Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.debian.user > #224422 > unrolled thread

Using .XCompose

Back to article view | Back to linux.debian.user

This discussion starts older than the indexed window; earlier articles aren't shown. The article labeled Started by below is the oldest one visible, not the original post.

Contents

  Using .XCompose Ajith R <ajithramayyan@yahoo.co.in> - 2020-07-06 07:40 +0200
    Re: Using .XCompose Greg Wooledge <wooledg@eeg.ccf.org> - 2020-07-06 14:30 +0200
    Re: Using .XCompose Zenaan Harkness <zenaan@freedbms.net> - 2020-07-06 14:50 +0200
    Re: Using .XCompose davidson <davidson@freevolt.org> - 2020-07-06 21:20 +0200
      Re: Using .XCompose davidson <davidson@freevolt.org> - 2020-07-06 22:40 +0200
        Re: Using .XCompose Ajith R <ajithramayyan@yahoo.co.in> - 2020-07-07 17:20 +0200
      Re: Using .XCompose davidson <davidson@freevolt.org> - 2020-07-07 08:40 +0200
        Re: Using .XCompose Ajith R <ajithramayyan@yahoo.co.in> - 2020-07-07 17:20 +0200
          Re: Using .XCompose davidson <davidson@freevolt.org> - 2020-07-09 23:50 +0200
            Re: Using .XCompose [resend of xev_output_annotated.txt] davidson <davidson@freevolt.org> - 2020-07-10 03:40 +0200
            Using .XCompose Ajith R <ajithramayyan@yahoo.co.in> - 2020-07-10 10:50 +0200
              Re: Using .XCompose davidson <davidson@freevolt.org> - 2020-07-11 10:40 +0200
                Re: Using .XCompose davidson <davidson@freevolt.org> - 2020-07-11 12:30 +0200
                  Re: Using .XCompose Ajith R <ajithramayyan@yahoo.co.in> - 2020-07-12 17:00 +0200
                    Re: Using .XCompose David Wright <deblis@lionunicorn.co.uk> - 2020-07-13 06:00 +0200
                    Re: Using .XCompose Greg Wooledge <wooledg@eeg.ccf.org> - 2020-07-13 14:00 +0200
                      Re: Using .XCompose Ajith R <ajithramayyan@yahoo.co.in> - 2020-07-13 20:30 +0200
                        Re: Using .XCompose Greg Wooledge <wooledg@eeg.ccf.org> - 2020-07-13 20:40 +0200
                          Re: Using .XCompose Ajith R <ajithramayyan@yahoo.co.in> - 2020-07-13 20:50 +0200
                            Re: Using .XCompose David Wright <deblis@lionunicorn.co.uk> - 2020-07-13 21:50 +0200
                              Re: Using .XCompose Ajith R <ajithramayyan@yahoo.co.in> - 2020-07-14 05:40 +0200
                          Fw: Using .XCompose Ajith R <ajithramayyan@yahoo.co.in> - 2020-07-15 16:50 +0200
                          Fw: Using .XCompose Ajith R <ajithramayyan@yahoo.co.in> - 2020-07-15 17:10 +0200
                          Fw: Using .XCompose Ajith R <ajithramayyan@yahoo.co.in> - 2020-07-15 17:20 +0200
                    Re: Using .XCompose davidson <davidson@freevolt.org> - 2020-07-16 01:30 +0200
                      Re: Using .XCompose davidson <davidson@freevolt.org> - 2020-07-16 02:40 +0200
                      Re: Using .XCompose Ajith R <ajithramayyan@yahoo.co.in> - 2020-07-16 20:40 +0200
                        how u mine 4 utf8 [was Re: Using .XCompose] davidson <davidson@freevolt.org> - 2020-07-18 03:20 +0200
                          UTF-8 Everywhere -- Re: how u mine 4 utf8 [was Re: Using .XCompose] Zenaan Harkness <zenaan@freedbms.net> - 2020-07-18 07:40 +0200
                          Re: how u mine 4 utf8 [was Re: Using .XCompose] davidson <davidson@freevolt.org> - 2020-07-18 09:20 +0200
                            Re: how u mine 4 utf8 [was Re: Using .XCompose] Ajith R <ajithramayyan@yahoo.co.in> - 2020-07-19 13:00 +0200
                          Fw: how u mine 4 utf8 [was Re: Using .XCompose] Ajith R <ajithramayyan@yahoo.co.in> - 2020-07-19 12:50 +0200
                            Re: Fw: how u mine 4 utf8 [was Re: Using .XCompose] Tom Browder <tom.browder@gmail.com> - 2020-07-19 13:10 +0200
                              Re: how u mine 4 utf8 [was Re: Using .XCompose] Ajith R <ajithramayyan@yahoo.co.in> - 2020-07-19 13:50 +0200
                        sed:awk:perl::rock:paper:chainsaw [was Re: Using .XCompose] davidson <davidson@freevolt.org> - 2020-07-18 12:10 +0200
                          Re: sed:awk:perl::rock:paper:chainsaw [was Re: Using .XCompose] Zenaan Harkness <zenaan@freedbms.net> - 2020-07-18 12:50 +0200
                            Re: sed:awk:perl::rock:paper:chainsaw [was Re: Using .XCompose] William Michels <wjm1@caa.columbia.edu> - 2020-07-18 22:00 +0200
                          Re: sed:awk:perl::rock:paper:chainsaw [was Re: Using .XCompose] Andrei POPESCU <andreimpopescu@gmail.com> - 2020-07-18 13:20 +0200
                            Re: sed:awk:perl::rock:paper:chainsaw [was Re: Using .XCompose] songbird <songbird@anthive.com> - 2020-07-18 21:00 +0200
                          Re: sed:awk:perl::rock:paper:chainsaw [was Re: Using .XCompose] <tomas@tuxteam.de> - 2020-07-18 16:00 +0200
                            Re: sed:awk:perl::rock:paper:chainsaw [was Re: Using .XCompose] mick crane <mick.crane@gmail.com> - 2020-07-18 22:40 +0200
                              Re: sed:awk:perl::rock:paper:chainsaw [was Re: Using .XCompose] <tomas@tuxteam.de> - 2020-07-18 23:10 +0200
                                Re: sed:awk:perl::rock:paper:chainsaw [was Re: Using .XCompose] mick crane <mick.crane@gmail.com> - 2020-07-18 23:30 +0200
                                  Re: sed:awk:perl::rock:paper:chainsaw [was Re: Using .XCompose] <tomas@tuxteam.de> - 2020-07-19 10:40 +0200
                                    Re: sed:awk:perl::rock:paper:chainsaw [was Re: Using .XCompose] mick crane <mick.crane@gmail.com> - 2020-07-20 13:20 +0200
                          Re: sed:awk:perl::rock:paper:chainsaw [was Re: Using .XCompose] David Christensen <dpchrist@holgerdanske.com> - 2020-07-19 01:50 +0200
                            Re: sed:awk:perl::rock:paper:chainsaw [was Re: Using .XCompose] Ajith R <ajithramayyan@yahoo.co.in> - 2020-07-19 12:50 +0200
                              Re: sed:awk:perl::rock:paper:chainsaw [was Re: Using .XCompose] David Christensen <dpchrist@holgerdanske.com> - 2020-07-19 23:30 +0200
                          Re: sed:awk:perl::rock:paper:chainsaw [was Re: Using .XCompose] Ajith R <ajithramayyan@yahoo.co.in> - 2020-07-19 13:40 +0200
                            Re: sed:awk:perl::rock:paper:chainsaw [was Re: Using .XCompose] William Michels <wjm1@caa.columbia.edu> - 2020-07-19 21:50 +0200
                            Re: sed:awk:perl::rock:paper:chainsaw [was Re: Using .XCompose] David Wright <deblis@lionunicorn.co.uk> - 2020-07-23 04:00 +0200
                              Re: sed:awk:perl::rock:paper:chainsaw [was Re: Using .XCompose] songbird <songbird@anthive.com> - 2020-07-23 12:50 +0200
                    Re: Using .XCompose davidson <davidson@freevolt.org> - 2020-07-16 02:00 +0200
                      Re: Using .XCompose Ajith R <ajithramayyan@yahoo.co.in> - 2020-07-16 20:40 +0200
                Re: Using .XCompose David Wright <deblis@lionunicorn.co.uk> - 2020-07-11 17:00 +0200
                  Tex characters Teemu Likonen <tlikonen@iki.fi> - 2020-07-11 18:20 +0200
                  Re: Using .XCompose davidson <davidson@freevolt.org> - 2020-07-11 22:40 +0200
                    Re: Using .XCompose Greg Wooledge <wooledg@eeg.ccf.org> - 2020-07-13 14:00 +0200
                      Re: Using .XCompose davidson <davidson@freevolt.org> - 2020-07-16 07:50 +0200
                Re: Using .XCompose davidson <davidson@freevolt.org> - 2020-07-11 18:10 +0200
                Re: Using .XCompose Greg Wooledge <wooledg@eeg.ccf.org> - 2020-07-20 15:30 +0200
                Re: Using .XCompose Ajith R <ajithramayyan@yahoo.co.in> - 2020-07-20 15:30 +0200
                  Re: Using .XCompose David Wright <deblis@lionunicorn.co.uk> - 2020-07-20 18:20 +0200
            Re: Using .XCompose davidson <davidson@freevolt.org> - 2020-07-10 12:50 +0200
      Re: [offlist] posting style [was Re: Using .XCompose] Ajith R <ajithramayyan@yahoo.co.in> - 2020-07-08 05:00 +0200
        Re: [offlist] posting style [was Re: Using .XCompose] Andrei POPESCU <andreimpopescu@gmail.com> - 2020-07-08 07:00 +0200
        Re: [not-so-offlist] posting style [was Re: Using .XCompose] davidson <davidson@freevolt.org> - 2020-07-08 07:10 +0200
      Re: [offlist] Re: Using .XCompose Ajith R <ajithramayyan@yahoo.co.in> - 2020-07-08 05:00 +0200
        Re: [offlist] Re: Using .XCompose davidson <davidson@freevolt.org> - 2020-07-08 05:20 +0200
        Re: [offlist] Re: Using .XCompose Zenaan Harkness <zenaan@freedbms.net> - 2020-07-08 05:40 +0200
          Re: [offlist] Re: Using .XCompose davidson <davidson@freevolt.org> - 2020-07-08 08:00 +0200
    Re: Using .XCompose David Wright <deblis@lionunicorn.co.uk> - 2020-07-07 00:40 +0200
      tangent on man page locations [was Re: Using .XCompose] davidson <davidson@freevolt.org> - 2020-07-07 01:10 +0200
        Re: tangent on man page locations [was Re: Using .XCompose] David Wright <deblis@lionunicorn.co.uk> - 2020-07-07 03:50 +0200
      Re: Using .XCompose Ajith R <ajithramayyan@yahoo.co.in> - 2020-07-07 11:50 +0200
        Re: Using .XCompose Greg Wooledge <wooledg@eeg.ccf.org> - 2020-07-07 13:40 +0200
          Re: Using .XCompose Ajith R <ajithramayyan@yahoo.co.in> - 2020-07-08 16:30 +0200
        Re: Using .XCompose David Wright <deblis@lionunicorn.co.uk> - 2020-07-09 02:20 +0200
          Re: Using .XCompose Zenaan Harkness <zenaan@freedbms.net> - 2020-07-09 04:20 +0200
    Re: Using .XCompose Zenaan Harkness <zenaan@freedbms.net> - 2020-07-08 01:50 +0200
      Re: Using .XCompose davidson <davidson@freevolt.org> - 2020-07-08 02:30 +0200
      Fw: Using .XCompose Ajith R <ajithramayyan@yahoo.co.in> - 2020-07-08 05:10 +0200
      Re: Using .XCompose Ajith R <ajithramayyan@yahoo.co.in> - 2020-07-08 18:40 +0200
        Re: Using .XCompose Zenaan Harkness <zenaan@freedbms.net> - 2020-07-09 04:00 +0200
          Re: Using .XCompose Zenaan Harkness <zenaan@freedbms.net> - 2020-07-10 02:40 +0200
            Re: Using .XCompose The Wanderer <wanderer@fastmail.fm> - 2020-07-10 02:50 +0200
        Re: Using .XCompose davidson <davidson@freevolt.org> - 2020-07-10 05:00 +0200

Page 4 of 5 — ← Prev page 1 2 3 [4] 5  Next page →

#225038

On Mon, Jul 20, 2020 at 01:20:19PM +0000, Ajith R wrote:
> I tried to start firefox using the command above. Fire fox gave the error 
> (firefox-esr:1554): Gtk-WARNING **: 15:05:11:541 : GTK+ suports to output one char only: "a long substitution": <W> : "a long substitution"

> What next steps do you suggest for me? Should I submit a bug report? Where?

Upstream libgtk, sounds like.

<https://www.gtk.org/community/> exists and talks about an issue
tracker, but it's a corporate-kindergarten-looking page designed for
"mobile".  Good luck getting anything done there.

[toc] | [prev] | [next] | [standalone]

#225040

Hi,


> Try again, for firefox-esr (and with a ~/.XCompose file that is not
> befouled with nonbreaking spaces).
> 
> But make one change to the procedure. When you launch firefox-esr, do
> so like this:
> 
> $ env GTK_IM_MODULE=xim firefox-esr
> 
> Let us know how that goes.

I made a fresh install of Debian on an old 32 bit netbook. Before I did anything to the fresh install, I wrote just one line to the .XCompose file:
<W> : "a long substitution"

Using Kite, I tried typing W while holding down the shift key. I got just the letter "a".
I tried to start firefox using the command above. Fire fox gave the error 
(firefox-esr:1554): Gtk-WARNING **: 15:05:11:541 : GTK+ suports to output one char only: "a long substitution": <W> : "a long substitution"
The error was repeated 8 times (with different timetsamps) before the firefox window came up. When I tried to type W, either the W remained as such or the field showed a list of string starting with W.

What next steps do you suggest for me? Should I submit a bug report? Where?

Thanks,
ajith

[toc] | [prev] | [next] | [standalone]

#225050

On Mon 20 Jul 2020 at 13:20:19 (+0000), Ajith R wrote:
[> Someone else wrote:]
> > Try again, for firefox-esr (and with a ~/.XCompose file that is not
> > befouled with nonbreaking spaces).
> > 
> > But make one change to the procedure. When you launch firefox-esr, do
> > so like this:
> > 
> > $ env GTK_IM_MODULE=xim firefox-esr
> > 
> > Let us know how that goes.
> 
> I made a fresh install of Debian on an old 32 bit netbook. Before I did anything to the fresh install, I wrote just one line to the .XCompose file:
> <W> : "a long substitution"
> 
> Using Kite, I tried typing W while holding down the shift key. I got just the letter "a".
> I tried to start firefox using the command above. Fire fox gave the error 
> (firefox-esr:1554): Gtk-WARNING **: 15:05:11:541 : GTK+ suports to output one char only: "a long substitution": <W> : "a long substitution"
> The error was repeated 8 times (with different timetsamps) before the firefox window came up. When I tried to type W, either the W remained as such or the field showed a list of string starting with W.

Yes, my experiment with multiple characters produced the same warning.
So it's no surprise that it doesn't work. We're trusting to luck that
some applications appear to support it, at least in the case where
certain other conditions hold. What those conditions are—who knows.

> What next steps do you suggest for me? Should I submit a bug report? Where?

Not much point. The obvious reply will be to refer you to the man page:

   "Compose(5)                File Formats Manual                Compose(5)

    NAME
       Compose - X client mappings for multi-key input sequences

    DESCRIPTION
       The X library, libX11, provides a simple input method for characters
       beyond those represented on typical keyboards using sequences of key
       strokes that are combined to enter a single character."

                                            ↑↑↑↑↑↑

Whether XKB can support what you're trying to do, I don't know.
Some of the static phenomena you described seem to be entirely
handled by the rendering of Unicode, and the fonts. Where the
dynamic phenomenon I've read about (the modal behaviour of Delete
and Backspace) is supported, I don't know.

My own interference with the keyboard layout is done in console-setup,
not the X configuration. I can refer you back to my first reply and
https://lists.debian.org/debian-user/2019/07/msg00926.html
which gives you the entire contents of my /etc/console-setup/remap.inc
file. Note that it contains strings.

That's the file where I indulge in a little cargo-cult behaviour:
you have to configure it in a console, but previous experience has
suggested that it *might* not work properly when X is even running,
ie I either kill X, or don't startx it.

(You don't need X to test whether your edits worked, of course, so
an efficient workflow cycle of
  edit/dpkg-reconfigure/test/edit/dpkg-reconfigure/test/…
involves using a text-mode editor.)

Cheers,
David.

[toc] | [prev] | [next] | [standalone]

#224681

[Multipart message — attachments visible in raw view] — view raw

On Thu, 9 Jul 2020 davidson wrote:
> I have conducted an experiment with the following ~/.XCompose file,
> and tried to document it relatively thoroughly below.
[snip]

> %% THREE XCOMPOSE RULES to test
>
> $ cat ~/.XCompose
> include "%L"
> <U0D19>   : "ങ്ങ" # Ajith's auto-geminate rule, (U0D19) => (U0D19) (U0D4D) 
> (U0D19)
> <Multi_key> <s> <x> : "✄"  # (U2704) white scissors, h/t David Wright
> <Z>                 : "ARGA WARGA IN THE DARGA instead of Z"
>
[snip]

There is a mistake here. A missing step.

> %% ARGA WARGA
>
> The third rule ('Z' => "ARGA WARGA...Z") was easiest to test:
>
> 1. Open the application. (If necessary, navigate to a text entry field.)

STEP 1-AND-A-HALF. Hold down the Shift key.

> 2. (While under "us" keyboard layout) strike the key labeled "Z" on my
>    standard US laptop keyboard
> 3. Observe whether
>
>      ARGA WARGA IN THE DARGA instead of Z
>
>    is printed.
>
> The two terminal emulators passed this test. The graphical web browsers 
> failed
> in an interesting way. Instead of displaying
>
>     ARGA WARGA IN THE DARGA instead of Z
>
> in the text entry field, they displayed just
>
>     A

-- 
You won't feel the collar if you don't go anywhere.

[toc] | [prev] | [next] | [standalone]

#224560 — Re: [offlist] posting style [was Re: Using .XCompose]

Hi David,

>>>
Debian conformant style looks like this instead:
>>>

Do I understand the style correctly now? BTW, a query about how the original message is quoted with '<'. Is it done by hand or does any email clients do it automatically?  Is it necessary that each line s marked with '<'? Or is it okay to demarcate the quote at the start and end as I do?

Thanks,
ajith

[toc] | [prev] | [next] | [standalone]

#224567 — Re: [offlist] posting style [was Re: Using .XCompose]

[Multipart message — attachments visible in raw view] — view raw

On Mi, 08 iul 20, 02:52:02, Ajith R wrote:
> 
> Do I understand the style correctly now?

Mostly. Think of it as "conversational" style, with the reply following 
the portion of text it addresses.

Other text should be removed, unless it provides useful context (would 
your message make sense if read by someone not familiar with the issue 
and without access to previous messages?).

> BTW, a query about how the original message is quoted with '<'. Is it 
> done by hand or does any email clients do it automatically?

Most email clients should do this automatically when you select "plain 
text" mode.

HTML messages may look completely different at the receiving side, which 
is why they should be avoided. One can use other characters for 
emphasis, like *bold*, /italics/, _underline_.

> Is it necessary that each line s marked with '<'? Or is it okay to 
> demarcate the quote at the start and end as I do?

In my opinion the most important part is clearly distinguishing between 
quote and reply, as well as properly attributing quotes (who wrote 
what).

Using '>' makes it easier for list recipients to read your message, e.g.  
my email client is configured to use different colours depending on the 
quote level.

Kind regards,
Andrei
-- 
http://wiki.debian.org/FAQsFromDebianUser

[toc] | [prev] | [next] | [standalone]

#224568 — Re: [not-so-offlist] posting style [was Re: Using .XCompose]

[Multipart message — attachments visible in raw view] — view raw

On Wed, 8 Jul 2020 Ajith R wrote:
>
>
> Hi David,
>
>>>>
> Debian conformant style looks like this instead:
>>>>
>
> Do I understand the style correctly now?

LOL. Probably not. Maybe I should have allowed your discovery of the
mailing list's quoting conventions to take its natural course.

It was maybe silly of me to explain it, before we've figured out why
some of your seemingly reasonable XCompose lines, like this one

   <U0D19> : "ങ്ങ"  # Ajith's auto-geminate rule

aren't effective.

> BTW, a query about how the original message is quoted with '<'.

That left angle bracket '<' points in the wrong direction.

A conventional quoted-line prefix is

   "> "

which is a right angle bracket followed by a space.

A mail client that understands this convention may interpret these
indicators, and display any text so prefixed in some other distinctive
fashion. The prefix is mere markup that means "The following text on
this line is quoted material".

And so what one sees displayed in a mail client may not exactly
correspond to the plain text of the message body. Instead of literal
angle bracket prefixes, you might see some other visual representation
of the embedding.

Examining message bodies in a text editor (or with a pager like
"less") might help reduce confusion.

> Is it done by hand or does any email clients do it automatically?

Only the artisanal hand-quoter would be able to tell the difference!

And I'm sure there exists at least one maverick mail client that does
not do so automatically.

But I have not taken a survey of email clients. Mine (alpine) does do
so automatically whenever I include a message body in my reply to that
message. And even if it did not automatically prefix the lines of the
quoted message body, then it would not be difficult to tell any
decently functional text editor to employ

  "> "

as a comment string, and then select and comment blocks of quoted
text.

And these prefixes will stack, over the course of a thread. Naturally
enough, email lends itself to replies. So you have an initial message
(degree zero), replies to such a message (degree one), and replies to
replies (degree two), and so on.

A message of the Nth degree might retain lines of quoted material from
N previous messages. And if it does, then one expects to find each of
those lines constituting the most deeply embedded material prefixed
with N instances of

  "> "

> Is it necessary that each line s marked with '<'?

All the cool people prefix each line of a quoted message body with

  "> "

It is cool to be cool, but not strictly necessary. If it were
necessary, then it couldn't be cool.

> Or is it okay to demarcate the quote at the start and end as I do?

It disrupts the convention. I would not do this.

-- 
You won't feel the collar if you don't go anywhere.

[toc] | [prev] | [next] | [standalone]

#224561 — Re: [offlist] Re: Using .XCompose

Hi David,

>>>
As a rule, send replies to the listserv address:

    debian-user@lists.debian.org

2. In-Line replies, instead of Top-Posting
3. I apologise for being a Terrible Person

>>>

I didn't pay attention to the fact that the from address in the email was yours and not that of the list server. I just replied to the email. I will try to remember to correct the to address

I hope my replies confirm to the Bottom-Line style you suggested in another mail.

I thank you for showing the patience to explain to us the newcomers the conventions of this list.

Thanks,
ajith

[toc] | [prev] | [next] | [standalone]

#224563 — Re: [offlist] Re: Using .XCompose

[Multipart message — attachments visible in raw view] — view raw

On Wed, 8 Jul 2020 Ajith R wrote:
> Hi David,
>
>>>>
> As a rule, send replies to the listserv address:
>
>     debian-user@lists.debian.org
>
> 2. In-Line replies, instead of Top-Posting
> 3. I apologise for being a Terrible Person
>
>>>>

> I didn't pay attention to the fact that the from address in the
> email was yours and not that of the list server. I just replied to
> the email. I will try to remember to correct the to address

LOL. No big deal either way.

> I hope my replies confirm to the Bottom-Line style you suggested in another mail.

Working on a reply to that part.

> I thank you for showing the patience

No patience required.

> to explain to us the newcomers the conventions of this list.

They say that those who cannot, will teach. That might apply in this
case.

-- 
You won't feel the collar if you don't go anywhere.

[toc] | [prev] | [next] | [standalone]

#224564 — Re: [offlist] Re: Using .XCompose

On Wed, Jul 08, 2020 at 02:59:20AM +0000, Ajith R wrote:
> Hi David,
> 
> >>>
> As a rule, send replies to the listserv address:
> 
>     debian-user@lists.debian.org
> 
> 2. In-Line replies, instead of Top-Posting
> 3. I apologise for being a Terrible Person


Hey!  I thought I had a patent, trademark _and_ copyright on self deprecation ;)

[toc] | [prev] | [next] | [standalone]

#224575 — Re: [offlist] Re: Using .XCompose

[Multipart message — attachments visible in raw view] — view raw

On Wed, 8 Jul 2020 Zenaan Harkness wrote:
> On Wed, Jul 08, 2020 at 02:59:20AM +0000, Ajith R wrote:
>> Hi David,
>>
>>>>>
>> As a rule, send replies to the listserv address:
>>
>>     debian-user@lists.debian.org
>>
>> 2. In-Line replies, instead of Top-Posting
>> 3. I apologise for being a Terrible Person
>
> Hey!  I thought I had a patent, trademark _and_ copyright on self
> deprecation ;)

Neuroticism wants to be free.

-- 
You won't feel the collar if you don't go anywhere.

[toc] | [prev] | [next] | [standalone]

#224468

On Mon 06 Jul 2020 at 05:34:59 (+0000), Ajith R wrote:

> My .XCompose file in my home directory is---------------------include "%L" 
> <XK_Shift_L> <U0D19> : "ങ്ങ"
> <U0D19> : "ങ്ങ"
> ങ : "ങ്ങ"---------------------
> I found the name XK_Shift_L in keysymdef.h file. I tried the unicode character and its code as well to identify the keypress. I tried with only one of the lines as well. XCompose(3) — libx11-doc — Debian buster — Debian Manpages was consulted. I have tried restarting after making changes which didn't help. 
> Am I referring to the keypress correctly? What am I missing?

I presume XCompose(3) is a typo for 5, the file format section.
There you will find that you don't use the XK_ prefix here.

As I understand .XCompose, it is designed for "compositing", so each
key on the left side should be a "normal" keystroke. Typically the
LeftShift modifies the letter keys: in English, it makes "a" into "A".
So you wouldn't want to use it in a compositing sequence because you
would lose its normal shift function.

Usually you select a "Compose" key, and that is used here under the name
<Multi_key> in .XCompose, and in the system's Compose file,
/usr/share/X11/locale/…/Compose where … is a locale's name.
Mine, for example, is CapsLock, because I don't make a habit of typing
in All Caps. That is set in /etc/default/keyboard:
XKBOPTIONS="lv3:ralt_switch,compose:caps,terminate:ctrl_alt_bksp"
                            ↑↑↑↑↑↑↑↑↑↑↑↑

If you peruse your own system's /usr/share/X11/locale/…/Compose file,
which is the one being included by your "%L" above, you'll get an idea
of the best keys to use. I believe many European languages use the
"dead" keys a lot, but because British doesn't, I don't know how those
work. A European might like to comment on this.

In case it's any help, I posted some keyboard configuration files
last year at:

https://lists.debian.org/debian-user/2019/07/msg00926.html

The first section is an incomplete set of function keys for my window
manager; the next section defines keys for VCs (where I include the
system defaults to preserve them); following that is my .XCompose.
The system defaults already include most of what I want, so I add
only 14 more. The final section is a few definitions from my emacs
configuration.

Cheers,
David.

[toc] | [prev] | [next] | [standalone]

#224470 — tangent on man page locations [was Re: Using .XCompose]

On Mon, 6 Jul 2020 David Wright wrote:

> I presume XCompose(3) is a typo for 5, the file format section.

Probably not. Here we have

   $ zcat /usr/share/man/man3/XCompose.3.gz
   .so man5/Compose.5

   $ realpath /usr/share/man/man5/XCompose.5.gz
   /usr/share/man/man5/Compose.5.gz

and hence

   $ man -w 3 XCompose
   /usr/share/man/man5/Compose.5.gz

   $ man -w 5 XCompose
   /usr/share/man/man5/Compose.5.gz

Which is to say, XCompose(5) and XCompose(3) refer to the same
document.

-- 
What do you want to take off? [hrzF or ?*] F
You were wearing a +0 robe.  The frost giant turns to flee.

[toc] | [prev] | [next] | [standalone]

#224472 — Re: tangent on man page locations [was Re: Using .XCompose]

On Mon 06 Jul 2020 at 23:08:29 (+0000), davidson wrote:
> On Mon, 6 Jul 2020 David Wright wrote:
> 
> > I presume XCompose(3) is a typo for 5, the file format section.
> 
> Probably not. Here we have
> […]
> and hence
> 
>   $ man -w 3 XCompose
>   /usr/share/man/man5/Compose.5.gz
> 
>   $ man -w 5 XCompose
>   /usr/share/man/man5/Compose.5.gz
> 
> Which is to say, XCompose(5) and XCompose(3) refer to the same
> document.

Fair enough; I didn't know someone had dropped this alias into
section 3; it certainly looks out of place. But I keep well away
from X11 calls, and hence libx11-doc, and would advise the OP
to do so also:

Package: libx11-doc
Description-en: X11 client-side library (development documentation)
 This package provides a client interface to the X Window System, otherwise
 known as 'Xlib'.  It provides a complete API for the basic functions of the
 window system.
 This package contains supplemental documentation for the library found in
 libx11-6. Non-developers likely have little use for this package.

Having just looked at the OP again (via the HTML this time),
I see there's an external link to that library (which my lynx
suppresses). So it's possible the OP hasn't installed libx11-doc.

Cheers,
David.

[toc] | [prev] | [next] | [standalone]

#224484

[Multipart message — attachments visible in raw view] — view raw

 Hi David,
Thanks for your reply.>>>>I presume XCompose(3) is a typo for 5, the file format section.
There you will find that you don't use the XK_ prefix here.>>>>
I was looking at XCompose(3). Thanks for pointing out the mistake.
>>>>
As I understand .XCompose, it is designed for "compositing", so each
key on the left side should be a "normal" keystroke. Typically the
LeftShift modifies the letter keys: in English, it makes "a" into "A".
So you wouldn't want to use it in a compositing sequence because you
would lose its normal shift function.>>>>I understand that. My letter is in the 'shift' set of keystrokes. That was the reason I included it. Moreover, the Compose manual page states that 
'Each event consists of a specified input keysym, and optional modifier states:
[([!] ([~] MODIFIER)...) | None] <keysym>". '

I was just trying to find out the reason why the 'composting' was not working. I tried including the modifier name and also without it. I tried moving the key to the first layer of keyboard layout (without shift or other modifiers). Even that was not working. So, I was wondering if there was a syntax error in the way I was specifying my requirement. I understand the use of Compose key. But, I was not trying the usual 'composting' (by means of a compose key). I  want one character to be replaced with its geminate form whenever it is typed. If the keyboard layout specification allowed more than one character for a keystroke, I wouldn't have needed it. Hence the need for Compose mechanism. A similar approach is explained here https://dominiko.livejournal.com/20206.html . However in it, the author has defined a personal use Unicode character, but I use one of the Unicode defined characters. In the symbol file, he had used key <AD01> { type[group1] = "FOUR_LEVEL_C_HWERTY", 
                [            UF8FD,          UF8FE,           UF8FF,    Greek_alpha ] };
I didn't understand the use of type[group1 = "..."], but thought that it was not essential. Anyway, the layout is working as expected in that the characters returned by layout is as defined. 
The compose mechanism is not working.
I copied one line from your keyboard configuration file and  my keyboard file looks like this now :# KEYBOARD CONFIGURATION FILE

XKBMODEL="pc105"
XKBLAYOUT="in"
XKBVARIANT="eng"
XKBOPTIONS="lv3:ralt_switch,compose:caps,terminate:ctrl_alt_bksp"

BACKSPACE="guess"

and my .XCompose file now is 
include "%L" 
<Multi_key> <s> <x>                     : "✄"   U2704 # WHITE SCISSORS

However, when I press Caps Lock followed by s and then x or if I keep the Caps Lock pressed while I press s and x in sequence, I don't get scissors. I get S and X

So, my computer's  compose mechanism is not working. Is there something wrong in what I am doing or am I missing something? Other than creating the .XCompose file in the home folder, should I tell the system in some way that the compose mechanism should be activated? (BTW, I tried restarting the computer after the change).

I am running Debian 10, KDE Plasma  5.14.5, 64 bit

Thanks,ajith
    On Tuesday, 7 July, 2020, 4:00:50 am IST, David Wright <deblis@lionunicorn.co.uk> wrote:  
 
 On Mon 06 Jul 2020 at 05:34:59 (+0000), Ajith R wrote:

> My .XCompose file in my home directory is---------------------include "%L" 
> <XK_Shift_L> <U0D19> : "ങ്ങ"
> <U0D19> : "ങ്ങ"
> ങ : "ങ്ങ"---------------------
> I found the name XK_Shift_L in keysymdef.h file. I tried the unicode character and its code as well to identify the keypress. I tried with only one of the lines as well. XCompose(3) — libx11-doc — Debian buster — Debian Manpages was consulted. I have tried restarting after making changes which didn't help. 
> Am I referring to the keypress correctly? What am I missing?

I presume XCompose(3) is a typo for 5, the file format section.
There you will find that you don't use the XK_ prefix here.

As I understand .XCompose, it is designed for "compositing", so each
key on the left side should be a "normal" keystroke. Typically the
LeftShift modifies the letter keys: in English, it makes "a" into "A".
So you wouldn't want to use it in a compositing sequence because you
would lose its normal shift function.

Usually you select a "Compose" key, and that is used here under the name
<Multi_key> in .XCompose, and in the system's Compose file,
/usr/share/X11/locale/…/Compose where … is a locale's name.
Mine, for example, is CapsLock, because I don't make a habit of typing
in All Caps. That is set in /etc/default/keyboard:
XKBOPTIONS="lv3:ralt_switch,compose:caps,terminate:ctrl_alt_bksp"
                            ↑↑↑↑↑↑↑↑↑↑↑↑

If you peruse your own system's /usr/share/X11/locale/…/Compose file,
which is the one being included by your "%L" above, you'll get an idea
of the best keys to use. I believe many European languages use the
"dead" keys a lot, but because British doesn't, I don't know how those
work. A European might like to comment on this.

In case it's any help, I posted some keyboard configuration files
last year at:

https://lists.debian.org/debian-user/2019/07/msg00926.html

The first section is an incomplete set of function keys for my window
manager; the next section defines keys for VCs (where I include the
system defaults to preserve them); following that is my .XCompose.
The system defaults already include most of what I want, so I add
only 14 more. The final section is a few definitions from my emacs
configuration.

Cheers,
David.

  

[toc] | [prev] | [next] | [standalone]

#224486

On Tue, Jul 07, 2020 at 09:44:38AM +0000, Ajith R wrote:
> I copied one line from your keyboard configuration file and  my keyboard file looks like this now :# KEYBOARD CONFIGURATION FILE
> 
> XKBMODEL="pc105"
> XKBLAYOUT="in"
> XKBVARIANT="eng"
> XKBOPTIONS="lv3:ralt_switch,compose:caps,terminate:ctrl_alt_bksp"
> 
> BACKSPACE="guess"
> 
> and my .XCompose file now is 
> include "%L" 
> <Multi_key> <s> <x>                     : "✄"   U2704 # WHITE SCISSORS
> 
> However, when I press Caps Lock followed by s and then x or if I keep the Caps Lock pressed while I press s and x in sequence, I don't get scissors. I get S and X

I have no experience with modifying /etc/default/keyboard.
Does your Caps Lock key work for Compose if you set it with:

xmodmap -e "keysym Caps_Lock = Multi_key"

Also, as mentioned before, please make sure you open a new terminal
(or other X11 application) after modifying your .XCompose file.  That
file is only read when libx11 initializes during the start of a new
client application.

[toc] | [prev] | [next] | [standalone]

#224597

Hi Greg,

>>>>>
Does your Caps Lock key work for Compose if you set it with:

xmodmap -e "keysym Caps_Lock = Multi_key"
>>>>>

I tried this and found that the test definition in .XCopose file is working.
However, when I change the .XCompose definition to <Multi_key> <ങ>                     : "✄"
or to
<Multi_key> <ങ> <ങ>                      : "✄", it is not working
or to
<Multi_key> <U0D19>  <U0D19>                : "ങ്ങ"



>>>>>
Also, as mentioned before, please make sure you open a new terminal
(or other X11 application) after modifying your .XCompose file.  That
file is only read when libx11 initializes during the start of a new
client application.
>>>>>

Just an observation. When it works, the Compose definition works in the terminal from which the command is issued. I did test the definitions that don't work in a new terminal and Kate.

Thanks,
ajith

[toc] | [prev] | [next] | [standalone]

#224611

Sorry if the format looks a mess, but it's a consequence of replying to HTML.

On Tue 07 Jul 2020 at 09:44:38 (+0000), Ajith R wrote:

> I was looking at XCompose(3). Thanks for pointing out the mistake.

It wasn't a mistake on your part. It was pointed out to me that
section 3 has an alias redirecting to the real man page, which is
actually still in the right place, section 5. You just followed that
link. I was worried that you might have been reading a real section 3
man page, most of which are incomprehensible to us mere mortals.

> > As I understand .XCompose, it is designed for "compositing", […]

> I understand that. My letter is in the 'shift' set of
> keystrokes. That was the reason I included it. Moreover, the Compose
> manual page states that 'Each event consists of a specified input
> keysym, and optional modifier states:
> [([!] ([~] MODIFIER)...) | None] <keysym>". '

Sure. But typically, for ordinary letter keys, you would expect each
to have its own keysym when shifted. Thus (using English) you see
definitions sequences like

<Multi_key> <o> <c>                     : "©"   copyright # COPYRIGHT SIGN
<Multi_key> <o> <C>                     : "©"   copyright # COPYRIGHT SIGN
<Multi_key> <O> <c>                     : "©"   copyright # COPYRIGHT SIGN
<Multi_key> <O> <C>                     : "©"   copyright # COPYRIGHT SIGN
<Multi_key> <C> <o>                     : "©"   copyright # COPYRIGHT SIGN
<Multi_key> <C> <O>                     : "©"   copyright # COPYRIGHT SIGN

Obviously, I'm struggling when it comes to your own example as I don't
know how these characters relate to one another. So my examples will
be using Latin characters.

It's perfectly possible to write compositions like these:

<x>                                     : "xx"
<X>                                     : "XX"

where X is shifted x.

You can even have:

<y> <y>                                 : "yes"

but the problem is that whenever you type the key y, the system will
wait for another character to be typed. If you want to type the letter
y, you have to type y y Rubout Rubout. For the letter X, you have to
type X Rubout. Not very satisfactory.

What appears on the right is flexible, but the left part, before the
colon, is more constrained. To repeat myself, you should probably look
at the system file for your own locale (and perhaps others) to see
what they use. My own, /usr/share/X11/locale/en_US.UTF-8/Compose, has
the letters (in both cases) and digits (including <KP_3> as well as <3>,
surprisingly), and punctuation keys by name (<comma> etc). My Compose
key (CapsLock) is <Multi_key>.

I also see keys that I have no idea about how to generate them, like
<Greek_ALPHA> <Cyrillic_GHE> <Arabic_alef> and most of the
Unicode-defined ones. But noticeable is the fact that the *first* key is
almost always <Multi_key> or <dead something or other>. That's because
those few keys become unusable on their own, dead. (There are also 9
oddities that mean nothing to me.)

What I don't see are any modifiers like shift and ctrl. In fact, I
found none in /usr/share/X11/locale/*/Compose on my system. Perhaps
yours do, in which case you can follow their style. I can't help
you there.

> I was just trying to find out the reason why the 'composting' was
> not working. I tried including the modifier name and also without
> it. I tried moving the key to the first layer of keyboard layout
> (without shift or other modifiers). Even that was not working.

I think you're now talking about the /usr/share/X11/xkb/ files,
and overriding them in /etc/. I don't touch those, and can't help you
there. I read them, to help in configuring /etc/default/ files, but
no more than that.

> So, I
> was wondering if there was a syntax error in the way I was
> specifying my requirement. I understand the use of Compose key. But,
> I was not trying the usual 'composting' (by means of a compose
> key). I  want one character to be replaced with its geminate form
> whenever it is typed. If the keyboard layout specification allowed
> more than one character for a keystroke, I wouldn't have needed
> it. Hence the need for Compose mechanism. A similar approach is
> explained here https://dominiko.livejournal.com/20206.html . However
> in it, the author has defined a personal use Unicode character, but
> I use one of the Unicode defined characters.

If you're willing to lose the normal function of the key, then as
shown above, you can turn any keystroke into a string.

> In the symbol file, he had used key
> <AD01> { type[group1] = "FOUR_LEVEL_C_HWERTY", [ UF8FD, UF8FE, UF8FF, Greek_alpha ] };
> I didn't understand the use of type[group1 = "..."], but thought
> that it was not essential. Anyway, the layout is working as expected
> in that the characters returned by layout is as defined.

No idea about that stuff.

> The compose mechanism is not working.

Syntax errors are one possibility. I don't know how and when they are
handled. Another is making changes have any effect. (In the past,
I've edited the file in a VC, and started X from another VC after any
changes have been saved.)

A third is that every recipe at the left side must not entirely
include any other. Hence:

<Multi_key> <minus> <minus> <minus>     : "—"   U2014 # EM DASH
<Multi_key> <minus> <minus> <period>    : "–"   U2013 # EN DASH

and not:

<Multi_key> <minus> <minus> <minus>     : "—"   U2014 # EM DASH
<Multi_key> <minus> <minus>             : "–"   U2013 # EN DASH

> I copied one line from your keyboard configuration file and
> my keyboard file looks like this now :

> # KEYBOARD CONFIGURATION FILE
> XKBMODEL="pc105"
> XKBLAYOUT="in"
> XKBVARIANT="eng"
> XKBOPTIONS="lv3:ralt_switch,compose:caps,terminate:ctrl_alt_bksp"
> 
> BACKSPACE="guess"

I should have said that the usual way to write this file is to run
# dpkg-reconfigure keyboard-configuration
(and that command can also be run after editing it in order to
validate/correct it). I should also add that in my experience,
it's best done in a VC, without an X session even running.

(In systemd, there's also this mysterious command
# service keyboard-setup restart
about which I know nothing. I've never used it.)

> and my .XCompose file now is 
> include "%L" 
> <Multi_key> <s> <x>                     : "✄"   U2704 # WHITE SCISSORS
> 
> However, when I press Caps Lock followed by s and then x or if I
> keep the Caps Lock pressed while I press s and x in sequence, I
> don't get scissors. I get S and X.

That appears to indicate that CapsLock is not your Compose Key, but is
behaving as normal, turning s and x into S and X.

> So, my computer's  compose mechanism is not working. Is there
> something wrong in what I am doing or am I missing something? Other
> than creating the .XCompose file in the home folder, should I tell
> the system in some way that the compose mechanism should be
> activated? (BTW, I tried restarting the computer after the change). 
> 
> I am running Debian 10, KDE Plasma  5.14.5, 64 bit

I have no idea how a DE interacts (interferes) with any normal
linux/X system's configuration or behaviour. Sorry.

Cheers,
David.

[toc] | [prev] | [next] | [standalone]

#224617

On Wed, Jul 08, 2020 at 07:14:33PM -0500, David Wright wrote:

> > I was just trying to find out the reason why the 'composting' was
> > not working. I tried including the modifier name and also without
> > it. I tried moving the key to the first layer of keyboard layout
> > (without shift or other modifiers). Even that was not working.
> 
> I think you're now talking about the /usr/share/X11/xkb/ files,
> and overriding them in /etc/. I don't touch those, and can't help you
> there. I read them, to help in configuring /etc/default/ files, but
> no more than that.

I added a new "FunkyZen" layout file to /usr/share/X11/xkb/, a modification of the US keymap I think.

Can we create new layouts by just adding a file to /etc ??


> I should have said that the usual way to write this file is to run
> # dpkg-reconfigure keyboard-configuration
> (and that command can also be run after editing it in order to
> validate/correct it). I should also add that in my experience,
> it's best done in a VC, without an X session even running.

Ooh!  Now _that_ could be highly useful next time he (this friend of mine who seems to be oddly wearing a brown paper bag right now) looks at my, I mean his custom keyboard layout...


> (In systemd, there's also this mysterious command
> # service keyboard-setup restart
> about which I know nothing. I've never used it.)

Must investigate.

If my friend were to write an ad hock end user custom keyboard layout customizer, what would be the minimal set of perms required to do its job?  (My friend does not use `sudo` you see...)

> > and my .XCompose file now is 
> > include "%L" 
> > <Multi_key> <s> <x>                     : "✄"   U2704 # WHITE SCISSORS
> > 
> > However, when I press Caps Lock followed by s and then x or if I
> > keep the Caps Lock pressed while I press s and x in sequence, I
> > don't get scissors. I get S and X.
> 
> That appears to indicate that CapsLock is not your Compose Key, but is
> behaving as normal, turning s and x into S and X.

Although various desktops have GUIs to choose Compose key, it is Better ™©® to set this at a lower level so it also kinda works in the Linux console.  I guess my friend's question is how best to arrange the perms/sudo side of things for a GUI that wants to make this easy for end users?

[toc] | [prev] | [next] | [standalone]

#224545

I'm not sure if it's been asked or stated by you, but which desktop are you using?  If you're new to Debian, perhaps just the default.

I am using XFCE and in the menu:

   Applications → Settings → Keyboard

there are three tabs "Behaviour", "Application Shortcuts" and "Layout".

Under the "Layout" tab, I have the option for "Compose key" - this is a drop down list, and I chose the "Scroll Lock" key as my compose key.

As you can see in the above, I am producing arrows, by pressing:

   "Scroll Lock"  and then  ">"

For my standard US layout, ">" is "Shift + Period".

I don't know this will be useful, but I did not catch that you had checked your desktop settings.

Good luck,




On Mon, Jul 06, 2020 at 05:34:59AM +0000, Ajith R wrote:
> Hi,
> I am new to Linux and Debian.I am trying to build a custom layout for my mother tongue Malayalam (India, Kerala).
> The problem I am trying to tackle:
> One of the Malayalam letters, ങ (U+0D19), is used much more commonly in its geminate form which is composed of three unicode charcters ങ ് ങ (U+0D19 U+0D4D U+0D19)  which will yield the geminate form ങ്ങ So, when I type, I want the ങ to be replaced with ങ്ങ.The keyboard layout is working fine as far as I can see. It does yield the correct letters. What doesn't seem to work is the .XCompose mechanism. Please tell me what I am missing.
> Relevant part of my layout file /usr/share/X11/xkb/symbols/in
> -----------------------
> partial alphanumeric_keys
> xkb_symbols "mal_puthuniraA" {
>      name[Group1] = "Malayalam (Puthu Nira Aarambham)";
>      key.type="FOUR_LEVEL";
>  key <AC09>   { [   U0D41, U0D19, U0D4E, U0D71] }; //  ു MALAYALAM VOWEL SIGN U, ങ്ങ MALAYALAM LETTER NGA geminate form,  ൎMALAYALAM LETTER DOT REPH,  ൱ MALAYALAM NUMBER ONE HUNDRED include "level3(ralt_switch)"
> };----------------------
> My .XCompose file in my home directory is---------------------include "%L" 
> <XK_Shift_L> <U0D19> : "ങ്ങ"
> <U0D19> : "ങ്ങ"
> ങ : "ങ്ങ"---------------------
> I found the name XK_Shift_L in keysymdef.h file. I tried the unicode character and its code as well to identify the keypress. I tried with only one of the lines as well. XCompose(3) — libx11-doc — Debian buster — Debian Manpages was consulted. I have tried restarting after making changes which didn't help. 
> Am I referring to the keypress correctly? What am I missing?
> Thanks for your help,ajith
> 
> 
> | 
> | 
> | 
> |  |  |
> 
>  |
> 
>  |
> | 
> |  | 
> XCompose(3) — libx11-doc — Debian buster — Debian Manpages
> 
> 
>  |
> 
>  |
> 
>  |
> 
> 
> 
> 

[toc] | [prev] | [next] | [standalone]

Page 4 of 5 — ← Prev page 1 2 3 [4] 5  Next page →

Back to top | Article view | linux.debian.user

csiph-web