Kick start on VT100 setup?

[Neil Higgins]

Although screen purports to emulate a VT100, and although RT11 V5 (boot script 3) starts with:

set tt scope

   and

set edit ked

... I am having no luck with either command line recall, or editing with ked.

I seem to recall that RT-11 V5 did have command line recall (with the uparrow). Doesn't work for me.

ked starts ok, but I can't get the keypad to do "stuff". I seem to remember that you could exit from ked by pressing the "gold" key plus an adjacent key (then maybe a command), but I have tried all the adjacent keys without success.

Same results whether coming in via VNC, vanilla SSH or Putty.

Are these functions working for anyone? Hints?

TIA,

Neil

P.S. Good-ol' edit seems to work fine, but I find I have become something of a full-screen-editing junky :-(

[Jerry Weiss]

On 8/20/18 5:56 AM, Neil Higgins wrote:

Although screen purports to emulate a VT100, and although RT11 V5 (boot script 3) starts with:

set tt scope

   and

set edit ked

... I am having no luck with either command line recall, or editing with ked.

.SET SL KED, ON

I seem to recall that RT-11 V5 did have command line recall (with the uparrow). Doesn't work for me.

ked starts ok, but I can't get the keypad to do "stuff". I seem to remember that you could exit from ked by pressing the "gold" key plus an adjacent key (then maybe a command), but I have tried all the adjacent keys without success.

GOLD 7  , then type "EXIT" or "QUIT".

GOLD HELP   for keypad layout

CONTROL-W   (or R) to refresh the screen.  This should reset the keypad as well.


Note: if your emulation screen size width is not even, KED may crash.  Depending on the emulator you have have to toggle "keypad" features.. Gold is also PF1,  Help is PF2.

Same results whether coming in via VNC, vanilla SSH or Putty.

Are these functions working for anyone? Hints?

see ftp://bitsavers.informatik.uni-stuttgart.de/pdf/dec/pdp11/rt11/v5.6_Aug91/AA-H853C-TC_PDP-11_Keypad_Editor_Users_Guide_Aug91.pdf

TIA,

Neil

P.S. Good-ol' edit seems to work fine, but I find I have become something of a full-screen-editing junky :-(

Good Luck,
Jerry

[Neil Higgins]

On Tuesday, 21 August 2018 01:28:53 UTC+10, Jerry Weiss wrote:

On 8/20/18 5:56 AM, Neil Higgins wrote:

Although screen purports to emulate a VT100, and although RT11 V5 (boot script 3) starts with:

set tt scope

   and

set edit ked

... I am having no luck with either command line recall, or editing with ked.

.SET SL KED, ON

The disk for script3 does not contain SL.SYS. I'll have to find an alternative distribution.

 

[Neil Higgins]

Progress (of a sort). rk5.dsk in the Nankervis collection contains SL.SYS; with that distro, command line recall works as per your instructions (using the "arrows" on the numeric keypad, provided that NumLock is off; although not the dedicated arrow keys). Two problems remain:

1. The arrow keys on the numeric keypad work in ked, but the PF keys do not: NumLock still sets num lock on and off, GOLD-7 keys me a 7 character on-screen. Have tried CTRL-W and CTRL-R - the screen repaints but the PF keys remain stubbornly non-functional).

2. The Nankervis (RT-11 V5.4) disk does not have FORTRAN on it, whereas the bootscript 3 disk (RT-11 V5.04) does. What are the chances of getting both by copying SL.SYS one way, or the FORTRAN files the other way (nil, I would expect)?

[Jerry Weiss]

On 8/20/18 8:47 PM, Neil Higgins wrote:

Progress (of a sort). rk5.dsk in the Nankervis collection contains SL.SYS; with that distro, command line recall works as per your instructions (using the "arrows" on the numeric keypad, provided that NumLock is off; although not the dedicated arrow keys). Two problems remain:

1. The arrow keys on the numeric keypad work in ked, but the PF keys do not: NumLock still sets num lock on and off, GOLD-7 keys me a 7 character on-screen. Have tried CTRL-W and CTRL-R - the screen repaints but the PF keys remain stubbornly non-functional).

Sorry, that should be GOLD KP7  (key pad 7).   If you haven't already, give ALT or similar keys a try with the PFn.  Every emulator seems to invent a new way to do this...

2. The Nankervis (RT-11 V5.4) disk does not have FORTRAN on it, whereas the bootscript 3 disk (RT-11 V5.04) does. What are the chances of getting both by copying SL.SYS one way, or the FORTRAN files the other way (nil, I would expect)?

Mark B has a nice utility that will manage the contents of a RT11 disk image using Python.  You should be able to use it move the files you want between images  given sufficient free space. See https://gitlab.com/NF6X_Retrocomputing/pyRT11


Jerry

[Neil Higgins]

Yes, keypad 7. But no dice, with or without extra keys. This might be a Linux thing. Never mind, I'll keep playing around.

[Neil Higgins]

<Solution_Preamble>

There's quite a bit of ancient history around this, when you dig. VTxxx emulation was a big thing in the days when PDP-11s and VAXes ruled the world. Since then, the support seems to have steadily eroded and gone flaky.

There's confusion (certainly in my mind, which favours comprehensibility to infinite configurability) around VTxxx emulation versus ANSI terminal emulation.

There is even even some confusion (again in my mind, if nobody else's) around what really constitutes VTxxx emulation. I can't shed any light on that, except to note that VTxxx capabilities evolved (as you would expect), somewhat in parallel with ANSI capabilities.

An old-timer needs to write a history paper on Smart Terminals, the rise and fall of DEC, and the rise and rise of anything but DEC.

With respect to NIX/NUX programs like gnome-terminal and XTerm, you see claims of "out-of-the-box" VTxxx emulation, which I cannot substantiate. Where is the (goddamn) -VT220 command line option? Without it, the rest is just layer upon layer of ugly, ugly detail.

Understanding these latter products is greatly complicated by the fizz and bubble and (Aussie accent) "loike, rooly roooly deep crap" around termcap, which is an old-time companion of EMACS, the virtues of writing editing macros in LISP, etc., etc., etc. (sorry - NOT a true believer).

Anyway ... here's one solution:

</Solution_Preamble>

The attached script, which I have tested only on an ancient Dell Dimension E521 running Ubuntu 16.04LTS, starts XTerm with VTxxx key bindings (don't ask me which xxx - I don't know - one that works with ked). Note that the host has to be specified with a command line option, e.g.

./vtxxx-term.sh -host pi@rpi303w

Some smart person could change it so that doesn't need "-host".

Remember to set executable permission on this script.

Then everything in ked appears to work.

Happy days.

Neil

[Neil Higgins]

I forgot to mention the "funny" bit:

One punter was SURE that the PF keys had stopped working after he installed Python 3. "Just remove python3, reinstall gnome-terminal and all will be well".

It turns out that removing python3 has the side effect of removing Firefox, gedit, aisleriot patience, and umpteen other, "minor" dependencies (I'm sure I'll find out what else in time) ...

... and it doesn't fix the keypad problem.

Sheesh.

[Johnny Billquist]

Things really are not that complicated. But it does require a bit of
understanding. Just reading some howto, and running some script is never
going to be a satisfactory solution. But I understand that this is what
most people are looking for.

Since these are two incompatible requirements, the confusion will
persist. (Understanding and quick fixes, that is.)

However, just to point some silly thing out, termcap is not an old-time
companion of Emacs. Unix is not really the natural environment to Emacs
to start with. Emacs comes from the PDP-10 world, and was written in
TECO, and not Lisp.
When PDP-10 declined, some people missed Emacs sorely, and wrote ports
for Unix. Since it was ported through reimplementation, TECO did not
come along, so a different implementation language was needed, and
preferably one that was easy to extend things in, and which preferably
didn't differentiate between data and code. As such, Lisp was not such a
surprising choice.

VT capabilities did expand over the years, but this was not in parallel
to, or in conflict with ANSI. The VT terminals tries to be fully ANSI
compliant, but there are a few things extra, because ANSI allows that,
and the VT100 series was actually created before the ANSI standard existed.

Anyway, feel free to ask questions surrounding xterm, and I'll happily
give all the silly answers. Note that pretty much all other terminal
emulators I've tried have (for me) annoying incompatibilities compared
to the terminals they try to emulate.
But in most cases, most terminal emulators are sortof "good enough". But
you do need to understand what you want, and what the different options
means.

Johnny

On 2018-08-21 09:11, Neil Higgins wrote:
> <Solution_Preamble>
>
> There's quite a bit of ancient history around this, when you dig. VTxxx
> emulation was a big thing in the days when PDP-11s and VAXes ruled the
> world. Since then, the support seems to have steadily eroded and gone flaky.
>
> There's confusion (certainly in my mind, which favours comprehensibility
> to infinite configurability) around VTxxx emulation versus ANSI terminal
> emulation.
>
> There is even even some confusion (again in my mind, if nobody else's)
> around what really constitutes VTxxx emulation. I can't shed any light
> on that, except to note that VTxxx capabilities evolved (as you would
> expect), somewhat in parallel with ANSI capabilities.
>
> An old-timer needs to write a history paper on Smart Terminals, the rise
> and fall of DEC, and the rise and rise of anything but DEC.
>
> With respect to NIX/NUX programs like gnome-terminal and XTerm, you see
> claims of "out-of-the-box" VTxxx emulation, which I cannot substantiate.
> Where is the (goddamn) -VT220 command line option? Without it, the rest
> is just layer upon layer of ugly, ugly detail.
>
> Understanding these latter products is greatly complicated by the fizz
> and bubble and (Aussie accent) "loike, rooly roooly deep crap" around
> termcap, which is an old-time companion of EMACS, the virtues of writing
> editing macros in LISP, etc., etc., etc. (sorry - NOT a true believer).
>
> Anyway ... here's one solution:
> </Solution_Preamble>
>

> The attached script, which I have tested /*only*/ on an ancient Dell

> Dimension E521 running Ubuntu 16.04LTS, starts XTerm with VTxxx key
> bindings (don't ask me which xxx - I don't know - one that works with
> ked). Note that the host has to be specified with a command line option,
> e.g.
>
> ./vtxxx-term.sh -host pi@rpi303w
>
> Some smart person could change it so that doesn't need "-host".
>
> Remember to set executable permission on this script.
>
> Then everything in ked appears to work.
>
> Happy days.
> Neil
>

> --
> You received this message because you are subscribed to the Google
> Groups "[PiDP-11]" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to pidp-11+u...@googlegroups.com
> <mailto:pidp-11+u...@googlegroups.com>.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/pidp-11/74c15ae6-222d-4069-88eb-d9d9371f5f84%40googlegroups.com
> <https://groups.google.com/d/msgid/pidp-11/74c15ae6-222d-4069-88eb-d9d9371f5f84%40googlegroups.com?utm_medium=email&utm_source=footer>.
> For more options, visit https://groups.google.com/d/optout.


--
Johnny Billquist || "I'm on a bus
|| on a psychedelic trip
email: b...@softjar.se || Reading murder books
pdp is alive! || tryin' to stay hip" - B. Idol

[Johnny Billquist]

Of course it didn't. Python have nothing to do with this.

Johnny

[Neil Higgins]

The key question is: Do you have a simpler/better solution?

I don’t want to start (another) LISP war, but since every text on LISP that I have ever (tried to) read starts with the premise that LISP is the “best” language, and any other language can be described only with pejorative terms, I am well and truly “over” it. SICP goes even further, suggesting that MIT is the “best” school and any other school can be described only with pejorative terms; it is interesting only for the content of the footnotes.

I could be converted by a good read about Smart Terminals, even though (let’s face it) they are pretty much a thing of the past.

[jdm]

Well, not *everything* but DEC kept rising. Maybe instead of DEC you mean “minicomputers” to which their demise in favor of personal computers and workstations was just evolution and natural selection.

Speaking of old guys and history, I just reread Soul of a New Machine, which came out in 1982, the same year I started as a Unix system developer at Sidereal in Portland, OR. I never used a Data General machine as I was always on one flavor of Unix or another, ATT or BSD, both there and Graphics Software Systems in Beaverton, until 1991 when I started consulting and had to deal with this dain-bramaged thing called Windows 3.0. Now, working on Spark clusters on Linux, my vi-ingrained fingers are happy and autonomous again. Quite remarkable that, despite our industry’s reputation for rapid change, we’re still using Unix skills nearly 40 years on.

--jdm

[Johnny Billquist]

On 2018-08-21 10:36, Neil Higgins wrote:
> The key question is: Do you have a simpler/better solution?

That depends on your view. As I said, this is not really that
complicated, but it does require that you understand things, and don't
just expect a quick howto or a script to run.

Sure, I could provide you with my settings and configurations, but if
you don't understand what I'm doing, it will still be just cryptic magic
to you, and not really any more helpful.

So, if you want to really understand things, I'm happy to talk. But then
we should start at the right end, and talk about what you want to do,
and when your issues are.

By the way, I sortof disagree with the terminology "smart terminal" for
a VT-series text terminal. I never heard anyone call them that back in
the day. A VT1200 would maybe have been called an "intelligent
terminal", but not a VT100. I think we might have said dumb terminals
about them all. Something with no cursor addressing or anything else
might have been called a "glass tty", if you wanted to be specific,
though. (ttys obviously never had cursor addressing or the like.)

> I don’t want to start (another) LISP war, but since every text on LISP that I have ever (tried to) read starts with the premise that LISP is the “best” language, and any other language can be described only with pejorative terms, I am well and truly “over” it. SICP goes even further, suggesting that MIT is the “best” school and any other school can be described only with pejorative terms; it is interesting only for the content of the footnotes.

I wasn't trying to claim any kind of superiority of LISP. I was pointing
out your erroneous assumption/claim that Emacs was written in LISP in
the first place. (As well as the claim of Emacs and termcap, which
obviously also is wrong, since Emacs don't come from Unix.)
But Emacs does stand for editing macros, and that you could dynamically
change the behavior at runtime in Emacs have always been fundamental,
and of course TECO (the language Emacs *was* implemented in), make no
difference between text and data, and that property is pretty much
needed. So when Emacs was being ported, you needed some language in
order to be able to extend things, and most people did not feel TECO was
a good language to start with (it is sortof "special"). So was mostly
just a question of which language to pick. Can you name some others that
were around in the early 80s?

> I could be converted by a good read about Smart Terminals, even though (let’s face it) they are pretty much a thing of the past.

That they are indeed, no matter what they might be.

Johnny

[jdm]

no, vt100 and friends were not considered smart, whatever you take “smart” to mean. i remember writing the color graphics driver for the vt102 in 1984, and boy was I was thrilled to see “bounce” (a simple line graphics test) running on it! remember that the de facto standard for graphics terminals in those days was the Tektronix 4014 - vector graphics. ie. you draw some number of vector primitives, including vector fonts for text, then *clear the screen* to draw something different. think electronic etch-a-sketch.

funny thing is, I think that vt102 was the only DEC VT I ever saw. even by the early 1980’s they were too bulky and expensive, and “clones” from Wyse et al popped up everywhere, both for VT100 and VT52.

the first terminal I’d consider “smart” or at least heading in that direction was the famous AT&T 5620 “blit” terminal - portrait orientation, with windowed terminal sessions and downloaded programs. I loved it, and its big, red, half-dome mouse (a radical thing). the 5620 had the first raster operations, aka “bit-block transfer” or bitblit for short. Rob Pike was the genious there. I was lucky enough to write the graphics for the 5630, its successor, which, thnakfully, had the glorious 68000 CPU, which was a joy to program the necessary self-modifying code needed to copy rectangular regions (eg windows, text, cursors, ...) to other locations, with any alignment and in 16 boolean “writing modes.” my code generated the most optimal 68k code to do that (on the stack) and jumped to it. remember, we had no GPU, no graphics hardware at all - not even for the cursor! (remember XOR cursors on bw graphics terminals?)

so, one could puckishly argue that assembly code fits your requirement for data-as-code. ;-) more seriously, I’d toss smalltalk into that ring, tho when emacs was rewritten, smalltalk-80 was even more esoteric than lisp. (for the record, i enjoyed lisp for doing AI, and really loved smalltalk.)

of course, speaking of smalltalk, the famous Xerox Star and later Alto were “smart” but IMHO not terminals at all but the first GUI-based workstations. my first company, Sidereal, had a couple of ex-Xerox PARC-ers and in 1982 we came out with a 68k-based system, complete with GUI, supporting up to 4 users at a time - two years ahead of the original Mac (which, of course, we found brilliant). no one ever heard of the Sidereal Mic-10 because it was targeted to the telecom vertical, not the then-nascent microcomputer market. (btw, if you’re wondering how we could provide a GUI for 4 people using a single 68k, it was a trick: the 6850 character generator had a mode where two consecutive characters could be made 2x in size. so, we designed icons as two “characters,” split down the middle. thus, we could efficiently create icons, window borders, etc. 4x the size of the screen font.)

ah, good times!! sadly, none of the “kids” I hire now have *any* experience with bare-metal - and it’s easier than ever now, with Arduino and Pi. man, if I had those things as a kid in the ‘70s....

--jdm

> --
> You received this message because you are subscribed to the Google Groups "[PiDP-11]" group.

> To unsubscribe from this group and stop receiving emails from it, send an email to pidp-11+u...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/pidp-11/357266f6-a9e1-d936-e8b5-dfa35e82671c%40softjar.se.

[Neil Higgins]

Hi Johnny,

I tend to use florid language (bordering on trolling, sorry) when I'm trying to stir up a little passion. All of this is in the best possible spirit ...

On Wednesday, 22 August 2018 08:02:49 UTC+10, Johnny Billquist wrote:

On 2018-08-21 10:36, Neil Higgins wrote:
> The key question is: Do you have a simpler/better solution?

That depends on your view ... ...I'm happy to talk. But then 

we should start at the right end, and talk about what you want to do,
and when your issues are.

I admit that I fall into the "life's too short" category. Have solution, will edit.

 

By the way, I sortof disagree with the terminology "smart terminal" for

a VT-series text terminal ...

... only Warren from the PiDP-8 Group objected to the term "dumb terminal" for the VTxxx. VTxxx/ANSI was certainly an advance on the glass TTY.

> I don’t want to start (another) LISP war, but since every text on LISP that I have ever (tried to) read starts with the premise that LISP is the “best” language, and any other language can be described only with pejorative terms, I am well and truly “over” it. SICP goes even further, suggesting that MIT is the “best” school and any other school can be described only with pejorative terms; it is interesting only for the content of the footnotes.

Definitely trolling there. However LOL has the arrogance to describe non-LISP languages as BLUB, That makes me angry.

I wasn't trying to claim any kind of superiority of LISP. I was pointing
out your erroneous assumption/claim that Emacs was written in LISP in

the first place ...

I wasn't claiming that EMACS is written in LISP, only that EMACS macros are written in a LISP variant, which I believe to be a true statement.

(As well as the claim of Emacs and termcap, which
obviously also is wrong, since Emacs don't come from Unix.)

Ok, but if I read NIX/NUX Emacs man files, there is a now, and undoubtedly has been for a long time, a dependency on TERMCAP or TERMINFO. To that extent, I have not diverged from the truth.

I understand the reason, just don't like the complete absence of sugar coating to appease Muggles like me.

 

But Emacs does stand for editing macros, and that you could dynamically
change the behavior at runtime in Emacs have always been fundamental,
and of course TECO (the language Emacs *was* implemented in), make no
difference between text and data, and that property is pretty much
needed. So when Emacs was being ported, you needed some language in
order to be able to extend things, and most people did not feel TECO was
a good language to start with (it is sortof "special"). So was mostly
just a question of which language to pick. Can you name some others that
were around in the early 80s?

Ummmm ... FORTH? Tee hee. Maybe not.

There is some excellent history on Emacs in the PiDP-8 Group. TECO was a phenomenon, a write-once language, now behind us except for the stories. Legend has it that people would submit their full name to TECO as a MACRO, just to see what it would do.

Best regards,

Neil

[Johnny Billquist]

On 2018-08-22 02:38, Neil Higgins wrote:
> Hi Johnny,
>
> I tend to use florid language (bordering on trolling, sorry) when I'm
> trying to stir up a little passion. All of this is in the best possible
> spirit ...

Fair enough. As long as you can take corrections without blowing a fuse,
I think we should be good.

> On Wednesday, 22 August 2018 08:02:49 UTC+10, Johnny Billquist wrote:
>
> On 2018-08-21 10:36, Neil Higgins wrote:
> > The key question is: Do you have a simpler/better solution?
>
> That depends on your view ... ...I'm happy to talk. But then
>
> we should start at the right end, and talk about what you want to do,
> and when your issues are.
>
>
> I admit that I fall into the "life's too short" category. Have solution,
> will edit.

Many do nowadays. And for me, that is the biggest problem. But I'm not
going to try and fix the world. :-)

> By the way, I sortof disagree with the terminology "smart terminal" for
> a VT-series text terminal ...
>
>
> ... only Warren from the PiDP-8 Group objected to the term "dumb
> terminal" for the VTxxx. VTxxx/ANSI was certainly an advance on the
> glass TTY.

Marginal. They got more advanced functions for doing the same stupid
things, and nothing more. Look up the definition of intelligent
terminals on Wikipedia, for example, and you'll see that for those, you
were expected to be able to do local processing of input, possibly load
programs into them that were run, and so on.

And from that point of view, a VT100 is no better than any other dumb
terminal. It's just a little more efficient, but it does not offer any
more advanced functionality.

> > I don’t want to start (another) LISP war, but since every text on
> LISP that I have ever (tried to) read starts with the premise that
> LISP is the “best” language, and any other language can be described
> only with pejorative terms, I am well and truly “over” it. SICP goes
> even further, suggesting that MIT is the “best” school and any other
> school can be described only with pejorative terms; it is
> interesting only for the content of the footnotes.
>
>
> Definitely trolling there. However LOL has the arrogance to describe
> non-LISP languages as BLUB, That makes me angry.

You need to relax. :-)

> I wasn't trying to claim any kind of superiority of LISP. I was
> pointing
> out your erroneous assumption/claim that Emacs was written in LISP in
> the first place ...
>
>
> I wasn't claiming that EMACS is written in LISP, only that EMACS macros
> are written in a LISP variant, which I believe to be a true statement.

No, it is not. That was my whole point. LISP have absolutely nothing to
do with EMACS. LISP only got in touch with EMACS when EMACS was
reimplemented the n:th time.
It was all TECO originally.

> (As well as the claim of Emacs and termcap, which
> obviously also is wrong, since Emacs don't come from Unix.)
>
>
> Ok, but if I read NIX/NUX Emacs man files, there is a now, and
> undoubtedly has been for a long time, a dependency on TERMCAP or
> TERMINFO. To that extent, I have not diverged from the truth.

Right. Any EMACS running under Unix will be making use of termcap or
terminfo.
It's just that the original EMACS didn't run under Unix.

> I understand the reason, just don't like the complete absence of sugar
> coating to appease Muggles like me.

:-)

> But Emacs does stand for editing macros, and that you could dynamically
> change the behavior at runtime in Emacs have always been fundamental,
> and of course TECO (the language Emacs *was* implemented in), make no
> difference between text and data, and that property is pretty much
> needed. So when Emacs was being ported, you needed some language in
> order to be able to extend things, and most people did not feel TECO
> was
> a good language to start with (it is sortof "special"). So was mostly
> just a question of which language to pick. Can you name some others
> that
> were around in the early 80s?
>
>
> Ummmm ... FORTH? Tee hee. Maybe not.

Forth is an interesting language. But I can't remember if you can take a
string, and actually execute it as if it was Forth code... Remember,
we're not talking about interpreted languages, but languages that makes
no difference between data and code.

> There is some excellent history on Emacs in the PiDP-8 Group. TECO was a
> phenomenon, a write-once language, now behind us except for the stories.
> Legend has it that people would submit their full name to TECO as a
> MACRO, just to see what it would do.

Well, TECO have rather been described as a write-only language.
And it's not behind all of us. I still get TECO out from time to time
when I want to do a bunch of weird editing on a file, where the power of
TECO really can help me.
And I wrote a small Emacs clone for TECO-8 many years ago, which I still
make use of when I'm running on a PDP-8.

By the way, NEIL would result in a search for the string "EIL" through
the whole file. Unfortunately maybe not so interesting. :-)

Johnny

>
> Best regards,

> Neil
>
> --
> You received this message because you are subscribed to the Google
> Groups "[PiDP-11]" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to pidp-11+u...@googlegroups.com
> <mailto:pidp-11+u...@googlegroups.com>.
> To view this discussion on the web visit

> https://groups.google.com/d/msgid/pidp-11/2f099c59-a993-44a0-990e-10f927cb844a%40googlegroups.com
> <https://groups.google.com/d/msgid/pidp-11/2f099c59-a993-44a0-990e-10f927cb844a%40googlegroups.com?utm_medium=email&utm_source=footer>.

> For more options, visit https://groups.google.com/d/optout.

[Warren Young]

On Tuesday, August 21, 2018 at 4:02:49 PM UTC-6, Johnny Billquist wrote:

By the way, I sortof disagree with the terminology "smart terminal" for
a VT-series text terminal. I never heard anyone call them that back in
the day.

Just as with animal intelligence, "smart" in technology is a spectrum, not a discrete quality. 

The claim wasn't that VTxxx terminals were called "smart terminals" back in the day, but that a VTxxx should not be confused with a "dumb terminal."

My definition of "dumb terminal" is that its only control characters are those defined by ASCII, so that it can only do what a Teletype could, differing only in the medium, ink on paper vs. glowing phosphors behind glass.

You can see this on a modern Linux or Unix box by setting your TERM type to "dumb", which causes any program based on ncurses or similar to suppress all terminal control commands. This is useful for getting plain text output from a program that normally tries to control your terminal. I've said something like "TERM=dumb ./annoying-program > output.txt" more than once in my career.

Here's a manual for an S100 board ca.1980 that let you turn a television into a text and graphics terminal, which uses the term "dumb terminal" in the way I've set out above:

    https://archive.org/details/bitsavers_scionMicroualApr81_3387823

I think we might have said dumb terminals about them all.

I also remember "dumb terminal" being used more broadly than I set out above, but usually disparagingly, such as in comparison to a PC or workstation. "Mess DOS has corrupted the directory again. I'll just have to enter it via the dumb terminal over there instead."

This ad supports your view:

   https://books.google.com/books?id=jT2fQqJplN8C&lpg=PT3&pg=PT3#v=onepage&q&f=false

Whether that's the source of the broader use of the term or just a recognition of it, I couldn't say.

If a terminal can support a full-screen text editor, I'm willing to call it a "smart" terminal. We can argue about how smart a given terminal is separately.

That makes the ADM-3A a "smart terminal" in my book, since that's what Bill Joy used when creating vi. Even with the 3A's minimal smarts, moving the cursor around programmatically saves a full-screen refresh, which was painful below about 9600 bps.

[Neil Higgins]

I tended to think of block mode terminals like the IBM 3270 (?) as being dumb, just dumb in a different way. They didn’t really “process” anything (did they? e.g. validating that a field intended to contain a number didn’t contain letters?). But then, if you rule that out, what constitutes “smart”? By the time we got to thick clients, they weren’t “terminals” any more - they were PCs or workstations(??) I’m really winging it now. So maybe Warren is right - anything beyond “glass TTY” was smart (-ish).

[Johnny Billquist]

On 2018-08-22 10:26, Neil Higgins wrote:
> I tended to think of block mode terminals like the IBM 3270 (?) as being dumb, just dumb in a different way. They didn’t really “process” anything (did they? e.g. validating that a field intended to contain a number didn’t contain letters?). But then, if you rule that out, what constitutes “smart”? By the time we got to thick clients, they weren’t “terminals” any more - they were PCs or workstations(??) I’m really winging it now. So maybe Warren is right - anything beyond “glass TTY” was smart (-ish).

I never used any IBM equipment, but my understanding is that they indeed
could validate input already in the terminal. You filled in the form,
and the terminal only accepted input for the different fields based on
information provided. So when the form was passed back to the computer,
you already knew it fulfilled various requirements.

It would have been rather bad (I think), if the terminal managed forms,
but you could have input letters in a field that were supposed to be a
number.

[Nick Sargeant]

3270E datastream was a bit smart - allowed the download of extended character sets, colours, highlighting, programmed symbols; supported operations like "read modified" so the terminal, working in block mode would only return fields where the user had changed them and so on. The first terminals were implemented in random logic, and later the controller to which the terminals were attached provided some intelligence for deconstructing the datastream from the host and reconstructing the responses. In IBM when we talked about dumb terminals we really meant ASCII teletypes, not 3270. 

[Neil Higgins]

My only exposure (and only as a user) was to terminals for a Fujitsu FACOM (an IBM knockoff). They were honkin’ big things - built like battleships. What you say about controllers makes sense - there was a coaxial cable from each terminal to a controller. My employer somehow managed to to get PCs on serial lines to act like FACOM block mode terminals. We put them on a stat multiplexor. I have no idea where the controller functionality ended up, unless it was somehow integrated into the PC as well.

[Andrew Yeomans]

As I remember it, there were dumb terminals, and there were really dumb terminals which did crazy things when sent cursor move characters at the screen edges. Causing software difficulties when implementing a "curses" style display. I think the ADM-3A was one of those, but I might be mistaking it for a cheap clone. I blame brain damage causing me to forgot the details after banging my head against the wall!

(As an aside, I'm surprised the Wikipedia article has not been made more politically correct. Those terminals were definitely "special needs".)

If you want to have fun, see if you can get my upside-down fonts to work, as mentioned in PiDP-8 group.

Andrew

[Nick Sargeant]

Modesty previously prevented my admission that I was one of the engineers on the IBM 3279 colour display terminal, so I may have a bit of a rosy view of the 3270E datastream. 

I went on to work on a number of follow-on products, most of which never saw the light of day, and many of which never got beyond simulation, but we variously looked at the capabilities of dumb terminals. Smooth scrolling was a feature of VT220 that we decided would be valuable, and we developed a solution that could 'smooth scroll' a partition on a multi-window text display. It wasn't until those mavericks down in Boca designed the PC, and forced us into the world of ASCII and chunky graphics, as well as supporting EBCDIC, and then integrating the ability to support 5250 protocols as well, as we fought to design a terminal that would homogenise all of IBMs display families that we gave up and went with microprocessor based bit-mapped display products - by then, DRAM was affordable enough. 

[jdm]

Doh! Bad memory bit: that was a VT-125, not VT-102. And by “graphics driver” I don’t mean the device firmware - not on the VT, anyway, but the driver for Graphics Software Systems “VDI” (virtual device interface) and “CGI” ( computer graphics interface), which enabled graphics apps to work on any type (display, printer, plotter) or model.

We had a whole zoo of devices and systems to port our graphics subsystems to and write graphics drivers for that particular hardware - pretty much any display, printer, plotter, mouse, tablet, and microcomputer system, whether CP/M, DOS, OS/2, and Unix. That was our stuff in IBM’s Professional Graphics Adapter card for the PC - a whopping 640x480x8 graphics. We did do device firmware for many OEMs, too, including the aforementioned AT&T 5630. The last big thing I did there was the first X Windows system for the PC, essentially making the PC a pretty decent X Windows display; the actual applications (xterm, etc) ran on Unix minis and communicated to the display over ethernet.

GSS. never had a PDP but had a VAX 750 and later Sequent (which was acquired by IBM).

Speaking of the IBM 3270, in 2000 GSS merged with Spectragraphics, whose sole product was a 3270 emulator for PCs. That was the beginning of the end, as both companies products were sunsetting.

--jdm

[Drew Rogge]

I tend to think of smart terminals of the day as being able to move the cursor to some row/column and if you were really lucky they could insert/delete lines and clear to end of line. I seem to remember that the ADM3 was a really dumb terminal where as the ADM3A had cursor addressing. For the life of me I can't remember what we used for smart terminals.

With regards to termcap I think that, along with the curses library,  was a Berkeley invention.

Drew

On 8/22/18 1:26 AM, Neil Higgins wrote:
> I tended to think of block mode terminals like the IBM 3270 (?) as being dumb, just dumb in a different way. They didn’t really “process” anything (did they? e.g. validating that a field intended to contain a number didn’t contain letters?). But then, if you rule that out, what constitutes “smart”? By the time we got to thick clients, they weren’t “terminals” any more - they were PCs or workstations(??) I’m really winging it now. So maybe Warren is right - anything beyond “glass TTY” was smart (-ish).
>

--
Drew Rogge
dr...@dasrogges.com

Phone: 8934OOO629OO4829631OOOOOOO

[Neil Higgins]

“Modesty previously prevented my admission that I was one of the engineers on the IBM 3279 colour display terminal, so I may have a bit of a rosy view of the 3270E datastream”

Ahem (blush!) ... well ... creating any kind of block mode terminal would obviously be harder than creating a glass TTY. My apologioes for the “dumb” epithet :-\

[Steve Hatle]

On Tuesday, August 21, 2018 at 6:51:33 PM UTC-5, John Miller wrote:

  no one ever heard of the Sidereal Mic-10 because it was targeted to the telecom vertical, not the then-nascent microcomputer market. 

I'd like to hear more about this! :-) Did a quick google and didn't find anything - too much astronomy results with "Sidereal" to wade through

Steve

[Ian Schofield]

Dear All,

 Got to say something being a 3270 lover (Ummm). This gem was seriously annoying as the keyboard locked until the concentrater acknowledged the character. And, you knew the mainframe (our 360) was down (crashed) when the lock persisted .. usually in the middle of an edit! The 'dumb terminal' bit is a bit of a grey area but IMHO, is best defined as a terminal that sent back stuff that the user had not typed .. even as trivial as the terminal ID on answer-back. EG VT05=dumb, VT52 and successors not dumb. In regard of cursor addressing, we are at risk of getting into the vertical tab discussion....

Regards, Ian.

On Monday, August 20, 2018 at 11:56:34 AM UTC+1, Neil Higgins wrote:

Although screen purports to emulate a VT100, and although RT11 V5 (boot script 3) starts with:

set tt scope

   and

set edit ked

... I am having no luck with either command line recall, or editing with ked.

I seem to recall that RT-11 V5 did have command line recall (with the uparrow). Doesn't work for me.

ked starts ok, but I can't get the keypad to do "stuff". I seem to remember that you could exit from ked by pressing the "gold" key plus an adjacent key (then maybe a command), but I have tried all the adjacent keys without success.

Same results whether coming in via VNC, vanilla SSH or Putty.

Are these functions working for anyone? Hints?

TIA,

Neil

P.S. Good-ol' edit seems to work fine, but I find I have become something of a full-screen-editing junky :-(

[Neil Higgins]

“The vertical tab discussion ...”. Dare I ask?

Your comments reminded me of the TTYs in our computer centre when the PDP-10 went down. Typing a character would produce a very distinctive “squoik” as the character was sent, followed by the absence of the usual “thwack” when the character was echoed. It was a “sad” noise.