Score's RP07s were in January 1983, so RP07s were available by
>> >then. I doubt that they were available in the 1970s.
>> I know they weren't. TW's RP07 project was after SMP. Look
>> in the listing for the date.
>
>With all due respect, Barb; Score was a TOPS-20 system. So TW's RP07
>project for TOPS-10 is somewhat orthogonal.
Now listen, you young whippersnapper. When a new piece of gear
showed up, TW was the first one to make it work. And THEN the
-20 people wrote the code a second time.
That's how things happened.
/BAH
Mark Crispin
> >> >The new Score's RP07s were in January 1983, so RP07s were available by
> >> >then. I doubt that they were available in the 1970s.
> >> I know they weren't. TW's RP07 project was after SMP. Look
> >> in the listing for the date.
> >With all due respect, Barb; Score was a TOPS-20 system. So TW's RP07
> >project for TOPS-10 is somewhat orthogonal.
> Now listen, you young whippersnapper. When a new piece of gear
> showed up, TW was the first one to make it work. And THEN the
> -20 people wrote the code a second time.
> That's how things happened.
Fortunately, TOPS-20 kept edit histories.
RP07 support was added to TOPS-20 on May 6, 1978.
TOPS-10 sources do not have edit histories prior to 1980, and what exists
is just a reference to an MCO number.
If you still contend that TW added RP07 support to TOPS-10 before it was
added to TOPS-20, then you must withdraw your claim that TW did it in the
1980s and show evidence that TW did the work prior to May 6, 1978.
Human memories, while valuable in other contexts, do not constitute proof.
Tom94022
number of servo (or data) bytes index to index. 13,030 is the length
of a full track record 1 with no key field in IBM format. The 3330-11
was the same. DEC used a sector architecture which was more efficient
than IBM at that sector size, something like 2 additional sectors per
track.
I think the 3340 magic number was 20,160. I don't recall any later
ones but all of them had a fixed number of servo bytes index to index
to which the data was phase locked, usually at a multiple of 8.
I'm glad to see there are other preverts around :-)
glen herrmannsfeldt
(snip)
>>Good grief :-) I thought I was the only pervert on the planet
>>to have these figures (13,030 and 19,069) permanently implanted
>>in my brain.
> Actually isn't the magic number for the 3330 13,440, that is, the
> number of servo (or data) bytes index to index. 13,030 is the length
> of a full track record 1 with no key field in IBM format. The 3330-11
> was the same. DEC used a sector architecture which was more efficient
> than IBM at that sector size, something like 2 additional sectors per
> track.
The number I remember for the 3330 is 13165, then you subtract 135 for
each block, such that a full track block is 13030. Half track would
be (13165/2)-135. Other drives, such as the 3214, were not described
that way, but such that the gap overhead didn't apply to the first
block. The 13440 might include R0, which didn't hold user data.
The next number that was important was 3120, a multiple of 80 that
was reasonably efficient on both 2314 and 3330. (If you didn't
know which drive you were going to write on.)
-- glen
jmfb...@aol.com
>> >> >The new Score's RP07s were in January 1983, so RP07s were available by
>> >> >then. I doubt that they were available in the 1970s.
>> >> I know they weren't. TW's RP07 project was after SMP. Look
>> >> in the listing for the date.
>> >With all due respect, Barb; Score was a TOPS-20 system. So TW's RP07
>> >project for TOPS-10 is somewhat orthogonal.
>> Now listen, you young whippersnapper. When a new piece of gear
>> showed up, TW was the first one to make it work. And THEN the
>> -20 people wrote the code a second time.
>> That's how things happened.
>
>Fortunately, TOPS-20 kept edit histories.
So did TOPS-10. We didn't put them in the sources.
>
>RP07 support was added to TOPS-20 on May 6, 1978.
>
>TOPS-10 sources do not have edit histories prior to 1980, and what exists
>is just a reference to an MCO number.
Then you can find the MCO in the x.MCO that was shipped on the tape.
Now, if the RP07 was done before 1980, then there must have been
an RP07.LIR. I don't remember that being done.
>
>If you still contend that TW added RP07 support to TOPS-10 before it was
>added to TOPS-20, then you must withdraw your claim that TW did it in the
>1980s and show evidence that TW did the work prior to May 6, 1978.
>
>Human memories, while valuable in other contexts, do not constitute proof.
Yea, and mine is screwed up. I don't remember an RP07 hung on
the -20 machine. So, given your evidence and my knowledge of
how stuff worked, the folowing may be what happened.
The -20 got the RP07 and did their development and testing. Then
the RP07 was moved to the TOPS-10 machine and TW did his
work. I do know that the RP07 stayed on Kl1026.
If, as you say, it was just a rearrangement of the RP06 stuff, having
the -20 do it first would then make sense.
The usual procedures were for TW to do the first implementation of
a piece of new gear. Then the -20's implementation didn't have
to deal with the hardware types, the necessary ECO work was already
done and the -20 developer could use TW's experience for all the
stuff to avoid.
I simply don't remember the SMP work being interrupted by
the RP07.
/BAH
jmfb...@aol.com
>> >> >The new Score's RP07s were in January 1983, so RP07s were available by
>> >> >then. I doubt that they were available in the 1970s.
>> >> I know they weren't. TW's RP07 project was after SMP. Look
>> >> in the listing for the date.
>> >With all due respect, Barb; Score was a TOPS-20 system. So TW's RP07
>> >project for TOPS-10 is somewhat orthogonal.
>> Now listen, you young whippersnapper. When a new piece of gear
>> showed up, TW was the first one to make it work. And THEN the
>> -20 people wrote the code a second time.
>> That's how things happened.
>
>Fortunately, TOPS-20 kept edit histories.
>
>RP07 support was added to TOPS-20 on May 6, 1978.
Who did the -20 work?
<snip>
/BAH
Rich Alderson
Some context for the particular item MRC was quoting:
;<4.MONITOR>PHYP4.MAC.6, 5-Mar-79 15:53:41, EDIT BY KONEN
;UPDATE COPYRIGHT FOR RELEASE 4
;<4.MONITOR>PHYP4.MAC.5, 4-Dec-78 19:37:55, EDIT BY BOSACK
;TCO 4.2106 - DEFER ECC APPLICATION UNTIL DRIVE HAS SETTLED ON CYL
;<2MCLEAN>PHYP4.MAC.4, 30-Jul-78 15:16:13, Edit by MCLEAN
;<2MCLEAN>PHYP4.MAC.3, 30-Jul-78 14:46:26, Edit by MCLEAN
;<4.MONITOR>PHYP4.MAC.2, 22-Jul-78 22:11:23, Edit by MCLEAN
;<1MCLEAN>PHYP4.MAC.83, 6-May-78 21:45:28, Edit by MCLEAN
==> ;<1MCLEAN>PHYP4.MAC.82, 6-May-78 21:40:57, Edit by MCLEAN
==> ;ADD RP07 CODE
;<1MCLEAN>PHYP4.MAC.81, 6-May-78 21:40:37, Edit by MCLEAN
;<3.SM10-RELEASE-3>PHYP4.MAC.79, 3-Apr-78 18:49:52, Edit by MCLEAN
;FIX ATNXIT NOT TO CLEAR CHANNEL ON TRANSFERING DRIVE
;<3A.MONITOR>PHYP4.MAC.79, 8-Mar-78 23:58:51, Edit by MCLEAN
;CLEAR DRIVE AT ATNXIT
;<2BOSACK>PHYP4.MAC.77, 24-Feb-78 01:25:42, EDIT BY BOSACK
;CHANGE DEVICE/DATA ERROR FLAGGING
So the RP07 went into v.3A.
Mark Crispin
Mark Crispin
I don't have any of the MCO files.
> Yea, and mine is screwed up. I don't remember an RP07 hung on
> the -20 machine. So, given your evidence and my knowledge of
> how stuff worked, the folowing may be what happened.
I agree that the updated scenario that you outline is likely.
> If, as you say, it was just a rearrangement of the RP06 stuff, having
> the -20 do it first would then make sense.
Not quite a rearrangment; but the same code supports RP04/6/7.
jmfb...@aol.com
<alpine.WNT.0.82.0...@Tomobiki-Cho.CAC.Washington.EDU>,
Mark Crispin <M...@CAC.Washington.EDU> wrote:
>On Tue, 23 Jan 2007, jmfb...@aol.com wrote:
>> Then you can find the MCO in the x.MCO that was shipped on the tape.
>
>I don't have any of the MCO files.
>
>> Yea, and mine is screwed up. I don't remember an RP07 hung on
>> the -20 machine. So, given your evidence and my knowledge of
>> how stuff worked, the folowing may be what happened.
>
>I agree that the updated scenario that you outline is likely.
I also seem to recall that we would make some deal with a hardware
manufacturer and the suits would provide one, and only one,
piece of gear. You can't timeshare stand-alone time. The RP07
may have been in this category. It didn't happen often but was
really a RPITA w.r.t. scheduling major monitor releases. I don't
think the -20 had the infrastructure necessary for doing an LIR.
They had to do a full blown release. And this is never a trivial
matter.
>
>> If, as you say, it was just a rearrangement of the RP06 stuff, having
>> the -20 do it first would then make sense.
>
>Not quite a rearrangment; but the same code supports RP04/6/7.
You couldn't boot from that thing. The only functionality it
provided was that of a bit bucket. I can't recall if it
had dual ports. It made for a better boat anchor than a piece
of gear that included our new software functionality.
/BAH
jmfb...@aol.com
Tiger did that work?!!! It must have been done on the Hardware
Engineering system. That means it wasn't done by the monitor
group.
/BAH
jmfb...@aol.com
<alpine.WNT.0.82.0...@Tomobiki-Cho.CAC.Washington.EDU>,
Mark Crispin <M...@CAC.Washington.EDU> wrote:
>On Tue, 23 Jan 2007, jmfb...@aol.com wrote:
>> Who did the -20 work?
>
>Ron ("Bugs") McLean.
a.k.a. Tiger. That means the work wasn't done by the monitor group
and the gear had to have been hung off the Hardware Engineering
system....in 1978? Dick Helliwell "owned" that system.
/BAH
Al _Kossow
> a.k.a. Tiger. That means the work wasn't done by the monitor group
> and the gear had to have been hung off the Hardware Engineering
> system....in 1978? Dick Helliwell "owned" that system.
Barb, do you recall what the name of that HW system was, and the names
of the disc volumes? Ditto for the software group T10 machines (1026 and
SMP?). Appears 1026 may have been DSKA->DSKC or BLK? though this may
all be after you had left.
--
Posted via a free Usenet account from http://www.teranews.com
Mark Crispin
> >Not quite a rearrangment; but the same code supports RP04/6/7.
> You couldn't boot from that thing.
That's an overstatement.
The front end couldn't boot from an RP07; IIRC, because the RH11 was too
slow to handle an RP07. Hence the 2020 couldn't use an RP07 either.
However, once BOOT was loaded, there was nothing that prevented booting
the monitor from an RP07. Similarly, nothing prevented the use of an RP07
as the login/primary filesystem.
So, the only issue was that you either had to boot the front end from
floppy (s-l-o-w) or you had to have an RP06 to boot the front end.
Score had one RP06 (used for booting and holding sources) and a 4 RP07 PS:
for users.
> I can't recall if it
> had dual ports.
Yes, the RP07 could be dual ported. Quite a few of us did clusters that
way even before CFS existed.
Rich Alderson
Hmm. I'll have to get a phone number for Dick from Ralph Gorin and see what
he has to say about the RP07.
Antti Louko
> The front end couldn't boot from an RP07; IIRC, because the RH11 was
> too slow to handle an RP07. Hence the 2020 couldn't use an RP07
> either.
> However, once BOOT was loaded, there was nothing that prevented
> booting the monitor from an RP07. Similarly, nothing prevented the
> use of an RP07 as the login/primary filesystem.
> So, the only issue was that you either had to boot the front end from
> floppy (s-l-o-w) or you had to have an RP06 to boot the front end.
Almost like our experience with an RP07 and a VAX11/750 running
BSD. The RP07 came form the retiring DEC-2060 and the first problems
was the get it formatted for 512B blocksize. Local Digital support
didn't manage to do that. They had to replace the HDA. After that,
there was no boot support and we had to order a new boot ROM for
that. Before it finally arrived, we bootstrapped from TU78 tape with a
self-written boot program. And finally the new ROM (20(or so)-pin DIL)
arrived packed in a small plastic antistatic box (2"x3"x1") which was
inside a 4"x6"x2" cardboard box which was inside another 10"x15"x15"
box! After installing, the VAX booted natively from RP07.
jmfb...@aol.com
Al _Kossow <a...@spies.com> wrote:
>jmfb...@aol.com wrote:
>
>> a.k.a. Tiger. That means the work wasn't done by the monitor group
>> and the gear had to have been hung off the Hardware Engineering
>> system....in 1978? Dick Helliwell "owned" that system.
>
>Barb, do you recall what the name of that HW system was, and the names
>of the disc volumes?
It's in my head. Getting the bits retrieved is going to be the
hard part.
Their KLs were 1025, 1042. I'm going to have to wait for the
-20 cabinet numbers to pop in. I probably have this stuff
in my files somewhere, too.
> Ditto for the software group T10 machines (1026 and
>SMP?).
KAs #2 and #40
KIs #514 and #546
KLs - 1026 and 1042 (we inherited hardware's KL when were able to
get -20s off DEC's production line). For the third CPU it was
2276...no..2476
> Appears 1026 may have been DSKA->DSKC or BLK? though this may
>all be after you had left.
There were two phases. At first: DSKA was the system disk that
stayed with the machine; it was never moved. DSKB (2 packs) was the
system disk that was moved. DSKC: also moved but was one pack.
In the 1975-1979 (I'm guessing years) the machine room had three
systems: dual KA, dual KI and single (later dual) KL.
We had three "SYS"es. They were called the green packs, yellow
packs and red packs. The green packs were guaranteed to be
up all the time if any machine was running timesharing. If
two systems were up running timesharing, the second system
would have the yellow DSKB,C packs. If all three systems
(KA, KI, and KL) were up running timesharing then the red
packs would be up on the third system.
When the KA went away <sob>, the red packs went away.
Now, IIRC, DSKB was first in most people's search list.
In some cases, I can't recall why...probably something to
do with needing large quotas, some people would have
DSKC first in their login search list.
Now time has passed and the KL is the primary TOPS-10 system.
We have no space, both foot print and quota and spindles,
to keep all the black packs online. There are other apps
that require dedicated spindles. So we changed the
front end disks to be BLKX on #1026 and BLKY on #1042.
That way Release Engineering (and me) could start to develop
packaging procedures. In addition, the software library scheme
of managing all those CUSP sources could be available [(7x24)-
stand alone] could be implemented and then managed.
BLKK was the pack that had all the monitor software libraries.
In the beginning that pack only needed a dedicated spindle
when Magee was editing the monitor on Tuesdays. BLKK
used to be a set of DECtapes call The Black Tapes. RAP
(Alan Pommer) colored the white reels black with a magic
marker. Anybody who found a black tape knew that it
was in the wrong place and took it back to the guys
on ML3-5.
Now I've lost my train of thought. questions?
/BAH
Paul Repacholi
> Mark Crispin <m...@CAC.Washington.EDU> writes:
>> The front end couldn't boot from an RP07; IIRC, because the RH11
>> was too slow to handle an RP07. Hence the 2020 couldn't use an
>> RP07 either.
>> However, once BOOT was loaded, there was nothing that prevented
>> booting the monitor from an RP07. Similarly, nothing prevented the
>> use of an RP07 as the login/primary filesystem.
>> So, the only issue was that you either had to boot the front end
>> from floppy (s-l-o-w) or you had to have an RP06 to boot the front
>> end.
> Almost like our experience with an RP07 and a VAX11/750 running
> BSD. The RP07 came form the retiring DEC-2060 and the first problems
> was the get it formatted for 512B blocksize. Local Digital support
> didn't manage to do that. They had to replace the HDA.
Long time ago, but I think the on board formater gave you the pick of
512 or 576 bytes, and via an inch of super majik, interleaved or
`hi speed', aka the full 3M xfers/sec. Not for you 750 or 730!
780s needed a pair of interleaved memory controllers, KB-11s/RH-70s,
no problem!
> After that,
> there was no boot support and we had to order a new boot ROM for
> that. Before it finally arrived, we bootstrapped from TU78 tape with
> a self-written boot program. And finally the new ROM (20(or so)-pin
> DIL) arrived packed in a small plastic antistatic box (2"x3"x1")
> which was inside a 4"x6"x2" cardboard box which was inside another
> 10"x15"x15" box! After installing, the VAX booted natively from
> RP07.
From memory, the RP07 was more like a RM drive in it's programming,
not the rest of the late RPs.
I have a unisucks manual for the drive either here or over in `the
pile' at ACMS WA. So far I have found 2 empty binders, and 0 pages of
HW manual...
Mike Ross
wrote:
<vast and timeless snip>
>But, has ANYONE SEEN AN RP01 or RP02 recently?
FSSVO 'recently' - ten years ago - I had several rather rough old
RPR02s. They came with the RP15 controller which was attached to my
pdp-15 XVM system:
http://www.corestore.org/15-1.htm
Unfortunately they were all thrown away following a flood in the
property where they were stored.
I have recently (last year or two) be in touch with a guy in Canada
who has several interesting drives in storage, in not-too-bad shape
(I've seen pics): 1 RPR02, 2 RP03's, 2 RP04's and 4 RM03's. Must chase
that up; want these for my collection.
Cheers
Mike
--
http://www.corestore.org
'As I walk along these shores
I am the history within'
jmfb...@aol.com
>
>> a.k.a. Tiger. That means the work wasn't done by the monitor group
>> and the gear had to have been hung off the Hardware Engineering
>> system....in 1978? Dick Helliwell "owned" that system.
>
>Barb, do you recall what the name of that HW system was, and the names
>of the disc volumes? Ditto for the software group T10 machines (1026 and
>SMP?). Appears 1026 may have been DSKA->DSKC or BLK? though this may
>all be after you had left.
I can't remember. The only one that keeps popping up is System #1025
but logic says that isn't correct. I don't think 1025 worked well
enough to be used as hardware engineering's timesharing system.
And then I kept wondering if it wasn't 1042, the second CPU on
our system...It is possible that we got 1042 as our second
processor when hardware got a real -20 hardware machine, i.e.,
no external memory.
/BAH
Rich Alderson
> jmfb...@aol.com writes:
>> That means the work wasn't done by the monitor group
>> and the gear had to have been hung off the Hardware Engineering
>> system....in 1978? Dick Helliwell "owned" that system.
> Hmm. I'll have to get a phone number for Dick from Ralph Gorin and see what
> he has to say about the RP07.
So I did that (well, got RPH's e-mail address), and asked him. He didn't have
much to add, but suggested that I ask Ron McLean myself, and forwarded my query
on to Ron. With Ron's permission, I'm posting his response:
From: Ron McLean [ronmclean@....net]
Date: Wed 2/7/2007 10:25 AM
It's true that prior to about 1973 Wachs & Flemming did
almost all the new hardware development. Tops10 was the
primary operating system and they need the hardware working
there first for sale. After the release of KL10/20's this
was less the case since Tops20 had more funding I was
usually asked to do a lot of the new hardware implementation
(Disk/Tape/HSC). I generally got disks sooner than most
because they also had to run with the KL front end
software. The RP07's and later drives were therefore done
by me first. The changes were usually minimal since it was
only table entries for the disk and tape drivers and after
doing that I could test these drives. I also did the first
implementation of software for the KS10 since there was
little interest there for Tops10 on that platform.
I don't remember ever doing much on the hardware machines
for KL's so I presume I did the implementation on the Tops20
test machines. The diagnostic guys usually did there tests
with the hardware machines and then things were moved.
Ron (Tired of being retired ;-.])
Mark Crispin
my assessment based upon the historic records. In particular:
> The changes were usually minimal since it was
> only table entries for the disk and tape drivers and after
> doing that I could test these drives.
matches my knowledge of adding RP04-ish drives (RP04, RP05, RP06, RP07,
RM03, RM05) to PHYSIO.
For what it's worth, my KS has two RM05 drives. Sadly, one of these head
crashed about 9 years ago.
jmfb...@aol.com
Tom94022
> On 18 Jan 2007 21:10:50 -0800, "Tom94022" <t.gard...@computer.org>
> wrote:
>
> <vast and timeless snip>
>
> >But, has ANYONE SEEN AN RP01 or RP02 recently?
>
> FSSVO 'recently' - ten years ago - I had several rather rough old
> RPR02s. They came with the RP15 controller which was attached to my
> pdp-15 XVM system:
>
> http://www.corestore.org/15-1.htm
>
> Unfortunately they were all thrown away following a flood in the
> property where they were stored.
>
> I have recently (last year or two) be in touch with a guy in Canada
> who has several interesting drives in storage, in not-too-bad shape
> (I've seen pics): 1 RPR02, 2 RP03's, 2 RP04's and 4 RM03's. Must chase
> that up; want these for my collection.
>
> Cheers
>
> Mike
> 'As I walk along these shores
> I am the history within'
Hi Mike:
It turns out the RPR02 was a Memorex 660-0 (IBM plug compatible) that
had come off rent in the Memorex lease base and then was refurbished
into an RPR02 and resold by Digital. Memorex sold a bunch to Digital
in the late 1970's. I'd like to arm wrestle for the RPR02 if and when
you find the guy in Canada :-)
Stay in touch
Tom G
geezer
I think the guy in Canada you guys are talking about might be Dylan
Porter. A whois shows he still owns the pdp8.com site. I remember
years ago he had an HUGE collection. I did talk to him once but he had
some personal problems and disappeared for a while. Nice guy. I think
he was in Niagara Falls or Welland at the time.
Brian.
Al Kossow
> I think the guy in Canada you guys are talking about might be Dylan
> Porter. A whois shows he still owns the pdp8.com site.
"Dylan Porter" is John Bordynuik http://www.johnbordynuik.com/
who has my tu56, several of my scsi tape drives, and refuses to
answer my emails or phone calls
avoid him.
Mike Ross
>I think the guy in Canada you guys are talking about might be Dylan
>Porter. A whois shows he still owns the pdp8.com site. I remember
>years ago he had an HUGE collection. I did talk to him once but he had
>some personal problems and disappeared for a while. Nice guy. I think
>he was in Niagara Falls or Welland at the time.
I don't know 'Dylan Porter', unless that's an alias - you're talking
about a guy called John Bordynuik(sp?). Auctioned quite a few bits and
pieces on ebay a couple of years ago, I bought some.
He doesn't seem to participate in the 'scene' very much these days,
and I've heard varying opinions about the guy - but he's done some
pretty impressive stuff. Can't find the URL just now but there was a
website about a *very* thorough restoration of an Alto he did. He also
had a pdp-15 with a drum, which I think was the one that either came
from or went to Kevin Stumpf, I believe. Think a Mr. Garner has that
now...
And no, he's not the guy the RP disks :-)
Mike
--
Eric Smith
> but he's done some pretty impressive stuff.
Maybe, maybe not. He writes about doing some impressive stuff, but it's
not clear that much of it is real. Getting any actual hard information
out of him is impossible.
He's screwed over more than one person that frequents this newsgroup.
> Can't find the URL just now but there was a website about a *very*
> thorough restoration of an Alto he did.
He bought an Alto, but I don't think it's accurate to say that he did a
"very thorough restoration" as it was in good condition to start with.
Eric
glen herrmannsfeldt
(snip)
> In that format, the RP07 would come out to about 500MB. I don't think it was
> a 3350 equivalent--was the 3350 track size 19069 bytes? (3330-II was 13030,
> right?)
The 3330 is 13165 with 135 byte overhead. That leaves 13030 for a full
track block, though. Previous ones didn't seem to have the overhead for
the first block, or at least not in the published data. 7294 for the
2314. It might be that the difference is needed for Rotational Position
Sensing to get the block positions right. (Just a guess.)
-- glen