[TUHS] BSD/v8 TCP/IP

[TUHS] BSD/v8 TCP/IP

[Doug McIlroy]

Norman wrote

The earliest stream-I/O-system-based tty driver I'm aware of was
already in the Research kernel when I interviewed at Bell Labs
in early 1984.  I have a vague memory that it was a couple of
years older than that, but I cannot find any citations to
back up either of those memories.

Dennis described streams in the special Unix issue of the
BSTJ, Oct 1984, and noted that "it runs on about 20 machines in
... Research ... Although it is being investigated in other
parts of AT&T Bell Laboratories, it is not generally available."
The manuscript was received October 18, 1983.

Doug

Doug

[TUHS] BSD/v8 TCP/IP

[Larry McVoy]

I worked on the STREAMS based TCP/IP stack (at the time at Lachman
Associates but I think the code came from Convergent, not sure).
I ported it to the ETA-10 (weird machine) and also to SCO (yeah,
sorry if you have to use that mess).

I had quite a few conversations with dmr about it and he educated me
that there was STREAMS and there was streams.  His was the latter 
and it was like it was in the paper plus a few fixes.  The STREAMS
stuff was god awful.  I'm so glad that I've almost forgotten about
how awful.

On Mon, Sep 12, 2016 at 04:16:20PM -0400, Doug McIlroy wrote:
> Norman wrote
> 
> The earliest stream-I/O-system-based tty driver I'm aware of was
> already in the Research kernel when I interviewed at Bell Labs
> in early 1984.  I have a vague memory that it was a couple of
> years older than that, but I cannot find any citations to
> back up either of those memories.
> 
> Dennis described streams in the special Unix issue of the
> BSTJ, Oct 1984, and noted that "it runs on about 20 machines in
> ... Research ... Although it is being investigated in other
> parts of AT&T Bell Laboratories, it is not generally available."
> The manuscript was received October 18, 1983.
> 
> Doug
> 
> Doug

-- 
---
Larry McVoy            	     lm at mcvoy.com             http://www.mcvoy.com/lm 

[TUHS] BSD/v8 TCP/IP

[Lyndon Nerenberg]

> I worked on the STREAMS based TCP/IP stack (at the time at Lachman
> Associates but I think the code came from Convergent, not sure).

My memory of CTIX was that it was a native in-kernel BSD(-ish) 
implementation.  At least for the SVR2-derived CTIX kernels.  I don't ever 
recall a STREAMS-based implementation from Convergent, but I stopped using 
CTIX in the fall of 1988.

--lyndon

[TUHS] BSD/v8 TCP/IP

[Joerg Schilling]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 778 bytes --]

Doug McIlroy <doug at cs.dartmouth.edu> wrote:

> Dennis described streams in the special Unix issue of the
> BSTJ, Oct 1984, and noted that "it runs on about 20 machines in
> ... Research ... Although it is being investigated in other
> parts of AT&T Bell Laboratories, it is not generally available."
> The manuscript was received October 18, 1983.

Interesting, but then nobody did run a modern shell on one of these machines or 
everybody did type slowly, so the character lossage problem did not occur.


Jörg

-- 
 EMail:joerg at schily.net                  (home) Jörg Schilling D-13353 Berlin
       joerg.schilling at fokus.fraunhofer.de (work) Blog: http://schily.blogspot.com/
 URL:  http://cdrecord.org/private/ http://sourceforge.net/projects/schilytools/files/

[TUHS] BSD/v8 TCP/IP

[Larry McVoy]

On Tue, Sep 13, 2016 at 11:29:25AM +0200, Joerg Schilling wrote:
> Larry McVoy <lm at mcvoy.com> wrote:
> 
> > On Mon, Sep 12, 2016 at 06:52:02PM -0700, Lyndon Nerenberg wrote:
> > > --{ihnp4!alberta,pyramid,uunet,uwvax}!ncc!lyndon,auvax!lyndon,canada!lyndon
> >
> > Ahh, the memories.  I was 
> >
> > {ihnp4!alberta,pyramid,uunet,uwvax}!geophys!geowhiz!lm
> >
> > and then I was super stoked when I was 
> >
> > ...!uwvax!lm
> 
> In the 1980s I routed our whole companies (H. Berthold AG) mail traffic 
> through "sun" ;-), so you could reach me via sun!berthold!joerg
> 
> Well, we sold about 1/4 of all sun3 machines ever made, so Sun had an interest 
> to support us....

I'll bet.  By the late 1980's I left all this ! stuff behind and was
lm at sun.com, I loved having that email (and had to do some begging to
get it, everyone else was first.last at sun.com or something like that,
but I knew Guy Harris and he knew whoever ran sun.com so there ya go).

[TUHS] BSD/v8 TCP/IP

[Joerg Schilling]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 843 bytes --]

Larry McVoy <lm at mcvoy.com> wrote:

> On Mon, Sep 12, 2016 at 06:52:02PM -0700, Lyndon Nerenberg wrote:
> > --{ihnp4!alberta,pyramid,uunet,uwvax}!ncc!lyndon,auvax!lyndon,canada!lyndon
>
> Ahh, the memories.  I was 
>
> {ihnp4!alberta,pyramid,uunet,uwvax}!geophys!geowhiz!lm
>
> and then I was super stoked when I was 
>
> ...!uwvax!lm

In the 1980s I routed our whole companies (H. Berthold AG) mail traffic 
through "sun" ;-), so you could reach me via sun!berthold!joerg

Well, we sold about 1/4 of all sun3 machines ever made, so Sun had an interest 
to support us....

Jörg

-- 
 EMail:joerg at schily.net                  (home) Jörg Schilling D-13353 Berlin
       joerg.schilling at fokus.fraunhofer.de (work) Blog: http://schily.blogspot.com/
 URL:  http://cdrecord.org/private/ http://sourceforge.net/projects/schilytools/files/

[TUHS] BSD/v8 TCP/IP

[Joerg Schilling]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 950 bytes --]

Dan Cross <crossd at gmail.com> wrote:

> All modern shells use such an integrated history editor....
> >
>
> There is considerable difference on the meaning of "modern" with respect to
> this facility in recent shells, but this isn't the place for a holy war.

From what I could read in the net (I was not able to get an old ksh source)
older ksh versions did call "vi" as external program to edit the history. 

This is why there still is the idiosyncratic "fc" command in ksh. With this 
interface, you could not type ahead as you first need to call "fc" to be able 
to edit a command line. Without type aead, there is no character loss with the 
original AT&T streams concept.

Jörg

-- 
 EMail:joerg at schily.net                  (home) Jörg Schilling D-13353 Berlin
       joerg.schilling at fokus.fraunhofer.de (work) Blog: http://schily.blogspot.com/
 URL:  http://cdrecord.org/private/ http://sourceforge.net/projects/schilytools/files/

[TUHS] BSD/v8 TCP/IP

[Brantley Coile]

...!akgua!ganehd!bwc. 1983, the same week as "The Right Stuff" was released. In 1988 it was ..!research!dtscp1!bwc. 

Sent from my iPad

> On Sep 12, 2016, at 10:37 PM, Larry McVoy <lm at mcvoy.com> wrote:
> 
>> On Mon, Sep 12, 2016 at 06:52:02PM -0700, Lyndon Nerenberg wrote:
>> --{ihnp4!alberta,pyramid,uunet,uwvax}!ncc!lyndon,auvax!lyndon,canada!lyndon
> 
> Ahh, the memories.  I was 
> 
> {ihnp4!alberta,pyramid,uunet,uwvax}!geophys!geowhiz!lm
> 
> and then I was super stoked when I was 
> 
> ...!uwvax!lm
> 
> I think.  Might have been a uw-research or something like that in there.
> 
> Loved the Geophysics department but found my home at the Comp Sci department.
> 
> --lm

[TUHS] BSD/v8 TCP/IP

[Larry McVoy]

On Mon, Sep 12, 2016 at 06:52:02PM -0700, Lyndon Nerenberg wrote:
> --{ihnp4!alberta,pyramid,uunet,uwvax}!ncc!lyndon,auvax!lyndon,canada!lyndon

Ahh, the memories.  I was 

{ihnp4!alberta,pyramid,uunet,uwvax}!geophys!geowhiz!lm

and then I was super stoked when I was 

...!uwvax!lm

I think.  Might have been a uw-research or something like that in there.

Loved the Geophysics department but found my home at the Comp Sci department.

--lm

[TUHS] BSD/v8 TCP/IP

[Lyndon Nerenberg]

>> I cannot recollect, though, when dial-up uucp
> finally ceased.)

The final posts to comp.mail.maps went out in September, 2000.  Although by then the majority of UUCP traffic was on TCP port 540.

In Canada at least, cable modems started rolling out in late 1996 or early 1997.  I dropped my dialup links in the summer of 1997, so I would think by the end of 1998 the majority of UUCP was over TCP (if not moved to SMTP).

--{ihnp4!alberta,pyramid,uunet,uwvax}!ncc!lyndon,auvax!lyndon,canada!lyndon

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 801 bytes
Desc: Message signed with OpenPGP using GPGMail
URL: <http://minnie.tuhs.org/pipermail/tuhs/attachments/20160912/92348955/attachment.sig>

[TUHS] BSD/v8 TCP/IP

[Dan Cross]

On Mon, Sep 12, 2016 at 5:53 PM, Joerg Schilling <schily at schily.net> wrote:

> Doug McIlroy <doug at cs.dartmouth.edu> wrote:
>
> > > Interesting, but then nobody did run a modern shell on one of these
> machines or
> > > everybody did type slowly, so the character lossage problem did not
> occur.
> >
> > I'm afraid I don't get the point, apparently something about the
> > relative performance of stream- and non-stream tty drivers. How
> > do shells get into the act? And didn't uucp, which was certainly
> > not a slow typist, appear like any dial-up connection and thus
> > use /dev/ttyxx? (I cannot recollect, though, when dial-up uucp
> > finally ceased.)
>
> In 1982, I created a conceptional implementation and in 1984, I integrated
> a
> cursor editable history into my shell.
>
> As a result, this shell needed to switch the tty between raw and cooked
> mode.
> With the traditional UNIX tty driver, this was no problem, but with the
> unfixed
> AT&T strams based tty driver, this causes character loss.
>
> With such a shell, the conceptional bug in the original AT&T streams caused
> character loss when you type fast while the last command is going to
> terminate
> and the shell takes the input while switching the tty into raw mode.
>
> With the fix from Sun from around 1989, there is a new streams message that
> informs the lower side of the stream about how many characters re going to
> be
> read in raw mode. This permits to keep the other caracters in the edit
> buffer
> and avoids the character loss seen with the original AT&T streams driver
> concept.
>

AT&T STREAMS and research streams (note difference in case and specificity
of origin) were two separate things. v8 would have had the latter; you are
presumably referring to systems using the former. It is unsurprising that
bugs in the two would, in many cases, be disjoint: that is, the bug you are
referring to in AT&T STREAMS quite possibly wasn't in the research streams.

All modern shells use such an integrated history editor....
>

There is considerable difference on the meaning of "modern" with respect to
this facility in recent shells, but this isn't the place for a holy war.

        - Dan C.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://minnie.tuhs.org/pipermail/tuhs/attachments/20160912/605c1a0b/attachment.html>

[TUHS] BSD/v8 TCP/IP

[Joerg Schilling]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 1773 bytes --]

Doug McIlroy <doug at cs.dartmouth.edu> wrote:

> > Interesting, but then nobody did run a modern shell on one of these machines or
> > everybody did type slowly, so the character lossage problem did not occur.
>
> I'm afraid I don't get the point, apparently something about the
> relative performance of stream- and non-stream tty drivers. How
> do shells get into the act? And didn't uucp, which was certainly
> not a slow typist, appear like any dial-up connection and thus
> use /dev/ttyxx? (I cannot recollect, though, when dial-up uucp
> finally ceased.)

In 1982, I created a conceptional implementation and in 1984, I integrated a 
cursor editable history into my shell.

As a result, this shell needed to switch the tty between raw and cooked mode.
With the traditional UNIX tty driver, this was no problem, but with the unfixed
AT&T strams based tty driver, this causes character loss.

With such a shell, the conceptional bug in the original AT&T streams caused 
character loss when you type fast while the last command is going to terminate 
and the shell takes the input while switching the tty into raw mode.

With the fix from Sun from around 1989, there is a new streams message that 
informs the lower side of the stream about how many characters re going to be 
read in raw mode. This permits to keep the other caracters in the edit buffer 
and avoids the character loss seen with the original AT&T streams driver 
concept.

All modern shells use such an integrated history editor....

Jörg

-- 
 EMail:joerg at schily.net                  (home) Jörg Schilling D-13353 Berlin
       joerg.schilling at fokus.fraunhofer.de (work) Blog: http://schily.blogspot.com/
 URL:  http://cdrecord.org/private/ http://sourceforge.net/projects/schilytools/files/

[TUHS] BSD/v8 TCP/IP

[Dave Horsfall]

On Mon, 12 Sep 2016, Doug McIlroy wrote:

> (I cannot recollect, though, when dial-up uucp finally ceased.)

I'd not be surprised if in some obscure corner of the world it's still 
being used.

-- 
Dave Horsfall DTM (VK2KFU)  "Those who don't understand security will suffer."

[TUHS] BSD/v8 TCP/IP

[Doug McIlroy]

> Interesting, but then nobody did run a modern shell on one of these machines or
> everybody did type slowly, so the character lossage problem did not occur.

I'm afraid I don't get the point, apparently something about the
relative performance of stream- and non-stream tty drivers. How
do shells get into the act? And didn't uucp, which was certainly
not a slow typist, appear like any dial-up connection and thus
use /dev/ttyxx? (I cannot recollect, though, when dial-up uucp
finally ceased.)

DOug