On 2011-07-15 11.00, Kari Uusim ki wrote:
On 15.7.2011 4:36, Johnny Billquist wrote:
[...]
But since this is now starting to spin off into rather much PDP-11
stuff, maybe we should get this to some other place as well?
Anyone have any good suggestions? Otherwise I'd suggest alt.sys.pdp11 on
internet news.
Johnny
.
There seems to be many interesting discussions popping up every now and
then. Would a DEC Notes solution be a feasible one? Very easy to use and
as it supports DECnet it will be available for anyone on HECnet.
...who is running VMS... ;-)
But in a way yes, Notes would work just fine as well.
I could also create a separate list where we could move discussions that are not of general HECnet nature, but anything that spins off from here, and to which people who really want to, could subscribe to.
Johnny
On 15.7.2011 4:36, Johnny Billquist wrote:
On 2011-07-15 03.24, Zane H. Healy wrote:
At 1:03 AM +0200 7/15/11, Johnny Billquist wrote:
RSTS/E and RSX-11M+ (as well as Unix) do require an MMU, and some
additional things to run though.
RSTS/E offers some real challenges if installing on real hardware. While
I've installed RT-11 and RSX-11M+ off of CD-ROM, I'm not sure it's
possible with RSTS/E. Another tricky situation I ran into was that
RSTS/E would install off 4mm DAT tapes, but the layered products,
especially DECnet/E wouldn't.
I seem to remember that some installation pieces in RSTS/E actually
requires that tape drives identify as the "correct" models, or RSTS/E
will refuse. I don't remember the details, but I think John Wilson
documented in the e11 manual.
Disks on real hardware will be small and slow (an possibly expensive).
Well, not neccesarily true. There are SCSI controllers, and large fast
disks. But they are often not free.
How large disks you can actually use will also depend on what OS you
run, with RT-11 being the most limited.
What does the availability of disks look like for someone looking to put
together a PDP-11 at this point? Unless you're like a lot of us here,
and have been doing this for a long time, and built up a supply of
spares years ago, such things can be hard to come by. Even Narrow SCSI
disks are getting harder to obtain.
I have not had any problems with rather modern SCSI disks. No need to
get narrow disks. They are supposed to be backwards compatible anyway.
I have some wide RZ disks in storageworks bricks on my 11/93 as well as
my 11/84.
Just don't hope for anything if you have differential SCSI.
I also have RA disks running, and RL, but nothing else. I don't like RD
disks because they are small, and *slow*, if you are even lucky enough
to get them running.
If you have *ANY* third party controller, you're likely to have a
slightly easier time. Any MFM disks will be hard to find, and unlike
when I was putting my first system together ESDI disks are likely to be
even harder to get. While a SCSI board might set you back a nice chunk
of change, it's likely to be the best path. Besides then you can hook up
a CD-ROM drive.
Yes.
On emulated systems they can be much bigger and many many times faster.
With much more ease, and less cost, yes.
Back in the 90's I swore running on real hardware was the way to go.
Now, with a nice collection of spare parts, I think Emulation would have
been better! Not as fun, but a Linux box running an emulator takes up a
lot less space! Thankfully that's the route I went with the PDP-10...
He. A PDP-10 takes a little space, even for a KS...
But since this is now starting to spin off into rather much PDP-11
stuff, maybe we should get this to some other place as well?
Anyone have any good suggestions? Otherwise I'd suggest alt.sys.pdp11 on
internet news.
Johnny
.
There seems to be many interesting discussions popping up every now and then. Would a DEC Notes solution be a feasible one? Very easy to use and as it supports DECnet it will be available for anyone on HECnet.
Kari
On 2011-07-15 03.24, Zane H. Healy wrote:
At 1:03 AM +0200 7/15/11, Johnny Billquist wrote:
RSTS/E and RSX-11M+ (as well as Unix) do require an MMU, and some
additional things to run though.
RSTS/E offers some real challenges if installing on real hardware. While
I've installed RT-11 and RSX-11M+ off of CD-ROM, I'm not sure it's
possible with RSTS/E. Another tricky situation I ran into was that
RSTS/E would install off 4mm DAT tapes, but the layered products,
especially DECnet/E wouldn't.
I seem to remember that some installation pieces in RSTS/E actually requires that tape drives identify as the "correct" models, or RSTS/E will refuse. I don't remember the details, but I think John Wilson documented in the e11 manual.
Disks on real hardware will be small and slow (an possibly expensive).
Well, not neccesarily true. There are SCSI controllers, and large fast
disks. But they are often not free.
How large disks you can actually use will also depend on what OS you
run, with RT-11 being the most limited.
What does the availability of disks look like for someone looking to put
together a PDP-11 at this point? Unless you're like a lot of us here,
and have been doing this for a long time, and built up a supply of
spares years ago, such things can be hard to come by. Even Narrow SCSI
disks are getting harder to obtain.
I have not had any problems with rather modern SCSI disks. No need to get narrow disks. They are supposed to be backwards compatible anyway.
I have some wide RZ disks in storageworks bricks on my 11/93 as well as my 11/84.
Just don't hope for anything if you have differential SCSI.
I also have RA disks running, and RL, but nothing else. I don't like RD disks because they are small, and *slow*, if you are even lucky enough to get them running.
If you have *ANY* third party controller, you're likely to have a
slightly easier time. Any MFM disks will be hard to find, and unlike
when I was putting my first system together ESDI disks are likely to be
even harder to get. While a SCSI board might set you back a nice chunk
of change, it's likely to be the best path. Besides then you can hook up
a CD-ROM drive.
Yes.
On emulated systems they can be much bigger and many many times faster.
With much more ease, and less cost, yes.
Back in the 90's I swore running on real hardware was the way to go.
Now, with a nice collection of spare parts, I think Emulation would have
been better! Not as fun, but a Linux box running an emulator takes up a
lot less space! Thankfully that's the route I went with the PDP-10...
He. A PDP-10 takes a little space, even for a KS...
But since this is now starting to spin off into rather much PDP-11 stuff, maybe we should get this to some other place as well?
Anyone have any good suggestions? Otherwise I'd suggest alt.sys.pdp11 on internet news.
Johnny
At 1:03 AM +0200 7/15/11, Johnny Billquist wrote:
RSTS/E and RSX-11M+ (as well as Unix) do require an MMU, and some additional things to run though.
RSTS/E offers some real challenges if installing on real hardware. While I've installed RT-11 and RSX-11M+ off of CD-ROM, I'm not sure it's possible with RSTS/E. Another tricky situation I ran into was that RSTS/E would install off 4mm DAT tapes, but the layered products, especially DECnet/E wouldn't.
Disks on real hardware will be small and slow (an possibly expensive).
Well, not neccesarily true. There are SCSI controllers, and large fast disks. But they are often not free.
How large disks you can actually use will also depend on what OS you run, with RT-11 being the most limited.
What does the availability of disks look like for someone looking to put together a PDP-11 at this point? Unless you're like a lot of us here, and have been doing this for a long time, and built up a supply of spares years ago, such things can be hard to come by. Even Narrow SCSI disks are getting harder to obtain.
If you have *ANY* third party controller, you're likely to have a slightly easier time. Any MFM disks will be hard to find, and unlike when I was putting my first system together ESDI disks are likely to be even harder to get. While a SCSI board might set you back a nice chunk of change, it's likely to be the best path. Besides then you can hook up a CD-ROM drive.
On emulated systems they can be much bigger and many many times faster.
With much more ease, and less cost, yes.
Back in the 90's I swore running on real hardware was the way to go. Now, with a nice collection of spare parts, I think Emulation would have been better! Not as fun, but a Linux box running an emulator takes up a lot less space! Thankfully that's the route I went with the PDP-10...
Zane
--
| Zane H. Healy | UNIX Systems Administrator |
| healyzh at aracnet.com | OpenVMS Enthusiast |
| | Photographer |
+----------------------------------+----------------------------+
| My flickr Photostream |
| http://www.flickr.com/photos/33848088 at N03/ |
On Jul 14, 2011, at 4:53 PM, Steve Davidson wrote:
Mark,
Rare - probably depends on where you can look! :-)
RT-11 will run on any of the PDP-11's. The RSX-11 family of OS's has
specific requirements about memory amounts and memory management thus
the SPD should be reviewed. RSTS/E (and M+) require memory management
hardware.
RSTS V4 (not /E) runs on a non-MMU machine, but it's very limited -- Basic-Plus only.
The Pro-Series of systems could also be looked at for RT-11 and RSX
(P/OS). In this space the Pro-380 is probably preferred. The problem
here is finding a network card (DECNA) that does not cost as much (or
more) than the rest of the system.
Indeed. And they are quite slow because of the pathetic architecture. Also, all you get is the console display, unless you can find one of the 4-line UART cards that is so obscure that it's hard even to find a manual for it, or a reference to its name (it's PC3XC-BA... I had to look for that).
I did a RSTS port, but that wasn't distributed.
Disks on real hardware will be small and slow (an possibly expensive).
On emulated systems they can be much bigger and many many times faster.
The books you have detail the degree of expandability for each of these
HW platforms (except maybe the Pro). You will be surprised at just how
much can be done with these systems. The speed will be another story.
Pro expansion is possible in theory, out of the question in practice. The needed details are undocumented in critical places.
RT-11 Languages (from memory):
MACRO-11
APL-11
BASIC-11
MU/BASIC-11
BASIC-PLUS/RT-11
FORTRAN-IV (FORTRAN 66 STD)
PDP-11/C
DECUS-C
Also (via DECUS): Forth and Algol
RSX-11 Languages (also from memory):
MACRO-11
APL-11
BASIC-PLUS-2
FORTRAN-IV (FORTRAN 66 STD)
FORTRAN-77
PDP-11/C
DECUS-C
COBOL-81, COBOL-11
And Coral-66, whatever that is.
RSTS/E Languages (also from memory):
The same as RSX-11 for the most part.
That, plus BASIC-PLUS (the ancestor of BASIC-PLUS-2, but an interactive interpreter). And Forth (a port of the RT one with FIG-Forth features added, included in the most recent releases, unsupported). There was a port of DECUS Algol but some of the pieces may have gotten lost; I had it at DEC but I can't find some of the pieces, at least not all the sources. Also Dibol.
And don't forget TECO, for any PDP-11 and any OS (except DOS and probably DSM).
paul
On 2011-07-14 22.53, Steve Davidson wrote:
Mark,
Rare - probably depends on where you can look! :-)
Indeed.
RT-11 will run on any of the PDP-11's. The RSX-11 family of OS's has
specific requirements about memory amounts and memory management thus
the SPD should be reviewed. RSTS/E (and M+) require memory management
hardware. If you can find one, the best is probably the 11/73, 11/83 or
11/93 in a BA23 (QBUS) enclosure. These come in either 4, or 8-slot
backplanes. If you really want to experience pain in your electric bill
then the BA123 (also QBUS) with it's 645 watt dual power supplies can
double as a winter heater/loud white-noise generator. These come in a
12 slot backplane configuration.
Actually, RSX-11M will run on just about anything. I think the minimum memory requirements is 32K, but that's about it. You don't need an MMU, or any other fancy stuff, but such a system is not that much fun to run, so I wouldn't recommend it to anyone who is not already very familiar with the system, and have specific requirements.
There is also the diskless RSX-11S, but that requires a remote boot node (either RSX or VMS).
RSTS/E and RSX-11M+ (as well as Unix) do require an MMU, and some additional things to run though.
Size and powerwise, the BA23 with one of the above mentioned CPUs are probably pretty optimal. The BA123 is nice in that you have room for more disks and a much more serviceable box.
The Pro-Series of systems could also be looked at for RT-11 and RSX
(P/OS). In this space the Pro-380 is probably preferred. The problem
here is finding a network card (DECNA) that does not cost as much (or
more) than the rest of the system.
A PRO is a nice system in a way, unfortunately, if running P/OS, you'll probably feel somewhat frustrated by the feeling that the system probably can do more than you can figure out how to make it do.
Also, the hardware is incompatible with any other model of PDP-11, which puts some restrictions on you. But you'll get graphics instead. :-)
Disks on real hardware will be small and slow (an possibly expensive).
Well, not neccesarily true. There are SCSI controllers, and large fast disks. But they are often not free.
How large disks you can actually use will also depend on what OS you run, with RT-11 being the most limited.
On emulated systems they can be much bigger and many many times faster.
With much more ease, and less cost, yes.
The books you have detail the degree of expandability for each of these
HW platforms (except maybe the Pro). You will be surprised at just how
much can be done with these systems. The speed will be another story.
All of these machines have some weight to them. An emulator will be
much faster and in the end easier to deal with. When the PLUTO::
machine is running as an emulated machine it is running with NetBSD and
SimH on a 700MHz Pentium-III, otherwise it is a real PDP-11/23+ (22-bit
backplane). The emulated machine is faster than the real thing, but not
my much. I was going for similar performance so that when I switch
between the two it wasn't that noticeable/painful to return to the
actual hardware.
Wow, since a 11/23 is a pretty slow machine...
RT-11 Languages (from memory):
MACRO-11
APL-11
BASIC-11
MU/BASIC-11
BASIC-PLUS/RT-11
FORTRAN-IV (FORTRAN 66 STD)
PDP-11/C
DECUS-C
There was probably some more languages, although I haven't got much clue about RT-11. I know one of the earliest Prolog implementations was for RT-11.
RSX-11 Languages (also from memory):
MACRO-11
APL-11
BASIC-PLUS-2
FORTRAN-IV (FORTRAN 66 STD)
FORTRAN-77
PDP-11/C
DECUS-C
COBOL-81, COBOL-11
Let's see. You forgot Pascal, BCPL, Forth, SIMULA, Lisp (which are other languages I have installed), and I also know of Focal and Dibol, and I feel that I must be missing a few languages in here...
RSTS/E Languages (also from memory):
The same as RSX-11 for the most part.
Almost a combination of stuff from RT-11 and RSX.
Johnny
On 15-7-2011 0:40, Johnny Billquist wrote:
I'd say no, they are not rare. But it's constantly getting more difficult to find machines as time goes by. No new ones are being built. You either need to find some place that is throwing out old hardware, and those are getting fewer, or else get something from another collector or a dealer.
They're not rare, but they're hard to find...?
- MG
On 2011-07-14 21.49, Mark Wickens wrote:
Are the MicroPDPs (desktop, deskside, rackmount) rare beasts? They seem
to be of a fairly hobbyist-friendly size, although presumably they are
less expandable (much like a VAXstation versus VAXserver I would imagine)
I'd say no, they are not rare. But it's constantly getting more difficult to find machines as time goes by. No new ones are being built. You either need to find some place that is throwing out old hardware, and those are getting fewer, or else get something from another collector or a dealer.
MicroPDP is a designation that don't really say anything about expandability, and is not really comparable to VAXstation vs. VAXserver.
MicroPDP was a designation used for Q-bus machines. Originally they were prepackaged solutions, on which a "micro" version of an OS was designed to run on, but that was only to make a cheaper option available. The full software versions also runs just fine, and you can add any additional hardware on the machine.
Traditionally, "large" PDP systems were the Unibus systems. They are much less "hobbyist-friendly" in size, yes. :-)
The fastest machines produced by DEC were the MicroPDP systems, though.
Johnny
The 11/83 is not as rare as the rest might be. I purchased one (1) on Ebay a year or so ago. It is running in a BA123 enclosure with 2MB of memory and (4) RD54's.
-Steve
From: owner-hecnet at Update.UU.SE [mailto:owner-hecnet at Update.UU.SE] On Behalf Of Joe Ferraro
Sent: Thursday, July 14, 2011 16:44
To: hecnet at update.uu.se
Subject: Re: [HECnet] PDP Ignorance
Although I hate to mention it... I have about 30 micro PDP 11/83's sitting in a lab right beside me; unfortunately they do not (yet... nor might they ever) belong to me... are they, in fact, "rare"?
On Thu, Jul 14, 2011 at 3:49 PM, Mark Wickens <mark at wickensonline.co.uk> wrote:
Are the MicroPDPs (desktop, deskside, rackmount) rare beasts? They seem to be of a fairly hobbyist-friendly size, although presumably they are less expandable (much like a VAXstation versus VAXserver I would imagine)
Mark.
Mark,
Rare - probably depends on where you can look! :-)
RT-11 will run on any of the PDP-11's. The RSX-11 family of OS's has
specific requirements about memory amounts and memory management thus
the SPD should be reviewed. RSTS/E (and M+) require memory management
hardware. If you can find one, the best is probably the 11/73, 11/83 or
11/93 in a BA23 (QBUS) enclosure. These come in either 4, or 8-slot
backplanes. If you really want to experience pain in your electric bill
then the BA123 (also QBUS) with it's 645 watt dual power supplies can
double as a winter heater/loud white-noise generator. These come in a
12 slot backplane configuration.
The Pro-Series of systems could also be looked at for RT-11 and RSX
(P/OS). In this space the Pro-380 is probably preferred. The problem
here is finding a network card (DECNA) that does not cost as much (or
more) than the rest of the system.
Disks on real hardware will be small and slow (an possibly expensive).
On emulated systems they can be much bigger and many many times faster.
The books you have detail the degree of expandability for each of these
HW platforms (except maybe the Pro). You will be surprised at just how
much can be done with these systems. The speed will be another story.
All of these machines have some weight to them. An emulator will be
much faster and in the end easier to deal with. When the PLUTO::
machine is running as an emulated machine it is running with NetBSD and
SimH on a 700MHz Pentium-III, otherwise it is a real PDP-11/23+ (22-bit
backplane). The emulated machine is faster than the real thing, but not
my much. I was going for similar performance so that when I switch
between the two it wasn't that noticeable/painful to return to the
actual hardware.
PLUTO:: configuration (from memory)
BA23 backplane (8-slot version)
PDP-11/23+
4MB memory
FPU chip set (somewhat rare these days)
CIS chip set (extremely rare these days)
2 * RQDX3 controllers - 1 for RD54 and dual RX50 drives (internal) and
1 for RD5x drive (external)
KLESI controller (TK50 tape drive (external)
DELQA NIC
2 serial ports (on main CPU board)
4 serial ports (on DHQ11 board)
RT-11 Languages (from memory):
MACRO-11
APL-11
BASIC-11
MU/BASIC-11
BASIC-PLUS/RT-11
FORTRAN-IV (FORTRAN 66 STD)
PDP-11/C
DECUS-C
RSX-11 Languages (also from memory):
MACRO-11
APL-11
BASIC-PLUS-2
FORTRAN-IV (FORTRAN 66 STD)
FORTRAN-77
PDP-11/C
DECUS-C
COBOL-81, COBOL-11
RSTS/E Languages (also from memory):
The same as RSX-11 for the most part.
If you want to have a look around RSTS/E, I can create an account on
PLUTO:: for you to poke around. Just let me know off-list.
-Steve
-----Original Message-----
From: owner-hecnet at Update.UU.SE [mailto:owner-hecnet at Update.UU.SE] On
Behalf Of Mark Wickens
Sent: Thursday, July 14, 2011 15:50
To: hecnet at Update.UU.SE
Subject: Re: [HECnet] PDP Ignorance
Are the MicroPDPs (desktop, deskside, rackmount) rare beasts? They seem
to be of a fairly hobbyist-friendly size, although presumably they are
less expandable (much like a VAXstation versus VAXserver I would
imagine)
Mark.
