Small request for the new cygwin terminal

Small request for the new cygwin terminal

[Lluís Batlle i Rossell]

Hello,

thank you first for such a helpful software, for those that use Windows but are
used to the Linux interface.

I'd like to request some heavy annoyance I encounter with the new cygwin
terminal, compared to rxvt. In 'vim', it looks like I'm used to press ESC
*before* releasing *shift* (if I was inserting some text innvolving shift).

On xterm, rxvt, windows terminal, if I pressed "shift-esc" in vim, it had the
same effect as esc. But in the cygwin terminal, in insert mode it inserts some
weird character.

Could it be made so shift-esc has the same effect as esc, in cygwin terminal? I
imagine that the difference is not done on purpose.

Thank you,
Lluís.

--
Problem reports:       http://cygwin.com/problems.html
FAQ:                   http://cygwin.com/faq/
Documentation:         http://cygwin.com/docs.html
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple

Re: Small request for the new cygwin terminal

[Andy Koppe]

On 7 June 2012 09:40, Lluís Batlle i Rossell wrote:
> Hello,
>
> thank you first for such a helpful software, for those that use Windows but are
> used to the Linux interface.
>
> I'd like to request some heavy annoyance I encounter with the new cygwin
> terminal, compared to rxvt. In 'vim', it looks like I'm used to press ESC
> *before* releasing *shift* (if I was inserting some text innvolving shift).
>
> On xterm, rxvt, windows terminal, if I pressed "shift-esc" in vim, it had the
> same effect as esc. But in the cygwin terminal, in insert mode it inserts some
> weird character.
>
> Could it be made so shift-esc has the same effect as esc, in cygwin terminal? I
> imagine that the difference is not done on purpose.

It is done on purpose, to allow Shift+Esc to be bound to a different
function if desired. It's documented here:

[1] http://code.google.com/p/mintty/wiki/Keycodes#Special_keys

Shift+ESC sends the Unicode codepoint U+009B, aka the Control Sequence
Introducer (CSI) character.

In vim, you could map it to Esc like this:

map <CSI> <Esc>
map! <CSI> <Esc>

Andy

--
Problem reports:       http://cygwin.com/problems.html
FAQ:                   http://cygwin.com/faq/
Documentation:         http://cygwin.com/docs.html
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple

RE: Small request for the new cygwin terminal

[Nellis, Kenneth]

-----Original Message-----
From: Andy Koppe
<snip/>
It is done on purpose, to allow Shift+Esc to be bound to a different
function if desired. It's documented here:

[1] http://code.google.com/p/mintty/wiki/Keycodes#Special_keys

Shift+ESC sends the Unicode codepoint U+009B, aka the Control Sequence
Introducer (CSI) character.
<snip/>
-----END Original Message-----

My experience does not agree with the "Special keys" table at the
page referenced by that URL.

When pressing the unmodified Pause/Break key, I get ^], which maps 
to Pause in your table, and pressing Ctrl+Pause/Break, I get ^\, 
which maps to Break in your table, so this agrees with the statement, 
"Pause and Break usually share a key, whereby Ctrl has to be pressed 
to get the Break function."

But, when I press Shift+Pause/Break, because I'm not pressing Ctrl, 
I should get your table's Pause/Shift result, U+009C, but I get 9D. 
Similarly, when I press Ctrl+Shift+Pause/Break, because I am pressing 
Ctrl, I should get your table's Break/Shift result, U+009D, but I get 9C.

Also, when I press either Ctrl+Tab or Ctrl+Shift+Tab, instead of 
getting what the table indicates, I don't get anything at all.

--Ken Nellis

Re: Small request for the new cygwin terminal

[Helmut Karlowski]

Nellis, Kenneth, 07.06.2012 16:56:57:

> -----Original Message-----
> From: Andy Koppe
> <snip/>
> It is done on purpose, to allow Shift+Esc to be bound to a different
> function if desired. It's documented here:
>
> [1] http://code.google.com/p/mintty/wiki/Keycodes#Special_keys
>
> Shift+ESC sends the Unicode codepoint U+009B, aka the Control Sequence
> Introducer (CSI) character.
> <snip/>
> -----END Original Message-----
>
> My experience does not agree with the "Special keys" table at the
> page referenced by that URL.

On my german keyboard I receive ^[^[ (2x ESC) for Shift-ESC. Character Set  
is CP850 if that matters. As of my understanding CSI is something like  
ESC[ (at least on an ANSI-terminal, is this different on xterm?).

> Also, when I press either Ctrl+Tab or Ctrl+Shift+Tab, instead of
> getting what the table indicates, I don't get anything at all.

Same here.

-- 
Helmut Karlowski

--
Problem reports:       http://cygwin.com/problems.html
FAQ:                   http://cygwin.com/faq/
Documentation:         http://cygwin.com/docs.html
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple

Re: Small request for the new cygwin terminal

[Helmut Karlowski]

Helmut Karlowski, 07.06.2012 17:52:43:

> Nellis, Kenneth, 07.06.2012 16:56:57:
>
>> -----Original Message-----
>> From: Andy Koppe
>> <snip/>
>> It is done on purpose, to allow Shift+Esc to be bound to a different
>> function if desired. It's documented here:
>>
>> [1] http://code.google.com/p/mintty/wiki/Keycodes#Special_keys
>>
>> Shift+ESC sends the Unicode codepoint U+009B, aka the Control Sequence
>> Introducer (CSI) character.
>> <snip/>
>> -----END Original Message-----
>>
>> My experience does not agree with the "Special keys" table at the
>> page referenced by that URL.
>
> On my german keyboard I receive ^[^[ (2x ESC) for Shift-ESC. Character  
> Set is CP850 if that matters. As of my understanding CSI is something  
> like ESC[ (at least on an ANSI-terminal, is this different on xterm?).

Switching to ISO-8859-1 emits 0x9b which seems to be correct.

-- 
Helmut Karlowski

--
Problem reports:       http://cygwin.com/problems.html
FAQ:                   http://cygwin.com/faq/
Documentation:         http://cygwin.com/docs.html
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple

Re: Small request for the new cygwin terminal

[Andy Koppe]

On 7 June 2012 15:56, Nellis, Kenneth wrote:
> -----Original Message-----
> From: Andy Koppe
> <snip/>
> It is done on purpose, to allow Shift+Esc to be bound to a different
> function if desired. It's documented here:
>
> [1] http://code.google.com/p/mintty/wiki/Keycodes#Special_keys
>
> Shift+ESC sends the Unicode codepoint U+009B, aka the Control Sequence
> Introducer (CSI) character.
> <snip/>
> -----END Original Message-----
>
> My experience does not agree with the "Special keys" table at the
> page referenced by that URL.
>
> When pressing the unmodified Pause/Break key, I get ^], which maps
> to Pause in your table, and pressing Ctrl+Pause/Break, I get ^\,
> which maps to Break in your table, so this agrees with the statement,
> "Pause and Break usually share a key, whereby Ctrl has to be pressed
> to get the Break function."
>
> But, when I press Shift+Pause/Break, because I'm not pressing Ctrl,
> I should get your table's Pause/Shift result, U+009C, but I get 9D.
> Similarly, when I press Ctrl+Shift+Pause/Break, because I am pressing
> Ctrl, I should get your table's Break/Shift result, U+009D, but I get 9C.

Oops, documentation error. Now fixed. Thanks very much for testing that.

The intention is that the Shift combination sends the basic control
character plus 0x80. See also the preceding section (which I've now
added a link to from the Special Keys section):

http://code.google.com/p/mintty/wiki/Keycodes#Ctrl

> Also, when I press either Ctrl+Tab or Ctrl+Shift+Tab, instead of
> getting what the table indicates, I don't get anything at all.

You've probably got the shortcuts for switching among mintty windows
enabled on the Keys page of the options.

Andy

--
Problem reports:       http://cygwin.com/problems.html
FAQ:                   http://cygwin.com/faq/
Documentation:         http://cygwin.com/docs.html
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple

Re: Small request for the new cygwin terminal

[Andy Koppe]

On 7 June 2012 17:10, Helmut Karlowski wrote:
> Helmut Karlowski, 07.06.2012 17:52:43:
>
>
>> Nellis, Kenneth, 07.06.2012 16:56:57:
>>
>>> -----Original Message-----
>>> From: Andy Koppe
>>> <snip/>
>>> It is done on purpose, to allow Shift+Esc to be bound to a different
>>> function if desired. It's documented here:
>>>
>>> [1] http://code.google.com/p/mintty/wiki/Keycodes#Special_keys
>>>
>>> Shift+ESC sends the Unicode codepoint U+009B, aka the Control Sequence
>>> Introducer (CSI) character.
>>> <snip/>
>>> -----END Original Message-----
>>>
>>> My experience does not agree with the "Special keys" table at the
>>> page referenced by that URL.
>>
>>
>> On my german keyboard I receive ^[^[ (2x ESC) for Shift-ESC. Character Set
>> is CP850 if that matters. As of my understanding CSI is something like ESC[
>> (at least on an ANSI-terminal, is this different on xterm?).
>
>
> Switching to ISO-8859-1 emits 0x9b which seems to be correct.

See the preceding section on the wiki page:

http://code.google.com/p/mintty/wiki/Keycodes#Ctrl

"If Shift is held in addition to a control character combination, the
corresponding character from the so-called C1 control character set is
sent. The C1 control characters are shown as Unicode codepoints in the
table below. How exactly C1 control characters are sent depends on the
selected character encoding. In ISO-8859 encodings, they are encoded
as single bytes, e.g. \x81 for U+0081. With UTF-8, they are encoded as
two-byte sequences, which effectively means that the character code is
prefixed with a \xC2 byte, so for example U+0081 becomes \xC2\x81. C1
codepoints that are not supported by the selected character encoding
are sent by prefixing the corresponding ASCII control character with
an ESC character, so U+0081 would be sent as ^[^A."

CP850 doesn't support the C1 control characters, so you get the
fallback instead, which is the scheme for encoding the Meta modifier
when bit 7 isn't available. This means that Ctrl+Shift+letter
combinations can't be distinguished from Ctrl+Alt+letter, but since
Ctrl+Alt+letter combinations are suppposedly reserved for global
shortcuts in Windows, they might still come in handy occasionally.

There indeed is another scheme of encoding C1 control characters with
an ESC followed by the base character, e.g. ^[A instead of ^[^A, but
that would have made the Ctrl+Shift+letter keycode the same as
Alt+letter rather than Ctrl+Alt+letter, which is why I chose the Meta
scheme.

I hope that makes some sense.

Andy

--
Problem reports:       http://cygwin.com/problems.html
FAQ:                   http://cygwin.com/faq/
Documentation:         http://cygwin.com/docs.html
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple

Re: Small request for the new cygwin terminal

[Helmut Karlowski]

Andy Koppe, 07.06.2012 21:12:05:

> CP850 doesn't support the C1 control characters, so you get the
> fallback instead, which is the scheme for encoding the Meta modifier
> when bit 7 isn't available. This means that Ctrl+Shift+letter
> combinations can't be distinguished from Ctrl+Alt+letter, but since
> Ctrl+Alt+letter combinations are suppposedly reserved for global
> shortcuts in Windows, they might still come in handy occasionally.
>
> There indeed is another scheme of encoding C1 control characters with
> an ESC followed by the base character, e.g. ^[A instead of ^[^A, but
> that would have made the Ctrl+Shift+letter keycode the same as
> Alt+letter rather than Ctrl+Alt+letter, which is why I chose the Meta
> scheme.
>
> I hope that makes some sense.

Yes. One question remains: Does the Character Set (ISO, CP, etc.) have an  
effect on the terminfo-terminal-description? I've added

csin=\233,

but I'm not sure if it's really necessary, and if it's wrong when I switch  
the Character Set. tgetstr() returns the value defined in xterm-terminfo,  
which may be wrong for another character set. When csin is missing in  
xterm, tgetstr() returns 0. I'm not sure if this can be solved inside  
mintty, is there a fallback if an entry in terminfo is missing?

-- 
Helmut Karlowski

--
Problem reports:       http://cygwin.com/problems.html
FAQ:                   http://cygwin.com/faq/
Documentation:         http://cygwin.com/docs.html
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple

Re: Small request for the new cygwin terminal

[Andy Koppe]

On 7 June 2012 21:14, Helmut Karlowski wrote:
> Andy Koppe, 07.06.2012 21:12:05:
>
>
>> CP850 doesn't support the C1 control characters, so you get the
>> fallback instead, which is the scheme for encoding the Meta modifier
>> when bit 7 isn't available. This means that Ctrl+Shift+letter
>> combinations can't be distinguished from Ctrl+Alt+letter, but since
>> Ctrl+Alt+letter combinations are suppposedly reserved for global
>> shortcuts in Windows, they might still come in handy occasionally.
>>
>> There indeed is another scheme of encoding C1 control characters with
>> an ESC followed by the base character, e.g. ^[A instead of ^[^A, but
>> that would have made the Ctrl+Shift+letter keycode the same as
>> Alt+letter rather than Ctrl+Alt+letter, which is why I chose the Meta
>> scheme.
>>
>> I hope that makes some sense.
>
>
> Yes. One question remains: Does the Character Set (ISO, CP, etc.) have an
> effect on the terminfo-terminal-description?

I don't think you can have different terminfo entries for different
character sets for the same terminal type, i.e. you'd need to create
charset-specifc types such as "xterm-cp850".

> I've added
>
> csin=\233,
>
> but I'm not sure if it's really necessary, and if it's wrong when I switch
> the Character Set. tgetstr() returns the value defined in xterm-terminfo,
> which may be wrong for another character set. When csin is missing in xterm,
> tgetstr() returns 0.

Looking at the standard, 'csin' isn't meant for a keycode, but for a
command sent to the terminal, namely for the "Init sequence for
multiple codesets". I'm afraid I don't know what that means though,
nor why the xterm terminfo for it is empty.

> I'm not sure if this can be solved inside mintty, is
> there a fallback if an entry in terminfo is missing?

There isn't a fallback in the terminfo system itself, so I guess
that's up to each terminfo-using program (or its configuration).

Andy

--
Problem reports:       http://cygwin.com/problems.html
FAQ:                   http://cygwin.com/faq/
Documentation:         http://cygwin.com/docs.html
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple

RE: Small request for the new cygwin terminal

[Nellis, Kenneth]

-----Original Message-----
From: Andy Koppe
Sent: Thursday, June 07, 2012 2:45 PM

<snip/>
Oops, documentation error. Now fixed. Thanks very much for testing that.

The intention is that the Shift combination sends the basic control
character plus 0x80. See also the preceding section (which I've now
added a link to from the Special Keys section):

http://code.google.com/p/mintty/wiki/Keycodes#Ctrl

> Also, when I press either Ctrl+Tab or Ctrl+Shift+Tab, instead of
> getting what the table indicates, I don't get anything at all.

You've probably got the shortcuts for switching among mintty windows
enabled on the Keys page of the options.

-----END Original Message-----

Appreciate the doc fix and, indeed, your shortcut explanation
applied to me.

Regarding, the "@" key in the table referenced by your URL,
how does one enter Ctrl+Shift+@, when to get "@" you need to
already be holding down the Shift key. This maybe sounds like
a ridiculous question, but regardless, when I press the 2/@ 
key while holding down Ctrl+Shift, I get ^@, not U+0080, as
indicated by the table. 

Additionally, Ctrl+Shift+(6/^) gives me ^^, not U+009E;
Ctrl+(-/_) gives ^[ [ 1 ; 5 m, not ^_;
Ctrl+Shift+(-/_) gives ^_, not U+009F;
Ctrl+Shift+/ gives ^?, not U+009F.

--Ken Nellis

Re: Small request for the new cygwin terminal

[Andy Koppe]

On 8 June 2012 19:02, Nellis, Kenneth wrote:
> -----Original Message-----
> From: Andy Koppe
> Sent: Thursday, June 07, 2012 2:45 PM
>
> <snip/>
> Oops, documentation error. Now fixed. Thanks very much for testing that.
>
> The intention is that the Shift combination sends the basic control
> character plus 0x80. See also the preceding section (which I've now
> added a link to from the Special Keys section):
>
> http://code.google.com/p/mintty/wiki/Keycodes#Ctrl
>
>> Also, when I press either Ctrl+Tab or Ctrl+Shift+Tab, instead of
>> getting what the table indicates, I don't get anything at all.
>
> You've probably got the shortcuts for switching among mintty windows
> enabled on the Keys page of the options.
>
> -----END Original Message-----
>
> Appreciate the doc fix and, indeed, your shortcut explanation
> applied to me.
>
> Regarding, the "@" key in the table referenced by your URL,
> how does one enter Ctrl+Shift+@, when to get "@" you need to
> already be holding down the Shift key.

Hold down both Shift keys. Yes, really.

(For the record, I wouldn't expect anyone to really use that, and I'm
a bit surprised that I bothered to implement it. Remember though that
different keyboard layouts have the relevant characters in different
positions.)

Andy

--
Problem reports:       http://cygwin.com/problems.html
FAQ:                   http://cygwin.com/faq/
Documentation:         http://cygwin.com/docs.html
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple

RE: Small request for the new cygwin terminal

[Nellis, Kenneth]

-----Original Message-----
From: Andy Koppe
Subject: Re: Small request for the new cygwin terminal

On 8 June 2012 19:02, Nellis, Kenneth wrote:
> -----Original Message-----
> From: Andy Koppe
> Sent: Thursday, June 07, 2012 2:45 PM
>
<snip/>
>
> Regarding, the "@" key in the table referenced by your URL,
> how does one enter Ctrl+Shift+@, when to get "@" you need to
> already be holding down the Shift key.

Hold down both Shift keys. Yes, really.

(For the record, I wouldn't expect anyone to really use that, and I'm
a bit surprised that I bothered to implement it. Remember though that
different keyboard layouts have the relevant characters in different
positions.)
-----END Original Message-----

Okay, thanx. This leaves one unexplained discrepancy: I don't get
the U+009F result from the "/" key from any combination of Ctrl
and Shift. Not that I'm planning on using it, but, if it's a
documentation error, thought you might appreciate knowing about it.
--Ken Nellis

Re: Small request for the new cygwin terminal

[Helmut Karlowski]

Nellis, Kenneth, 11.06.2012 14:55:02:

> -----Original Message-----
> From: Andy Koppe
> Subject: Re: Small request for the new cygwin terminal
>
> On 8 June 2012 19:02, Nellis, Kenneth wrote:
>> -----Original Message-----
>> From: Andy Koppe
>> Sent: Thursday, June 07, 2012 2:45 PM
>>
> <snip/>
>>
>> Regarding, the "@" key in the table referenced by your URL,
>> how does one enter Ctrl+Shift+@, when to get "@" you need to
>> already be holding down the Shift key.
>
> Hold down both Shift keys. Yes, really.

I don't get anything for Shift-F4. Is this a windows-shortcut? I haven't  
found anything about it. Maybe I have defined it for something in the  
windows-desktop?

-- 
Helmut Karlowski

--
Problem reports:       http://cygwin.com/problems.html
FAQ:                   http://cygwin.com/faq/
Documentation:         http://cygwin.com/docs.html
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple

Re: Small request for the new cygwin terminal

[Andy Koppe]

On 11 June 2012 13:55, Nellis, Kenneth wrote:
> I don't get
> the U+009F result from the "/" key from any combination of Ctrl
> and Shift.

On a US keyboard, Shift+"/" is "?", and Ctrl+"?" is itself a valid
control character combination, producing ^? (i.e. 0x7F).

Andy

--
Problem reports:       http://cygwin.com/problems.html
FAQ:                   http://cygwin.com/faq/
Documentation:         http://cygwin.com/docs.html
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple

Re: Small request for the new cygwin terminal

[Andy Koppe]

On 11 June 2012 21:07, Helmut Karlowski wrote:
> I don't get anything for Shift-F4.

Works for me, producing "^[[1;2S".

> Is this a windows-shortcut?

Not that I know of.

Andy

--
Problem reports:       http://cygwin.com/problems.html
FAQ:                   http://cygwin.com/faq/
Documentation:         http://cygwin.com/docs.html
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple

RE: Small request for the new cygwin terminal

[Nellis, Kenneth]

-----BEGIN Original Message-----
From: Andy Koppe
On 11 June 2012 13:55, Nellis, Kenneth wrote:
> I don't get
> the U+009F result from the "/" key from any combination of Ctrl
> and Shift.

On a US keyboard, Shift+"/" is "?", and Ctrl+"?" is itself a valid
control character combination, producing ^? (i.e. 0x7F).

Andy
-----END Original Message-----
I hate to prolong this relatively uninteresting thread, but I
don't see how you addressed my point, which is that the table at
http://code.google.com/p/mintty/wiki/Keycodes#Ctrl indicates 
that one can generate U+009F with the "/", Ctrl, and Shift keys...
+-----+------+------------+
| Key | Ctrl | Ctrl+Shift |
+-----+------+------------+
|  /  |  ^_  |   U+009F   |
+-----+------+------------+
...but this doesn't seem to be the case, at least with a 
standard US keyboard.
--Ken Nellis

Re: Small request for the new cygwin terminal

[Andy Koppe]

On 12 June 2012 14:11, Nellis, Kenneth wrote:
> -----BEGIN Original Message-----
> From: Andy Koppe
> On 11 June 2012 13:55, Nellis, Kenneth wrote:
>> I don't get
>> the U+009F result from the "/" key from any combination of Ctrl
>> and Shift.
>
> On a US keyboard, Shift+"/" is "?", and Ctrl+"?" is itself a valid
> control character combination, producing ^? (i.e. 0x7F).
>
> Andy
> -----END Original Message-----
> I hate to prolong this relatively uninteresting thread, but I
> don't see how you addressed my point, which is that the table at
> http://code.google.com/p/mintty/wiki/Keycodes#Ctrl indicates
> that one can generate U+009F with the "/", Ctrl, and Shift keys...
> +-----+------+------------+
> | Key | Ctrl | Ctrl+Shift |
> +-----+------+------------+
> |  /  |  ^_  |   U+009F   |
> +-----+------+------------+
> ...but this doesn't seem to be the case, at least with a
> standard US keyboard.

Can you take a look at the latest revision of that wiki page section?
Note the 'Char' instead of 'Key' in the table header, along with a bit
of extra text.

Thanks,
Andy

--
Problem reports:       http://cygwin.com/problems.html
FAQ:                   http://cygwin.com/faq/
Documentation:         http://cygwin.com/docs.html
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple

Re: Small request for the new cygwin terminal

[Andy Koppe]

On 8 June 2012 13:55, Helmut Karlowski wrote:
>
> Andy Koppe wrote:
>> > I've added
>> >
>> > csin=\233,
>> >
>> > but I'm not sure if it's really necessary, and if it's wrong when I switch
>> > the Character Set. tgetstr() returns the value defined in xterm-terminfo,
>> > which may be wrong for another character set. When csin is missing in xterm,
>> > tgetstr() returns 0.
>>
>> Looking at the standard, 'csin' isn't meant for a keycode, but for a
>> command sent to the terminal, namely for the Init sequence for
>> multiple codesets. I'm afraid I don't know what that means though,
>
> That was the closest I've found in the terminfo-manpage. What is the
> terminfo-code for CSI (if any)?

I don't know of one, and I'm not sure it's necessary, since CSI is a
well-defined control character.

> How does vi know about CSI and its
> sequence as mentioned in your previous post?

It's hardcoded to U+009B (decimal 155) in vim:

http://vimdoc.sourceforge.net/htmldoc/intro.html#%3CCSI%3E

And since it works in both UTF-8 and ISO-8859-x, it would appear that
vim does the charset decoding before matching mappings, which is
great.

Andy

--
Problem reports:       http://cygwin.com/problems.html
FAQ:                   http://cygwin.com/faq/
Documentation:         http://cygwin.com/docs.html
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple

Re: Small request for the new cygwin terminal

[Helmut Karlowski]

Andy Koppe wrote:

> I don't think you can have different terminfo entries for different
> character sets for the same terminal type, i.e. you'd need to create
> charset-specifc types such as xterm-cp850.

I suspected that ...

> > I've added
> >
> > csin=\233,
> >
> > but I'm not sure if it's really necessary, and if it's wrong when I switch
> > the Character Set. tgetstr() returns the value defined in xterm-terminfo,
> > which may be wrong for another character set. When csin is missing in xterm,
> > tgetstr() returns 0.
>
> Looking at the standard, 'csin' isn't meant for a keycode, but for a
> command sent to the terminal, namely for the Init sequence for
> multiple codesets. I'm afraid I don't know what that means though,

That was the closest I've found in the terminfo-manpage. What is the
terminfo-code for CSI (if any)? How does vi know about CSI and its
sequence as mentioned in your previous post?

> There isn't a fallback in the terminfo system itself, so I guess
> that's up to each terminfo-using program (or its configuration).

Ok, in that case $TERM is xterm-cp850, xterm-ISO... etc. But I don't
think it's worth the effort.

-Helmut




--
Problem reports:       http://cygwin.com/problems.html
FAQ:                   http://cygwin.com/faq/
Documentation:         http://cygwin.com/docs.html
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple