On 2012-06-23 00:25, Peter Coghlan wrote:
Peter is this readable?
Hans
Sorry to be jumping into this late but I've been away.
I am confused as to the source of the problem.
I read the hecnet list using VMS Mail and I therefore get to see all the
headers in their gorey detail. I've never seen any mail from the list arrive
with base64 encoding.
I've had a quick look through the headers for all the hecnet mail for the
last few days. I find some mails have Content-transfer-encoding: 7BIT and others
have Content-transfer-encoding: 8BIT (including some that are not from Hans).
Just one mail (the HTML section of a mail from Jordi Guillaumes i Pons)
had Content-transfer-encoding: quoted-printable and none at all
had Content-transfer-encoding: base64. In particular, the ones from Hans do not
arrive here with Content-transfer-encoding: base64
I can't see anything special about the mails from Hans that would be causing
those and only those to result in difficulties on a 7bit only system.
Perhaps a mail transfer agent processing mail for the system experiencing the
problem is making incorrect assumptions about some of the mails going through
it and somewhat arbitrarily deciding to convert some of them to base64?
(I do see some headers such as X-Spam-Status: containing references to base64
but this should have no bearing on how the mail is processed.)
No. Some mails from Hans was encoded in Base64. If you didn't see that, then there is something going on at your side to hide it.
But I think Hans solved this.
Johnny
--
Johnny Billquist || "I'm on a bus
|| on a psychedelic trip
email: bqt at softjar.se || Reading murder books
pdp is alive! || tryin' to stay hip" - B. Idol
RSX would be great too. There was a rumour that somebody got it on a Pro, but I never managed to get a copy...
--
Wolfgang Eichberger - OE5EWL
Operating System Collector
Blog: 5ewl.blogspot.com
Homepage: www.eichberger.org
Ahnenforschung / Genealogy: A(E)ichberger, B(P)ruckmayr Raum Leonding/Alkoven/Eferding; Schmeisser Raum Attergau
2012/6/24 Johnny Billquist <bqt at softjar.se>
On 2012-06-24 11:53, Wolfgang Eichberger wrote:
I would greatly appreciate RSTS for my Pro380 too if you would like to
share it....
For myself P/OS was never my thing... RSTS would be much nicer on this
box...
Speaking as an RSX fanatic - was P/OS every the right thing for anyone...???
Now, if I could get proper RSX on that machine...
Johnny
Regards,
Wolfgang
--
Wolfgang Eichberger - OE5EWL
Operating System Collector
Blog: 5ewl.blogspot.com <http://5ewl.blogspot.com>
Homepage: www.eichberger.org <http://www.eichberger.org>
Ahnenforschung / Genealogy: A(E)ichberger, B(P)ruckmayr Raum
Leonding/Alkoven/Eferding; Schmeisser Raum Attergau
2012/6/23 Dave McGuire <mcguire at neurotica.com
<mailto:mcguire at neurotica.com>>
On 06/23/2012 04:43 PM, Paul_Koning at Dell.com wrote:
>> Urr? Doesn't the CNA use an Am7990 "Lance" chip? As I recall
>> that isn't too bad to program. I must be misremembering what chip
>> it uses..?
>
> Intel 82586, which has the worst programming interface in the history
> of Ethernet. It uses a single linked list, which is accessed both by
> the driver and by the chip, without interlocking. Race conditions
> everywhere...
Ohhhhhhhhh, that one! I remember that chip. That was an "ie"
(Intel Ethernet) in Sun parlance. The "le" (Lance Ethernet) found on
some machines was far preferable. I had to write a driver for an Am7990
for an embedded application a long time ago; I remember that wasn't too
bad. I've never really looked at the 82586 at the programming level; I
just remembered not to try to put too much traffic through them when
building networks of Sun machines. :)
There's Linux support for it, for which source code is readily
available, and I have some earlier Solaris source code which will have
some code. Maybe pick through that for some tricks. But man that
sounds dangerous, no access clash protection...ugh.
>> Umm...are you willing to share that stuff? =) I run RSTS 10.1 on
>> many of my PDP-11s; I'd LOVE to run it on my Pro-350!
>
> Sure, if I can figure out a way to do it reasonably cleanly.
That would be great...Thank you!
-Dave
--
Dave McGuire, AK4HZ
New Kensington, PA
--
Johnny Billquist || "I'm on a bus
|| on a psychedelic trip
email: bqt at softjar.se || Reading murder books
pdp is alive! || tryin' to stay hip" - B. Idol
On 2012-06-24 11:53, Wolfgang Eichberger wrote:
I would greatly appreciate RSTS for my Pro380 too if you would like to
share it....
For myself P/OS was never my thing... RSTS would be much nicer on this
box...
Speaking as an RSX fanatic - was P/OS every the right thing for anyone...???
Now, if I could get proper RSX on that machine...
Johnny
Regards,
Wolfgang
--
Wolfgang Eichberger - OE5EWL
Operating System Collector
Blog: 5ewl.blogspot.com <http://5ewl.blogspot.com>
Homepage: www.eichberger.org <http://www.eichberger.org>
Ahnenforschung / Genealogy: A(E)ichberger, B(P)ruckmayr Raum
Leonding/Alkoven/Eferding; Schmeisser Raum Attergau
2012/6/23 Dave McGuire <mcguire at neurotica.com
<mailto:mcguire at neurotica.com>>
On 06/23/2012 04:43 PM, Paul_Koning at Dell.com wrote:
>> Urr? Doesn't the CNA use an Am7990 "Lance" chip? As I recall
>> that isn't too bad to program. I must be misremembering what chip
>> it uses..?
>
> Intel 82586, which has the worst programming interface in the history
> of Ethernet. It uses a single linked list, which is accessed both by
> the driver and by the chip, without interlocking. Race conditions
> everywhere...
Ohhhhhhhhh, that one! I remember that chip. That was an "ie"
(Intel Ethernet) in Sun parlance. The "le" (Lance Ethernet) found on
some machines was far preferable. I had to write a driver for an Am7990
for an embedded application a long time ago; I remember that wasn't too
bad. I've never really looked at the 82586 at the programming level; I
just remembered not to try to put too much traffic through them when
building networks of Sun machines. :)
There's Linux support for it, for which source code is readily
available, and I have some earlier Solaris source code which will have
some code. Maybe pick through that for some tricks. But man that
sounds dangerous, no access clash protection...ugh.
>> Umm...are you willing to share that stuff? =) I run RSTS 10.1 on
>> many of my PDP-11s; I'd LOVE to run it on my Pro-350!
>
> Sure, if I can figure out a way to do it reasonably cleanly.
That would be great...Thank you!
-Dave
--
Dave McGuire, AK4HZ
New Kensington, PA
--
Johnny Billquist || "I'm on a bus
|| on a psychedelic trip
email: bqt at softjar.se || Reading murder books
pdp is alive! || tryin' to stay hip" - B. Idol
I would greatly appreciate RSTS for my Pro380 too if you would like to share it....
For myself P/OS was never my thing... RSTS would be much nicer on this box...
Regards,
Wolfgang
--
Wolfgang Eichberger - OE5EWL
Operating System Collector
Blog: 5ewl.blogspot.com
Homepage: www.eichberger.org
Ahnenforschung / Genealogy: A(E)ichberger, B(P)ruckmayr Raum Leonding/Alkoven/Eferding; Schmeisser Raum Attergau
2012/6/23 Dave McGuire <mcguire at neurotica.com>
On 06/23/2012 04:43 PM, Paul_Koning at Dell.com wrote:
>> Urr? Doesn't the CNA use an Am7990 "Lance" chip? As I recall
>> that isn't too bad to program. I must be misremembering what chip
>> it uses..?
>
> Intel 82586, which has the worst programming interface in the history
> of Ethernet. It uses a single linked list, which is accessed both by
> the driver and by the chip, without interlocking. Race conditions
> everywhere...
Ohhhhhhhhh, that one! I remember that chip. That was an "ie"
(Intel Ethernet) in Sun parlance. The "le" (Lance Ethernet) found on
some machines was far preferable. I had to write a driver for an Am7990
for an embedded application a long time ago; I remember that wasn't too
bad. I've never really looked at the 82586 at the programming level; I
just remembered not to try to put too much traffic through them when
building networks of Sun machines. :)
There's Linux support for it, for which source code is readily
available, and I have some earlier Solaris source code which will have
some code. Maybe pick through that for some tricks. But man that
sounds dangerous, no access clash protection...ugh.
>> Umm...are you willing to share that stuff? =) I run RSTS 10.1 on
>> many of my PDP-11s; I'd LOVE to run it on my Pro-350!
>
> Sure, if I can figure out a way to do it reasonably cleanly.
That would be great...Thank you!
-Dave
--
Dave McGuire, AK4HZ
New Kensington, PA
On 06/23/2012 04:43 PM, Paul_Koning at Dell.com wrote:
Urr? Doesn't the CNA use an Am7990 "Lance" chip? As I recall
that isn't too bad to program. I must be misremembering what chip
it uses..?
Intel 82586, which has the worst programming interface in the history
of Ethernet. It uses a single linked list, which is accessed both by
the driver and by the chip, without interlocking. Race conditions
everywhere...
Ohhhhhhhhh, that one! I remember that chip. That was an "ie"
(Intel Ethernet) in Sun parlance. The "le" (Lance Ethernet) found on
some machines was far preferable. I had to write a driver for an Am7990
for an embedded application a long time ago; I remember that wasn't too
bad. I've never really looked at the 82586 at the programming level; I
just remembered not to try to put too much traffic through them when
building networks of Sun machines. :)
There's Linux support for it, for which source code is readily
available, and I have some earlier Solaris source code which will have
some code. Maybe pick through that for some tricks. But man that
sounds dangerous, no access clash protection...ugh.
Umm...are you willing to share that stuff? =) I run RSTS 10.1 on
many of my PDP-11s; I'd LOVE to run it on my Pro-350!
Sure, if I can figure out a way to do it reasonably cleanly.
That would be great...Thank you!
-Dave
--
Dave McGuire, AK4HZ
New Kensington, PA
On Jun 22, 2012, at 8:56 PM, Dave McGuire wrote:
On 06/22/2012 05:33 PM, Paul_Koning at Dell.com wrote:
...
What's missing is CNA support (Ethernet). The chip used on that card
sucks quite severely, and I've never had enough energy to write the
driver.
Urr? Doesn't the CNA use an Am7990 "Lance" chip? As I recall that
isn't too bad to program. I must be misremembering what chip it uses..?
Intel 82586, which has the worst programming interface in the history of Ethernet. It uses a single linked list, which is accessed both by the driver and by the chip, without interlocking. Race conditions everywhere...
But I have disk, and display, and serial ports including the
obscure 4-port card, DDCMP support in both sync and async mode...
Wow, that is fantastic! (I keep saying "wow"! ;))
Umm...are you willing to share that stuff? =) I run RSTS 10.1 on many
of my PDP-11s; I'd LOVE to run it on my Pro-350!
Sure, if I can figure out a way to do it reasonably cleanly.
paul
On Jun 23, 2012, at 2:51 AM, Rob Jarratt wrote:
-----Original Message-----
From: owner-hecnet at Update.UU.SE [mailto:owner-hecnet at Update.UU.SE]
On Behalf Of Paul_Koning at Dell.com
Sent: 23 June 2012 01:49
To: hecnet at Update.UU.SE
Subject: Re: [HECnet] DECnet User Mode Router - Encapsulation Formats
...
Version 2? Do you mean Level 2?
No I mean version 2, as in this spec here:
http://www.openvms.compaq.com/wizard/decnet/route20.txt
I forgot about those confusing version numbers. That's DECnet Phase IV. (The "routing 1.x" spec is DECnet phase III, because that's when the routing layer first appeared -- DECnet phase II didn't have one.)
paul
On 06/22/2012 02:10 AM, Gregg Levine wrote:
Hello! Pontus I've met Adam (LINC-8) at the VCF earlier this year, that's where David and I met (officially that is) but not the 2060. What was both of them being used at the school for?
The LINC-8 was used by the department of psychology to control experiments related to human perception.
The 2060 was the main computer for the Computer Science department, used by researchers, teachers and students and later by the Update computer club for several years..
/P
It is going to be a while before I am ready to debug seriously against an
existing tunnel, but if you could make one available in any case that would
be great so I can start to see what things look like. Beware though that I
am very likely to cause stuff to break!
Regards
Rob
-----Original Message-----
From: owner-hecnet at Update.UU.SE [mailto:owner-hecnet at Update.UU.SE]
On Behalf Of Peter Lothberg
Sent: 23 June 2012 15:49
To: hecnet at Update.UU.SE
Cc: hecnet at Update.UU.SE
Subject: RE: [HECnet] DECnet User Mode Router - Encapsulation Formats
The Multinet ecap Stuart explained a while ago here on HECnet, I'll resend
it.
GRE is a IP protocol, like TCP and UDP.
If you encap anything in GRE you have:
IP header
GRE Header
(Protocol you are transporting part of GRE header represented with the
16 bit ether-type)
Payload.
csum
In the case of DECnet the cisco encap is take the stuff you put in a
etchernet
packet, but skip headers and start at the protocol ID. Remember check what
hapens with packets that decnet multicats on a lan.
http://en.wikipedia.org/wiki/Generic_Routing_Encapsulation
Let me knew if you run in to problems or need a tunnel to debug against.
--Peter
-----Original Message-----
From: owner-hecnet at Update.UU.SE [mailto:owner-
hecnet at Update.UU.SE]
On Behalf Of Dave McGuire
Sent: 22 June 2012 19:34
To: hecnet at Update.UU.SE
Subject: Re: [HECnet] DECnet User Mode Router - Encapsulation
Formats
On 06/22/2012 02:31 PM, Rob Jarratt wrote:
I am making a little progress on a user mode DECnet router that
will run on Raspberry Pi and on Windows. There is still a long way
to go, but I want to get the basic design right in terms on the
network
interfacing.
Can someone tell me how Cisco and Multinet encapsulate DECnet
packets sent over the internet?
Ciscos can encapsulate DECnet within either GRE or L2TP.
-Dave
--
Dave McGuire, AK4HZ
New Kensington, PA
I hope GRE and L2TP aren't too complex. It *looks* like GRE can be
transported over UDP. L2TP looks a bit more complex, but I have
skimmed the Wikipedia page for all of 15 seconds so I may have missed
something.
Regards
Rob