3 Oct
2016
3 Oct
'16
8:37 a.m.
From: Al Kossow
Stack Limit
That's there - that's the KJ11.
From: Tony Duell
KW11-L line time clock?
I don't really consider the LTC as a CPU option. It's on the actual UNIBUS,
it's just in a specially wired UNIBUS slot that's only one slot wide, and only
has BR/G6 and D6/7 wired to it. Yes, in this case that slot is in the CPU, so
one can make an argument...
Noel
3 Oct
3 Oct
9:07 a.m.
On Mon, Oct 3, 2016 at 11:37 AM, Noel Chiappa <jnc at mercury.lcs.mit.edu>
wrote:
I should have been more precise, but my point is that this is a non-parity
RAM basic 11/35 system useful for peripheral interfacing and
communications. As is, not good for something like RT11.
b
From: Al
Kossow
Stack Limit
That's there - that's the KJ11.
From: Tony Duell
KW11-L line time clock?
I don't really consider the LTC as a CPU option. It's on the actual UNIBUS,
it's just in a specially wired UNIBUS slot that's only one slot wide, and
only
has BR/G6 and D6/7 wired to it. Yes, in this case that slot is in the CPU,
so
one can make an argument...
Noel
9:13 a.m.
On Oct 3, 2016, at 12:07 PM, william degnan
<billdegnan at gmail.com> wrote:
...
I should have been more precise, but my point is that this is a non-parity
RAM basic 11/35 system useful for peripheral interfacing and
communications. As is, not good for something like RT11.
I wouldn't look at it this way. Non-parity memory is fine for any application if you
judge the risk of memory error to be low enough for that application. Peripheral
interfacing or communications may or may not be such an application. An application
running on RT11 may or may not be. The OS isn't what decides this, but rather the
requirements of the application: the expected bit error rate for the memory technology in
use vs. the failure rate that's tolerable for the application.
paul
9:46 a.m.
On Mon, Oct 3, 2016 at 12:13 PM, Paul Koning <paulkoning at comcast.net> wrote:
16K RAM does not make for much of an RT11 system. The industrial/beefy
11/40's all have parity RAM. My PDP 11/05 has the same core stack as this
11/35.
On Oct 3, 2016, at 12:07 PM, william degnan
<billdegnan at gmail.com>
wrote:
...
I should have been more precise, but my point is that this is a
non-parity
RAM basic 11/35 system useful for peripheral
interfacing and
communications. As is, not good for something like RT11.
I wouldn't look at it this way. Non-parity memory is fine for any
application if you judge the risk of memory error to be low enough for that
application. Peripheral interfacing or communications may or may not be
such an application. An application running on RT11 may or may not be.
The OS isn't what decides this, but rather the requirements of the
application: the expected bit error rate for the memory technology in use
vs. the failure rate that's tolerable for the application.
paul
11:05 a.m.
On 03/10/2016 17:46, william degnan wrote:
On Mon, Oct 3, 2016 at 12:13 PM, Paul Koning
<paulkoning at comcast.net> wrote:
>> On Oct 3, 2016, at 12:07 PM, william degnan <billdegnan at gmail.com>
> wrote:
>> ...
>> I should have been more precise, but my point is that this is a non-parity
>> RAM basic 11/35 system useful for peripheral interfacing and
>> communications. As is, not good for something like RT11.
>
> I wouldn't look at it this way. Non-parity memory is fine for any
> application if you judge the risk of memory error to be low enough
16K RAM does not make for much of an RT11 system. The
industrial/beefy
11/40's all have parity RAM. My PDP 11/05 has the same core stack as this
11/35.
I agree with Paul. I've seen dozens of RT-11 systems and non-parity MOS
memory has been more common in the ones I've seen. I see no reason to
think it would be a problem. True, 16K is small even for RT-11, but
I've seen plenty like that too.
--
Pete
Pete Turnbull
11:27 a.m.
On Oct 3, 2016, at 2:05 PM, Pete Turnbull <pete at
dunnington.plus.com> wrote:
...
I agree with Paul. I've seen dozens of RT-11 systems and non-parity MOS memory has
been more common in the ones I've seen. I see no reason to think it would be a
problem. True, 16K is small even for RT-11, but I've seen plenty like that too.
16 kwords? That's at least 2x the minimum for older versions. I did my honors work
in college on an 11/20 with 8 kW... which was big enough not just for RT11, but for
RT-BASIC and some reasonable sized applications.
paul
12:51 p.m.
On Mon, Oct 3, 2016 at 2:27 PM, Paul Koning <paulkoning at comcast.net> wrote:
fine. I am not going to win. Is it *possible* yes...Hoping you can take
the spirit of my original comments - this is not a beefed up 11/35 is all I
am trying to say. If you want to run some old copy of RT 11 on this
machine by all means do it.
On Oct 3, 2016, at 2:05 PM, Pete Turnbull
<pete at dunnington.plus.com>
wrote:
...
I agree with Paul. I've seen dozens of RT-11 systems and non-parity MOS
memory has been more common in the ones I've seen. I see no reason to
think it would be a problem. True, 16K is small even for RT-11, but I've
seen plenty like that too.
16 kwords? That's at least 2x the minimum for older versions. I did my
honors work in college on an 11/20 with 8 kW... which was big enough not
just for RT11, but for RT-BASIC and some reasonable sized applications.
paul
2783
days inactive
2783
days old
6 comments
4 participants
participants (4)
-
billdegnan＠gmail.com -
jnc＠mercury.lcs.mit.edu -
paulkoning＠comcast.net -
pete＠dunnington.plus.com