Dan Cross writes:
> This came up on the TUHS list back in 2021 (you were on the
> thread, Lars). That pointed to this:
>
> https://www.tuhs.org/cgi-bin/utree.pl?file=BBN-Vax-TCP/history
>
> Which stronly implies that there was some "NCP" in VAX Unix
> sometime in 1980. Whether that was the Network Control Protocol
> or just an affectation for "networking code" (as implied by Noel
> Chiappa in the TUHS thread) is unknown.

Thanks! Yes, that would be unknow. That is almost an optimum point in
time for ambiguity as to what NCP means. DECnet, Chaosnet, and IBM also
picked up the term as roughly equivalent to "network stack". So this
BBN VAX NCP could be either the old Arpanet NCP, or a new TCP stack.

As per the 1981 Transition book, there were already TCP-only hosts on
the Arpanet, implying that some sites developed and deployed TCP well
before that. I get the sense that everyone was aware the switch was
going to happen and new development was towards TCP. But I don't know
the exact timeline.

On Thursday, September 28, 2023 at 3:49:37 PM UTC-4, Lars Brinkhoff wrote:
> Johnny Billquist wrote:
> > There were of course development, and testing done between machines
> > and so on. But that was not "ARPANET". ARPANET was running NCP until
> > flag day, when it officially switched to IP.
> NCP and TCP operated in parallel on the ARPANET for a while. The
> Internet Protocol Transition Workbook from November 1981 encouraged new
> hosts to only implement TCP, not NCP, and says at that point there were
> TCP-only hosts. On several occasions during 1982, NCP was temporarily
> blocked, but TCP was allowed. What happened on flag day was that NCP
> was permanently blocked.
>
> So what I was wondering was: were there any VAXen talking NCP, or did
> they jump straight to TCP? I'd like to see evidence, not handwaving.

Most people reading this thread will all ready know much of the following facts:
1) ARPANET research begins in 1966 (ARPA becomes DARPA in 1972)
2) A lot of people were writing their own client/server modifications before 1982, and much of it was in assembler
3) The various network modifications were not compatible, so DARPA wanted to develop a newer technology which would allow the various networks to interconnect (this is where the second name, internet, comes from)
4) Many people today do not know that UDP was developed 5-6 years 'after' TCP (many think it was the other way around; UDP was primarily developed to aid in packet routing but today it has many other uses (SIP springs to mind))
5) DARPA needed standardized protocols and code; this would best come from one team. Not sure of all the politics, but much of this work eventually came from a gifted programmer at UC Berkeley by the name of Bill Joy. He did a lot of his work on a VAX running BSD UNIX.
6) I'm not certain who moved all the assembler code into C, but once that was done, it was distributed amongst all the universities who were running UNIX systems.
7) I knew a lot of people who were running third party stacks on their Windows and Macintosh systems between 1994 and 1998. At that time network communication interfaces all cost a lot of money, so most newbies were asking questions like "how can I be using this internet stack for free?" My answer was always "anything developed by the US tax payer is usually placed into the public domain".

Neil

In article <7w7co9ijmc.fsf@junk.nocrew.org>,
Lars Brinkhoff <lars.spam@nocrew.org> wrote:
>Dan Cross writes:
>> This came up on the TUHS list back in 2021 (you were on the
>> thread, Lars). That pointed to this:
>>
>> https://www.tuhs.org/cgi-bin/utree.pl?file=BBN-Vax-TCP/history
>>
>> Which stronly implies that there was some "NCP" in VAX Unix
>> sometime in 1980. Whether that was the Network Control Protocol
>> or just an affectation for "networking code" (as implied by Noel
>> Chiappa in the TUHS thread) is unknown.
>
>Thanks! Yes, that would be unknow. That is almost an optimum point in
>time for ambiguity as to what NCP means. DECnet, Chaosnet, and IBM also
>picked up the term as roughly equivalent to "network stack". So this
>BBN VAX NCP could be either the old Arpanet NCP, or a new TCP stack.
>
>As per the 1981 Transition book, there were already TCP-only hosts on
>the Arpanet, implying that some sites developed and deployed TCP well
>before that. I get the sense that everyone was aware the switch was
>going to happen and new development was towards TCP. But I don't know
>the exact timeline.

I, too, find it hard to imagine that a lot of effort would have
been put into a VAX NCP implementation since it was clear that
TCP/IP was coming, but as a stopgap or some special purpose? I
could see it.

I found a copy of the hosts table from 1983:
https://github.com/ttkzw/hosts.txt/blob/master/pub/hosts/19830119/HOSTS.TXT

This lists a number of VAX systems that appear to have been
assigned "ARPANET host numbers", but they could also be "RCCnet"
numbers (I don't think I've ever heard of RCCnet).

This:
https://github.com/ttkzw/hosts.txt/blob/master/pub/hosts/19820615/SYSHST%3B%20HOSTS%20PRETTY
appears to show at several VAXen on ARPANET (BBNF running VMS at
01/05 and UCLA-SECURITY running Unix at 2/01, among others).

Again, it's not entirely clear if these are NCP-hosts, possibly
running TCP/IP, or what. I do feel comfortable assuming they're
not running DECnet (or at least, that's irrelevant to them being
listed in these host tables).

- Dan C.

In article <uf4s1t$cle$6@news.misty.com>,
Johnny Billquist <bqt@softjar.se> wrote:
>On 2023-09-28 21:49, Lars Brinkhoff wrote:
>> Johnny Billquist wrote:
>>> There were of course development, and testing done between machines
>>> and so on. But that was not "ARPANET". ARPANET was running NCP until
>>> flag day, when it officially switched to IP.
>>
>> NCP and TCP operated in parallel on the ARPANET for a while. The
>> Internet Protocol Transition Workbook from November 1981 encouraged new
>> hosts to only implement TCP, not NCP, and says at that point there were
>> TCP-only hosts. On several occasions during 1982, NCP was temporarily
>> blocked, but TCP was allowed. What happened on flag day was that NCP
>> was permanently blocked.
>>
>> So what I was wondering was: were there any VAXen talking NCP, or did
>> they jump straight to TCP? I'd like to see evidence, not handwaving.
>
>How would they interoperate? TCP and NCP are not exactly compatible in
>any way.
>
>You would basically have to have two different parallel networks, and
>them you might have some machines that would act as gateway between the
>two networks.

Well, the statement was that they existed in parallel, not that
they necessarily inter-operated. But communication between them
was documented in the transition guide, and there were gateways
for a while. I even see some evidence in early sendmail
configurations that there were provisions for sending mail to
NCP gateways (e.g., `usr.lib/sendmail/cf/ncphosts.m4` in 4.1c
BSD; UDEL was the NCP gateway and the file has a comment at the
top that says, "When NCP goes away, so should this file").

- Dan C.

On Friday, September 29, 2023 at 5:24:53 AM UTC-7, Neil Rieck wrote:

(snip)

> 7) I knew a lot of people who were running third party stacks on their Windows and Macintosh systems between 1994 and 1998. At that time network communication interfaces all cost a lot of money, so most newbies were asking questions like "how can I be using this internet stack for free?" My answer was always "anything developed by the US tax payer is usually placed into the public domain".
It was just about then, that NIC prices came down to really affordable prices.

I was about then working on school networking projects, where we really could put Ethernet into a school.

But yes, both MS-DOS and MacOS had little support for Ethernet. There was NCSA Telnet, which connected directly to the Ethernet card, with no OS support. (That was free, government funded. There were some non-free versions around.)

After not so long, MacOS had some support, and we ran a different NCSA Telnet. But also about then, Netscape 2.0, which was small enough to run on smaller Macintosh systems.

I do remember buying networking parts on eBay, and often could buy 10 for less than the price of 1.
People who wanted one, didn't bid on 10. (Less likely now, but maybe it still works.)

On 9/29/2023 3:03 PM, gah4 wrote:
> On Friday, September 29, 2023 at 5:24:53 AM UTC-7, Neil Rieck wrote:
>> 7) I knew a lot of people who were running third party stacks on
>> their Windows and Macintosh systems between 1994 and 1998. At that
>> time network communication interfaces all cost a lot of money, so
>> most newbies were asking questions like "how can I be using this
>> internet stack for free?" My answer was always "anything developed
>> by the US tax payer is usually placed into the public domain".
>
> It was just about then, that NIC prices came down to really
> affordable prices.
>
> I was about then working on school networking projects, where we
> really could put Ethernet into a school.
>
> But yes, both MS-DOS and MacOS had little support for Ethernet.
> There was NCSA Telnet, which connected directly to the Ethernet card,
> with no OS support. (That was free, government funded. There were
> some non-free versions around.)
>
> After not so long, MacOS had some support, and we ran a different
> NCSA Telnet. But also about then, Netscape 2.0, which was small
> enough to run on smaller Macintosh systems.

I would assume a lot of the people here ran PathWorks on DOS PC's.

Arne

On Friday, September 29, 2023 at 12:28:46 PM UTC-7, Arne Vajhøj wrote:

(snip)

> > But yes, both MS-DOS and MacOS had little support for Ethernet.
> > There was NCSA Telnet, which connected directly to the Ethernet card,
> > with no OS support. (That was free, government funded. There were
> > some non-free versions around.)
> > After not so long, MacOS had some support, and we ran a different
> > NCSA Telnet. But also about then, Netscape 2.0, which was small
> > enough to run on smaller Macintosh systems.

> I would assume a lot of the people here ran PathWorks on DOS PC's.

I do remember that one. And like NCSA Telnet, had no OS support.

But as noted previously, unless I forgot, you had to pay for that one.
Otherwise, yes, it allowed for DECnet connections.

I do remember having an account on an across the country MicroVAX
reachable by DECnet but not TCP/IP. I had to use the numeric address,
as the PC didn't know it by name.

On 9/29/2023 3:46 PM, gah4 wrote:
> On Friday, September 29, 2023 at 12:28:46 PM UTC-7, Arne Vajhøj wrote:
>>> But yes, both MS-DOS and MacOS had little support for Ethernet.
>>> There was NCSA Telnet, which connected directly to the Ethernet card,
>>> with no OS support. (That was free, government funded. There were
>>> some non-free versions around.)
>
>>> After not so long, MacOS had some support, and we ran a different
>>> NCSA Telnet. But also about then, Netscape 2.0, which was small
>>> enough to run on smaller Macintosh systems.
>
>> I would assume a lot of the people here ran PathWorks on DOS PC's.
>
> I do remember that one. And like NCSA Telnet, had no OS support.

MS did not provide anything. Drivers from the NIC vendor. The rest from
DEC.

> But as noted previously, unless I forgot, you had to pay for that one.

Yes. If I remember correct then one bought a N user license for the
server on VMS and then one could use it on N PC's.

> Otherwise, yes, it allowed for DECnet connections.
>
> I do remember having an account on an across the country MicroVAX
> reachable by DECnet but not TCP/IP. I had to use the numeric address,
> as the PC didn't know it by name.

Same as IP - it is either name or number. Not 256 x 256 x 256 x 256 but
just 64 x 1024.

Arne

On Friday, September 29, 2023 at 12:51:48 PM UTC-7, Arne Vajhøj wrote:

(snip, I wrote)

> > I do remember having an account on an across the country MicroVAX
> > reachable by DECnet but not TCP/IP. I had to use the numeric address,
> > as the PC didn't know it by name.

> Same as IP - it is either name or number. Not 256 x 256 x 256 x 256 but
> just 64 x 1024.

Yes, a five digit number.

And if I remember right, it was through a 56 kbit/s link.

Slow, but fast enough for remote login. Sometimes there was
a delay, though.

That was about 1987, when a fast cross country network used
a T1 line, but this wasn't one. Actually, some of the links might
have been T1, but just not the one I was on.

On 2023-09-29 00:19, gah4 wrote:
> On Thursday, September 28, 2023 at 2:43:28 PM UTC-7, Johnny Billquist wrote:
>
> (snip)
>
>> You would basically have to have two different parallel networks, and
>> them you might have some machines that would act as gateway between the
>> two networks.
>
> I don't remember back to NCP, but I do remember exactly that
> for DECnet and TCP/IP. There were gateways that would transfer
> mail between the two. I believe also some that would allow
> remote login between the two.

There were definitely software that forwarded mail between the two, yes.
Forwarding terminal traffic would also be doable, but it's a bit harder
as there is no conventional way to inform the intermediate hop what end
destination you would like to connect to. So usually, what people would
do, is connect and log in to the host that talked both protocols, and
then use the other protocol to establish the connection to the next host.

Now, you could argue that this means it was possible to remote login
between the two netwowrks, but I think that is sortof stretching the
definition gateway between protocols.

(Heck, I've written a mail package that talks both SMTP and MAIL-11, and
which can forward mail between the two. So that is something that still
happens to this day...)

But that don't mean that DECnet is suddenly a part of TCP/IP.

Same with NCP vs. TCP/IP. You could definitely have gateways, and I'm
sure there were. But ARPANET was talking NCP until flag day, at which
point it switched over to talk TCP/IP.

If you talked TCP/IP before flag day you would have needed some kind of
gateway in order to interact in any way with ARPANET.
After flag day it was the reverse. (If anyone stayed on NCP.)

But I guess, in a sense, you could say that this becomes a question of
the definition of what is/was ARPANET.

Johnny

On 2023-09-29 07:07, Lars Brinkhoff wrote:
> Johnny Billquist wrote:
>>> NCP and TCP operated in parallel on the ARPANET for a while.
>> How would they interoperate? TCP and NCP are not exactly compatible in
>> any way.
>
> Details are found in the "Internet Protocol Transition Workbook".
>
>> you might have some machines that would act as gateway between the two
>> networks.
>
> That's exactly what they did.

That's what I expect. SO we have ARPANET, which is talking NCP, and you
have hosts that talk TCP/IP that can communicate with hosts on the
ARPANET via a gateway. Does that mean the TCP/IP hosts are on ARPANET? I
would say not. Just as hosts on my hobbyst DECnet are not neccesarily on
the internet, but they can communicate with hosts on the internet when
there is some gateway in between that can forward stuff for them.

Johnny

Johnny Billquist wrote:
> But ARPANET was talking NCP until flag day, at which point it switched
> over to talk TCP/IP. [...] But I guess, in a sense, you could say
> that this becomes a question of the definition of what is/was ARPANET.

Already in 1981 it was encouraged that new ARPANET hosts only implement
TCP, leap-frogging NCP. It doesn't seem plausible to me it was thought
that those hosts were operating outside ARPANET.

On 2023-10-02 12:45, Lars Brinkhoff wrote:
> Johnny Billquist wrote:
>> But ARPANET was talking NCP until flag day, at which point it switched
>> over to talk TCP/IP. [...] But I guess, in a sense, you could say
>> that this becomes a question of the definition of what is/was ARPANET.
>
> Already in 1981 it was encouraged that new ARPANET hosts only implement
> TCP, leap-frogging NCP. It doesn't seem plausible to me it was thought
> that those hosts were operating outside ARPANET.

I would consider it to be a case of:
We know we are going to switch to TCP/IP soon, so it makes no sense that
you implement NCP. Until we switch, you can get partial participation
via gateways. And you can of course talk directly with others who are
also running TCP/IP.

Johnny

On Monday, October 2, 2023 at 3:16:30 AM UTC-7, Johnny Billquist wrote:
> On 2023-09-29 00:19, gah4 wrote:

(snip)

> > I don't remember back to NCP, but I do remember exactly that
> > for DECnet and TCP/IP. There were gateways that would transfer
> > mail between the two. I believe also some that would allow
> > remote login between the two.

> There were definitely software that forwarded mail between the two, yes.
> Forwarding terminal traffic would also be doable, but it's a bit harder
> as there is no conventional way to inform the intermediate hop what end
> destination you would like to connect to. So usually, what people would
> do, is connect and log in to the host that talked both protocols, and
> then use the other protocol to establish the connection to the next host.
> Now, you could argue that this means it was possible to remote login
> between the two netwowrks, but I think that is sortof stretching the
> definition gateway between protocols.

It is some years now, so I don't remember the details, but I am pretty
sure that there was one that worked even if you didn't have an account.

It might have been logically the same as logging in, one didn't
actually log in.

It might be that you supplied the user@host on the LOGIN prompt,
which then did the connection. Or, the other way, you put the
HOST::USER in the LOGIN: prompt.

It is pretty many years now, and not so many hosts did it.

On Monday, October 2, 2023 at 3:16:30 AM UTC-7, Johnny Billquist wrote:

(snip)

> There were definitely software that forwarded mail between the two, yes.
> Forwarding terminal traffic would also be doable, but it's a bit harder
> as there is no conventional way to inform the intermediate hop what end
> destination you would like to connect to. So usually, what people would
> do, is connect and log in to the host that talked both protocols, and
> then use the other protocol to establish the connection to the next host.

It is described here:

http://www.bitsavers.org/pdf/dec/vax/ultrix-32/DECnet_Ultrix_4.0/AA-JQ71C-TE_DECnet_Ultrix_4.0_DECnet-Internet_Gateway_Use_and_Management_May1990.pdf

You connect to the gateway with either SET HOST or telnet, then put

host::

or

host!

into the login: prompt on the Ultrix system.

In article <ufe5e0$nbu$7@news.misty.com>,
Johnny Billquist <bqt@softjar.se> wrote:
>On 2023-09-29 07:07, Lars Brinkhoff wrote:
>> Johnny Billquist wrote:
>>>> NCP and TCP operated in parallel on the ARPANET for a while.
>>> How would they interoperate? TCP and NCP are not exactly compatible in
>>> any way.
>>
>> Details are found in the "Internet Protocol Transition Workbook".
>>
>>> you might have some machines that would act as gateway between the two
>>> networks.
>>
>> That's exactly what they did.
>
>That's what I expect. SO we have ARPANET, which is talking NCP, and you
>have hosts that talk TCP/IP that can communicate with hosts on the
>ARPANET via a gateway. Does that mean the TCP/IP hosts are on ARPANET? I
>would say not. Just as hosts on my hobbyst DECnet are not neccesarily on
>the internet, but they can communicate with hosts on the internet when
>there is some gateway in between that can forward stuff for them.

The layering is not quite right here. NCP was essentially a
transport protocol, and the IMPs provided lower-level network
protocol services; in this sense, NCP is closer to TCP than to
IP. The IMPs, in turn, used a protocol that was commonly called
"1822" (from a BBN technical report) to communicate with ARPANET
hosts; the initial TCP/IP implementations hosted on ARPANET fed
IP datagrams directly to IMPs using 1822.

See, e.g., IEN 28, sec 1.4 ["Interfaces"]. To quote:
|In the ARPANET case, for example, the Internet module would
|call on a local net module which would add the 1822 leader [6]
|to the internet segment creating an ARPANET message to transmit
|to the IMP.
(From: https://www.rfc-editor.org/ien/ien28.pdf)

The ARPANET was the first backbone for internetworking using IP,
but TCP/IP and NCP sort of existed in quasi-parallel at the
time. So TCP/IP hosts were very much "on the ARPANET", in the
sense that they used the packet network of IMPs for
communication in the same way that NCP-only hosts did.

- Dan C.

On 2023-10-02 13:39, gah4 wrote:
> On Monday, October 2, 2023 at 3:16:30 AM UTC-7, Johnny Billquist wrote:
>> On 2023-09-29 00:19, gah4 wrote:
>
> (snip)
>
>>> I don't remember back to NCP, but I do remember exactly that
>>> for DECnet and TCP/IP. There were gateways that would transfer
>>> mail between the two. I believe also some that would allow
>>> remote login between the two.
>
>> There were definitely software that forwarded mail between the two, yes.
>> Forwarding terminal traffic would also be doable, but it's a bit harder
>> as there is no conventional way to inform the intermediate hop what end
>> destination you would like to connect to. So usually, what people would
>> do, is connect and log in to the host that talked both protocols, and
>> then use the other protocol to establish the connection to the next host.
>
>> Now, you could argue that this means it was possible to remote login
>> between the two netwowrks, but I think that is sortof stretching the
>> definition gateway between protocols.
>
> It is some years now, so I don't remember the details, but I am pretty
> sure that there was one that worked even if you didn't have an account.
>
> It might have been logically the same as logging in, one didn't
> actually log in.
>
> It might be that you supplied the user@host on the LOGIN prompt,
> which then did the connection. Or, the other way, you put the
> HOST::USER in the LOGIN: prompt.

That would possibly have been how I would do it.

But that is basically just the same as logging in to the intermediate
machine and starting a new session from there. There isn't really a
protocol translation directly between the two sides as such.

> It is pretty many years now, and not so many hosts did it.

I'm trying to remember if I saw/heard of something like that. I might
have, but I might just also be making that up as I write this. Too long
ago...

Johnny

On 2023-10-02 14:56, Dan Cross wrote:
> In article <ufe5e0$nbu$7@news.misty.com>,
> Johnny Billquist <bqt@softjar.se> wrote:
>> On 2023-09-29 07:07, Lars Brinkhoff wrote:
>>> Johnny Billquist wrote:
>>>>> NCP and TCP operated in parallel on the ARPANET for a while.
>>>> How would they interoperate? TCP and NCP are not exactly compatible in
>>>> any way.
>>>
>>> Details are found in the "Internet Protocol Transition Workbook".
>>>
>>>> you might have some machines that would act as gateway between the two
>>>> networks.
>>>
>>> That's exactly what they did.
>>
>> That's what I expect. SO we have ARPANET, which is talking NCP, and you
>> have hosts that talk TCP/IP that can communicate with hosts on the
>> ARPANET via a gateway. Does that mean the TCP/IP hosts are on ARPANET? I
>> would say not. Just as hosts on my hobbyst DECnet are not neccesarily on
>> the internet, but they can communicate with hosts on the internet when
>> there is some gateway in between that can forward stuff for them.
>
> The layering is not quite right here. NCP was essentially a
> transport protocol, and the IMPs provided lower-level network
> protocol services; in this sense, NCP is closer to TCP than to
> IP. The IMPs, in turn, used a protocol that was commonly called
> "1822" (from a BBN technical report) to communicate with ARPANET
> hosts; the initial TCP/IP implementations hosted on ARPANET fed
> IP datagrams directly to IMPs using 1822.
>
> See, e.g., IEN 28, sec 1.4 ["Interfaces"]. To quote:
> |In the ARPANET case, for example, the Internet module would
> |call on a local net module which would add the 1822 leader [6]
> |to the internet segment creating an ARPANET message to transmit
> |to the IMP.
> (From: https://www.rfc-editor.org/ien/ien28.pdf)
>
> The ARPANET was the first backbone for internetworking using IP,
> but TCP/IP and NCP sort of existed in quasi-parallel at the
> time. So TCP/IP hosts were very much "on the ARPANET", in the
> sense that they used the packet network of IMPs for
> communication in the same way that NCP-only hosts did.

It again goes into what do we mean when we say "ARPNANET".
Just because you had other protocols using the same underlying
infrastructure, does it mean they are part of the same network?
I would say not.
Just as with VPNs. They are all on the same physical network, but you
can't speak directly between them without a gateway that forwards the
traffic between the two.
But the "problem" with ARPANET is that it was speaking another protocol
that isn't interoperable with TCP before flag day. So, ARPANET was NCP.
The fact that other protocols also operated over the same infrastructure
don't mean they were ARPANET.

Johnny

In article <ufeno9$g48$2@news.misty.com>,
Johnny Billquist <bqt@softjar.se> wrote:
>On 2023-10-02 14:56, Dan Cross wrote:
>> In article <ufe5e0$nbu$7@news.misty.com>,
>> Johnny Billquist <bqt@softjar.se> wrote:
>>> On 2023-09-29 07:07, Lars Brinkhoff wrote:
>>>> Johnny Billquist wrote:
>>>>>> NCP and TCP operated in parallel on the ARPANET for a while.
>>>>> How would they interoperate? TCP and NCP are not exactly compatible in
>>>>> any way.
>>>>
>>>> Details are found in the "Internet Protocol Transition Workbook".
>>>>
>>>>> you might have some machines that would act as gateway between the two
>>>>> networks.
>>>>
>>>> That's exactly what they did.
>>>
>>> That's what I expect. SO we have ARPANET, which is talking NCP, and you
>>> have hosts that talk TCP/IP that can communicate with hosts on the
>>> ARPANET via a gateway. Does that mean the TCP/IP hosts are on ARPANET? I
>>> would say not. Just as hosts on my hobbyst DECnet are not neccesarily on
>>> the internet, but they can communicate with hosts on the internet when
>>> there is some gateway in between that can forward stuff for them.
>>
>> The layering is not quite right here. NCP was essentially a
>> transport protocol, and the IMPs provided lower-level network
>> protocol services; in this sense, NCP is closer to TCP than to
>> IP. The IMPs, in turn, used a protocol that was commonly called
>> "1822" (from a BBN technical report) to communicate with ARPANET
>> hosts; the initial TCP/IP implementations hosted on ARPANET fed
>> IP datagrams directly to IMPs using 1822.
>>
>> See, e.g., IEN 28, sec 1.4 ["Interfaces"]. To quote:
>> |In the ARPANET case, for example, the Internet module would
>> |call on a local net module which would add the 1822 leader [6]
>> |to the internet segment creating an ARPANET message to transmit
>> |to the IMP.
>> (From: https://www.rfc-editor.org/ien/ien28.pdf)
>>
>> The ARPANET was the first backbone for internetworking using IP,
>> but TCP/IP and NCP sort of existed in quasi-parallel at the
>> time. So TCP/IP hosts were very much "on the ARPANET", in the
>> sense that they used the packet network of IMPs for
>> communication in the same way that NCP-only hosts did.
>
>It again goes into what do we mean when we say "ARPNANET".

The historical record shows that the players at the time meant
the network of IMPs and the hosts that connected to them. It
seems pretty clear that they didn't _just_ mean NCP.

>Just because you had other protocols using the same underlying
>infrastructure, does it mean they are part of the same network?
>I would say not.

This is arguing semantics to an extent, but to answer this
question, I would describe such an arrangement as different
applications of the underlying network.

>Just as with VPNs. They are all on the same physical network, but you
>can't speak directly between them without a gateway that forwards the
>traffic between the two.
>But the "problem" with ARPANET is that it was speaking another protocol
>that isn't interoperable with TCP before flag day.

Yes, it was speaking 1822. :-) It spoke 1822 after flag day,
too.

>So, ARPANET was NCP.

In this case, the historical record is clear: ARPANET was the
physical set of IMPs and their interconnecting lines. It was
not just NCP.

>The fact that other protocols also operated over the same infrastructure
>don't mean they were ARPANET.

Well, in this case, NCP came after 1822; initially, hosts used
1822 directly for host-to-host communication in the context of
ARPANET, but that proved unsatisfactory, so NCP was designed and
implemented. So given that "ARPANET" predated NCP, it seems
unfair to redefine the former to mean the latter, particularly
when it's pretty clear that that was not how the people working
on it at the time thought of it (as can be seen from the
afore referenced IENs, for example).

- Dan C.

gah4 <gah4@u.washington.edu> wrote:
>It is some years now, so I don't remember the details, but I am pretty
>sure that there was one that worked even if you didn't have an account.

Decnet to arpa? Sure, there were lots of them and none that I know
required an account. It was just a polite service people provided.
The best one was at Columbia which had really good connectivity (and also
bitnet connectivity) so you could do "fredbox::fred@columbia.edu" as I
recall.
--scott

--
"C'est un Nagra. C'est suisse, et tres, tres precis."

On Monday, October 2, 2023 at 4:58:49 PM UTC-7, Scott Dorsey wrote:
> gah4 <ga...@u.washington.edu> wrote:
> >It is some years now, so I don't remember the details, but I am pretty
> >sure that there was one that worked even if you didn't have an account.

> Decnet to arpa? Sure, there were lots of them and none that I know
> required an account. It was just a polite service people provided.
> The best one was at Columbia which had really good connectivity (and also
> bitnet connectivity) so you could do "fredbox::fr...@columbia.edu" as I
> recall.

Seems to be a feature of Ultrix.

I presume it can be turned on and off.

On Monday, September 25, 2023 at 5:15:50 PM UTC-7, Dan Cross wrote:
> The early days of Windows NT are well-documented in the book,
> "Show Stopper!" by G. Pascal Zachary. In short, they used OS/2
> and 386 machines; NT was self-hosting within a couple of years.

Zachary wasn't very technical and made a number of mistakes in that book, although overall his story is well-researched and compelling. NT was originally brought up on single-board Intel i860 hardware, followed by MIPS DECstations and then i386 hardware; Cutler insisted that the team not focus on i386 because he wanted to keep NT from becoming wedded to the x86 architecture.

On Sunday, September 24, 2023 at 9:10:04 AM UTC-7, John Dallman wrote:
> Nor did they look at the history of the art of making computers
> faster. The VAX architecture was implemented readily enough at first, but
> made pipelining, out-of-order and other ideas that had been invented in
> the 1950s and 1960s hard to add.

I don't think it's reasonable to say they "didn't look at the history..." so much as made a set of architectural choices that later cornered them. The VAX roadmap had pipelining planned for the 8X series before 11/780 even shipped; I think it's more accurate to say they made a set of serious architectural mistakes because they were optimizing for tradeoffs that seemed reasonable at the time.

On Sunday, September 24, 2023 at 11:04:53 AM UTC-7, John Dallman wrote:

> > If DEC had went after the low end market with the C-VAX, I really
> > feel that DEC would still be with us today.
> Maybe. The MS-DOS hardware and software industry was already very well
> established, and competition had driven hardware prices down a lot. DEC
> would have had to pick some niches to target and win several of them.

Agreed. By the time CVAX launched, the damage had already been done -- CVAX was not price-competitive with x86. Even Alpha wasn't really price-competitive with Pentium.

The opportunity (if it ever existed) was squandered much earlier, before the first MicroVAX parts shipped, when Olsen decided to kill the strategy to sell them to OEMs and drive volume. John Mashey describes a number of ways where even that strategy might have failed, but there was zero chance of pulling it off by 1987.

On Monday, October 2, 2023 at 10:43:19 PM UTC-7, jimc...@gmail.com wrote:

(snip)

> The issue with VAX 9000 wasn't that it was an implementation of the VAX CISC architecture; the issue was chasing IBM with a massive ECL implementation of VAX, with all the associated costs in power and cooling and engineering required to address them. By the time VAX 9000 launched, CMOS VAX was already faster and dramatically cheaper.

This is what everyone had to figure out.

ECL (and STTL) have a different scaling law from MOS.

ECL and STTL stay at the same voltage, which is related to the
band gap, as transistors get smaller.

Well, you can do that with MOS, too, but Denard scaling shrinks
the oxide along with other dimensions. That reduces supply voltage,
and so power.

In 1978, CMOS was slower than TTL, and harder to build than NMOS.

So, there is a transition when CMOS gets faster, and lower power,
(and power density) the transition was made.

The other problem, at least for some years, with CMOS is
parasitic SCRs. The way the PN and NP junctions combine,
can lead to the configuration of an SCR. If you manage to
turn it in, it is a direct connection across the power supply,
sometimes destructively.

I don't remember the timeline for CMOS VAX, though.
(Even though I have a MicroVAX 2000 and 3100.)