MobaXterm - how to get the correct backspace?

1,272 views
Skip to first unread message

Neal G.

unread,
Oct 29, 2019, 12:07:57 AM10/29/19
to [PiDP-11]
I thought to give the recommended MobaXterm a try for connecting from Windows.
Sadly, the telnet session window to RSX-11M doesn't seem to work properly.
If I backspace over a mistyped command, then retype to correct it, that command will always fail.
For example,

$ dir *.ftn
DIRECTORY  --  Extraneous input - Check for proper command structure
DIR *.FTN
         ^
$

This problem does not happen when using telnet from Linux.
I imagine that there is a configuration table somewhere within MobaXterm, that can be used to fix this; does anyone know where to find it?

Thanks,
Neal G.

Jon Brase

unread,
Oct 29, 2019, 2:33:20 AM10/29/19
to [PiDP-11]
I think this section of the documentation for MobaXterm is what you're looking for:


Now, this just seems to be a checkbox with Ctrl-H being the character sent when checked and Ctrl-? being sent when unchecked. Under Mate-terminal on Linux there are a few other options.

Also, bear in mind that some systems of the era used other keys than the terminal backspace/delete key for backspacing, even printable characters. On 2.11 BSD, for example, the stty man page (see the * note below) lists "@" as the default erase character, "#" as the default line-kill character (erase the whole line you just typed), and Ctrl-? (often used in later times for backspace) as the interrupt character (now usually Ctrl-C). However, 2.11 BSD as configured on the image we use does not use those characters. I don't know if RSX did anything like that, or if it had any equivalent to the Unix stty command.

The problem is that in those days different terminal manufacturers would set up control keys with a given name (like "backspace" or "delete") to produce different control characters when they were pressed, and different operating systems would respond to control characters in different ways. Everything is well standardized now: Pretty much everybody is running either Windows or a Unix clone (and most of the Unix clone market is using Linux) on either x86 or ARM, and computers are mostly their own terminals, using a well standardized keyboard layout, and every platform that's not Windows or Unix knows that it has to interoperate with Windows and Unix. Back in the day, however, problems like what you're running into were common.


*While the default characters used for things like backspace, interrupt, and line-kill are well standardized nowadays, Unix has had the stty command from very early on, which, among other things, allows any user at any time to set any character they desire to be used for such things (at least when using applications that let the OS do their terminal handling, rather than doing it themselves), so that users with odd terminals could access the system. So, to this day, if you want, you can set some completely daft character (like "e") as your backspace character (of cour, tn you won't ab to ty " into your tminal, so don't do that!) <--- "Of course, then you won't be able to type "e"...".

Johnny Billquist

unread,
Oct 29, 2019, 3:42:38 AM10/29/19
to pid...@googlegroups.com
Yes. You need to configure it to sen DEL when deleting, and not
BACKSPACE. It's a rather common type of problem nowadays.

Johnny
> --
> 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-11/5096c561-b817-4e35-a027-787d8929bcc9%40googlegroups.com
> <https://groups.google.com/d/msgid/pidp-11/5096c561-b817-4e35-a027-787d8929bcc9%40googlegroups.com?utm_medium=email&utm_source=footer>.

--
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

Neal G.

unread,
Oct 30, 2019, 12:54:09 PM10/30/19
to [PiDP-11]
Thanks Jon for your historical persective.
I was a part of that era and have first-hand experience with a number of these old machines and operating systems.
Although the document doesn't provide any information not already displayed on the UI; changing the option to issue ASCII DEL (confusingly mislabeled ^? ) for BACKSPACE corrects the problem.

- Neal G.

Neal G.

unread,
Oct 30, 2019, 12:58:39 PM10/30/19
to [PiDP-11]

Thanks Johnny for the reminder. Yes, a common problem these days, especially for those of us who routinely bounce between Windows, MacOS, and various Linux distros: no terminal windows is the same.

On Tuesday, October 29, 2019 at 2:42:38 AM UTC-5, Johnny Billquist wrote:
Yes. You need to configure it to sen DEL when deleting, and not
BACKSPACE. It's a rather common type of problem nowadays.

...

Jon Brase

unread,
Oct 30, 2019, 1:50:57 PM10/30/19
to [PiDP-11]


On Wednesday, October 30, 2019 at 11:54:09 AM UTC-5, Neal G. wrote:
Thanks Jon for your historical persective.
I was a part of that era and have first-hand experience with a number of these old machines and operating systems.
Although the document doesn't provide any information not already displayed on the UI; changing the option to issue ASCII DEL (confusingly mislabeled ^? ) for BACKSPACE corrects the problem.

^? is standard caret notation for DEL. And if you think of the caret as flipping bit seven instead of zeroing it, it makes perfect sense.

Mike Katz

unread,
Oct 30, 2019, 3:06:25 PM10/30/19
to Jon Brase, [PiDP-11]
To expand on that, the "caret" notation usually indicates control, as in ^A (control-A).  The ASCII control characters occupy 0x00 to 0x1F of the basic (7 bit) ASCII codes.  The letter A is 0x41 and Control-A is 0x01.

From a binary standpoint the caret '^' usually means to bitwise and the ASCII code with 0x3F.  This allows both ^A and ^a to yield the same 0x01 result.

This applies for '^A'  (0x01) through '^_' (0x1f) and '^a' (0x01) through '^~' (0x1E).  With the addition of '^?' as an exception representing 0x7F instead of the expected 0x3F. 

DEL is ASCII code 0x7F (^?)
BS is ASCII code 0x08 (^H)

The use of the caret to indicate ASCII  control characters is by convection and not part of the ASCII (American Standard Code for Information Interchange) standard.

For additional information see:

https://en.wikipedia.org/wiki/Caret_notation

The caret notation was originally the up arrow character (↑) and was used to represent control characters on output devices that only could display/print the standard ASCII character set.  The dates back to the late 1950's or early 1960's.  The up arrow was replaced with the caret when the ASCII code was introduced in 1963.

And we return from our "way back" machine.
--
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/ec3df257-ab9d-4aa6-8d3b-b900fce86a41%40googlegroups.com.

Johnny Billquist

unread,
Oct 30, 2019, 4:17:35 PM10/30/19
to pid...@googlegroups.com
On 2019-10-30 17:58, Neal G. wrote:
>
> Thanks Johnny for the reminder. Yes, a common problem these days,
> especially for those of us who routinely bounce between Windows, MacOS,
> and various Linux distros: no terminal windows is the same.

IBM have a large part of the blame for this. Until they made the IBM PC,
BS was not so commonly used for the purpose of deleting characters.

Now it has pervaded everything, which is even sadder considering that in
any windowing system, you don't even use ASCII anyhow.

Johnny

>
> On Tuesday, October 29, 2019 at 2:42:38 AM UTC-5, Johnny Billquist wrote:
>
> Yes. You need to configure it to sen DEL when deleting, and not
> BACKSPACE. It's a rather common type of problem nowadays.
>
> ...
>
> --
> 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-11/23fc70b7-d2e9-4137-be25-c963465f1694%40googlegroups.com
> <https://groups.google.com/d/msgid/pidp-11/23fc70b7-d2e9-4137-be25-c963465f1694%40googlegroups.com?utm_medium=email&utm_source=footer>.

Johnny Billquist

unread,
Oct 30, 2019, 5:09:10 PM10/30/19
to pid...@googlegroups.com
Caret notation for control characters are a pretty standard convention.
And yes, ^? is commonly used to represent DEL. But all control
characters actually do have proper names in ASCII. The caret notion is
sometimes used in print to make control characters visible, and often
done by software.

So I wouldn't perhaps go as far as saying "mislabeled", but I think it
would be more appropriate in documentation or settings to say DEL. Just
as I would prefer things to say BS rather than ^H.

As for making perfect sense... Well, sortof. Control, as the key on a
keyboard, is doing more than just fiddling with bit six. (I assume we
are counting from 0.)
Caret could be seen as just another way of talking about Control.
But ^a and ^A would be expected to be the same thing... But it implies
more than just bit 6 being manipulated.

Also, when generating control characters, NUL is often generated by
control-space. But that don't follow the bit 6 pattern at all. I could
go on about this, but meh... :-)

Johnny

Mike Katz

unread,
Oct 30, 2019, 5:25:27 PM10/30/19
to Johnny Billquist, pid...@googlegroups.com
I always generated a null with control-@ on my ADM-3 (yes my age is
showing) or SWTPC CT-64.

Johnny Billquist

unread,
Oct 30, 2019, 5:56:28 PM10/30/19
to Mike Katz, pid...@googlegroups.com
On 2019-10-30 22:25, Mike Katz wrote:
> I always generated a null with control-@ on my ADM-3 (yes my age is
> showing) or SWTPC CT-64.

That's also an option, when you have an easily accessible @. Depends on
keyboard. But I don't think I've ever had a terminal on which C-SPC
didn't also generate a NUL.

Other common things:
^2 for NUL
^3 for ESC
^4 for FS
^5 for GS
^6 for RS
^7 for US
^8 for DEL

Maybe not all, and not everywhere, but they seem pretty common in my
experience.

(There is some kind of logic behind these as well, even if it might seem
a little far fetched for some...)

Johnny

Jon Brase

unread,
Oct 30, 2019, 7:48:26 PM10/30/19
to [PiDP-11]
On Wednesday, October 30, 2019 at 4:56:28 PM UTC-5, Johnny Billquist wrote:
On 2019-10-30 22:25, Mike Katz wrote:
> I always generated a null with control-@ on my ADM-3 (yes my age is
> showing) or SWTPC CT-64.

That's also an option, when you have an easily accessible @. Depends on
keyboard. But I don't think I've ever had a terminal on which C-SPC
didn't also generate a NUL.

Other common things:
^2 for NUL

Well, on US keyboards, Shift-2 is @, and control characters are generally keyed as Ctrl + key without regard to shift state and displayed in caret notation as ^ + whatever the character the key emits when shifted (e.g, Ctrl-h -> ^H), with a few exceptions, so Ctrl-2 giving the same character as Ctrl-@ isn't surprising at all.
 
^3 for ESC
^4 for FS
^5 for GS
^6 for RS
^7 for US
^8 for DEL

These, however, are very surprising. 

Jon Brase

unread,
Oct 30, 2019, 8:47:22 PM10/30/19
to [PiDP-11]


On Wednesday, October 30, 2019 at 4:09:10 PM UTC-5, Johnny Billquist wrote:

Caret notation for control characters are a pretty standard convention.
And yes, ^? is commonly used to represent DEL. But all control
characters actually do have proper names in ASCII. The caret notion is
sometimes used in print to make control characters visible, and often
done by software.

So I wouldn't perhaps go as far as saying "mislabeled", but I think it
would be more appropriate in documentation or settings to say DEL. Just
as I would prefer things to say BS rather than ^H.

I'll even fairly often say "Delete" and "Backspace" to indicate the character names. But my point was that, when using caret notation to notate characters, ^? is a perfectly consistent notation for Delete.
 

As for making perfect sense... Well, sortof. Control, as the key on a
keyboard, is doing more than just fiddling with bit six. (I assume we
are counting from 0.)
Caret could be seen as just another way of talking about Control.
But ^a and ^A would be expected to be the same thing... But it implies
more than just bit 6 being manipulated.

Yeah, the notation tends to assume that you're tying bit 6 high and then flipping bit 7. If no manipulation of bit 6 were being done, ^a would be ! and ^A would be SOH.

Johnny Billquist

unread,
Oct 31, 2019, 3:58:09 AM10/31/19
to pid...@googlegroups.com
On 2019-10-31 00:48, Jon Brase wrote:
> On Wednesday, October 30, 2019 at 4:56:28 PM UTC-5, Johnny Billquist wrote:
>
> On 2019-10-30 22:25, Mike Katz wrote:
> > I always generated a null with control-@ on my ADM-3 (yes my age is
> > showing) or SWTPC CT-64.
>
> That's also an option, when you have an easily accessible @. Depends on
> keyboard. But I don't think I've ever had a terminal on which C-SPC
> didn't also generate a NUL.
>
> Other common things:
> ^2 for NUL
>
>
> Well, on US keyboards, Shift-2 is @, and control characters are
> generally keyed as Ctrl + key without regard to shift state and
> displayed in caret notation as ^ + whatever the character the key emits
> when shifted (e.g, Ctrl-h -> ^H), with a few exceptions, so Ctrl-2
> giving the same character as Ctrl-@ isn't surprising at all.

Like I said, there is some logic. It's a bit funny, but not too much.

> ^3 for ESC
> ^4 for FS
> ^5 for GS
> ^6 for RS
> ^7 for US
> ^8 for DEL
>
>
> These, however, are very surprising.

Think octal...

ESC is 033
FS is 034
GS is 035
RS is 036
US is 037

and then there was only DEL missing, so that one came next.

Johnny Billquist

unread,
Oct 31, 2019, 4:01:19 AM10/31/19
to pid...@googlegroups.com
On 2019-10-31 01:47, Jon Brase wrote:
>
>
> On Wednesday, October 30, 2019 at 4:09:10 PM UTC-5, Johnny Billquist wrote:
>
>
> Caret notation for control characters are a pretty standard convention.
> And yes, ^? is commonly used to represent DEL. But all control
> characters actually do have proper names in ASCII. The caret notion is
> sometimes used in print to make control characters visible, and often
> done by software.
>
> So I wouldn't perhaps go as far as saying "mislabeled", but I think it
> would be more appropriate in documentation or settings to say DEL. Just
> as I would prefer things to say BS rather than ^H.
>
>
> I'll even fairly often say "Delete" and "Backspace" to indicate the
> character names. But my point was that, when using caret notation to
> notate characters, ^? is a perfectly consistent notation for Delete.

I think ^? is a little "stranger" than all the others. But if you want
any kind of caret notation for DEL, then why not. Also, you could look
at it as it's not any known character that morphs into DEL, and then '?'
could be viewed as the unknown character. :-)

> As for making perfect sense... Well, sortof. Control, as the key on a
> keyboard, is doing more than just fiddling with bit six. (I assume we
> are counting from 0.)
> Caret could be seen as just another way of talking about Control.
> But ^a and ^A would be expected to be the same thing... But it implies
> more than just bit 6 being manipulated.
>
>
> Yeah, the notation tends to assume that you're tying bit 6 high and then
> flipping bit 7. If no manipulation of bit 6 were being done, ^a would be
> ! and ^A would be SOH.

Bit 5 and 6... Not 6 and 7. :-)

Jon Brase

unread,
Oct 31, 2019, 4:34:19 AM10/31/19
to [PiDP-11]


On Thursday, October 31, 2019 at 3:01:19 AM UTC-5, Johnny Billquist wrote:
On 2019-10-31 01:47, Jon Brase wrote:
>
>
> On Wednesday, October 30, 2019 at 4:09:10 PM UTC-5, Johnny Billquist wrote:
>
>
>     Caret notation for control characters are a pretty standard convention.
>     And yes, ^? is commonly used to represent DEL. But all control
>     characters actually do have proper names in ASCII. The caret notion is
>     sometimes used in print to make control characters visible, and often
>     done by software.
>
>     So I wouldn't perhaps go as far as saying "mislabeled", but I think it
>     would be more appropriate in documentation or settings to say DEL. Just
>     as I would prefer things to say BS rather than ^H.
>
>
> I'll even fairly often say "Delete" and "Backspace" to indicate the
> character names. But my point was that, when using caret notation to
> notate characters, ^? is a perfectly consistent notation for Delete.

I think ^? is a little "stranger" than all the others.

Of course it's strange. DEL is the eccentric loner among the control characters, hanging out in some seedy tavern in the furthest corner of the printable range, and keeping company with uncouth rabble like } and ~, instead of being part of NUL's entourage like all of the respectable control characters. 
 
But if you want
any kind of caret notation for DEL, then why not. Also, you could look
at it as it's not any known character that morphs into DEL, and then '?'
could be viewed as the unknown character. :-)

>     As for making perfect sense... Well, sortof. Control, as the key on a
>     keyboard, is doing more than just fiddling with bit six. (I assume we
>     are counting from 0.)
>     Caret could be seen as just another way of talking about Control.
>     But ^a and ^A would be expected to be the same thing... But it implies
>     more than just bit 6 being manipulated.
>
>
> Yeah, the notation tends to assume that you're tying bit 6 high and then
> flipping bit 7. If no manipulation of bit 6 were being done, ^a would be
> ! and ^A would be SOH.

Bit 5 and 6... Not 6 and 7. :-)

Silly me. Next I'll be saying that the two possible values of a bit are 1 and 2.

Tony Nicholson

unread,
Oct 31, 2019, 6:17:15 PM10/31/19
to [PiDP-11]

On Thursday, October 31, 2019 at 7:34:19 PM UTC+11, Jon Brase wrote:

[snip]
 
Of course it's strange. DEL is the eccentric loner among the control characters, hanging out in some seedy tavern in the furthest corner of the printable range, and keeping company with uncouth rabble like } and ~, instead of being part of NUL's entourage like all of the respectable control characters. 
 

The origins of the DEL character go back to the ASR-33 Teletype - where one would prepare paper tapes off-line.   If you made a mistake typing and punched an incorrect character you could backspace the tape and overstrike a DEL character by pressing the RUB OUT key (punch out all 7-bits of the mistaken character).  Software reading the paper tapes would ignore the DEL character.

There are more historical details on the Wikipedia ASCII page at https://en.wikipedia.org/wiki/ASCII

Tony


Reply all
Reply to author
Forward
0 new messages