Debian Linux keyboard mapping files ...

[Albretch Mueller]

there should be files mapping (most probably unicode) number <->
gliph for each language.
From where can I get them?
lbrtchx

[Will Mengarini]

* Albretch Mueller <lbr...@gmail.com> [21-07/01=Th 01:59 -0400]:

> there should be files mapping (most probably unicode) number

> <-> glyph for each language. From where can I get them?

Does </etc/console-setup/boottime.kmap.gz> have what you want?

If not, try `locate kmap`.

[David Wright]

I'm not quite sure I can reconcile your subject line and text.
For the body text, you could look at a site like
https://unicode-table.com/en/
(I googled, without quotes, "unicode charts glyphs".)

But I think you're more interested in keyboard scancodes and their
keycaps (engravings, which vary by language, manufacturer, etc).

AIUI these files are built as necessary from description master
files. After all, many mappings are built around common cores.
The eventual mapping is then placed in /etc/console-setup/, so mine
is in /etc/console-setup/cached_UTF-8_del.kmap.gz

There are scripts in /usr/share/console-setup/, but the raw
information appears to be under /usr/share/X11/xkb/ and
looks complicated. I've never gone beyond consulting these,
and have only /added/ to their Compose sequences through
the /etc/console-setup/remap.inc file.

I did just type
$ apt-file find keymaps
and it looks as though installing "console-data" might yield
some interesting files to peruse, though it doesn't appear
to be a package necessary for building "normal" keyboard
mappings (I don't have it installed).

One can of course dig deeper, and examine the kernel's header files,
but that's way above my paygrade.

Cheers,
David.

[Albretch Mueller]

$ ls -l /etc/console-setup/boottime.kmap.gz
ls: cannot access '/etc/console-setup/boottime.kmap.gz': No such file
or directory
$

Or did you mean in the /etc/console-setup of the installation CD/DVD?
~
$ file /etc/console-setup/compose.KOI8-R.inc
/etc/console-setup/compose.KOI8-R.inc: ASCII text

$ ls -l /etc/console-setup/compose.KOI8-R.inc &
-rw-r--r-- 1 root root 31 Apr 6 2017 /etc/console-setup/compose.KOI8-R.inc

$ wc -l /etc/console-setup/compose.KOI8-R.inc &
1 /etc/console-setup/compose.KOI8-R.inc

$ cat /etc/console-setup/compose.KOI8-R.inc &[2] 8630
# Compose sequences for KOI8-R
$

$ ls -l /etc/console-setup/
total 144
-rw-r--r-- 1 root root 819 Mar 10 2018 cached_ISO-8859-15.acm.gz
-rw-r--r-- 1 root root 4736 Mar 10 2018 cached_ISO-8859-15_del.kmap.gz
-rw-r--r-- 1 root root 2427 Mar 10 2018 cached_Lat15-Fixed16.psf.gz
-rwxr-xr-x 1 root root 473 Feb 17 06:49 cached_setup_font.sh
-rwxr-xr-x 1 root root 199 Feb 17 06:49 cached_setup_keyboard.sh
-rwxr-xr-x 1 root root 73 Feb 17 06:49 cached_setup_terminal.sh
-rw-r--r-- 1 root root 4024 Feb 17 06:49 cached_UTF-8_del.kmap.gz
-rw-r--r-- 1 root root 34 Apr 6 2017 compose.ARMSCII-8.inc
-rw-r--r-- 1 root root 31 Apr 6 2017 compose.CP1251.inc
-rw-r--r-- 1 root root 31 Apr 6 2017 compose.CP1255.inc
-rw-r--r-- 1 root root 31 Apr 6 2017 compose.CP1256.inc
-rw-r--r-- 1 root root 41 Apr 6 2017 compose.GEORGIAN-ACADEMY.inc
-rw-r--r-- 1 root root 36 Apr 6 2017 compose.GEORGIAN-PS.inc
-rw-r--r-- 1 root root 32 Apr 6 2017 compose.IBM1133.inc
-rw-r--r-- 1 root root 35 Apr 6 2017 compose.ISIRI-3342.inc
-rw-r--r-- 1 root root 36 Apr 6 2017 compose.ISO-8859-10.inc
-rw-r--r-- 1 root root 36 Apr 6 2017 compose.ISO-8859-11.inc
-rw-r--r-- 1 root root 3737 Apr 6 2017 compose.ISO-8859-13.inc
-rw-r--r-- 1 root root 3020 Apr 6 2017 compose.ISO-8859-14.inc
-rw-r--r-- 1 root root 3552 Apr 6 2017 compose.ISO-8859-15.inc
-rw-r--r-- 1 root root 36 Apr 6 2017 compose.ISO-8859-16.inc
-rw-r--r-- 1 root root 3596 Apr 6 2017 compose.ISO-8859-1.inc
-rw-r--r-- 1 root root 2893 Apr 6 2017 compose.ISO-8859-2.inc
-rw-r--r-- 1 root root 3387 Apr 6 2017 compose.ISO-8859-3.inc
-rw-r--r-- 1 root root 2805 Apr 6 2017 compose.ISO-8859-4.inc
-rw-r--r-- 1 root root 35 Apr 6 2017 compose.ISO-8859-5.inc
-rw-r--r-- 1 root root 35 Apr 6 2017 compose.ISO-8859-6.inc
-rw-r--r-- 1 root root 1217 Apr 6 2017 compose.ISO-8859-7.inc
-rw-r--r-- 1 root root 35 Apr 6 2017 compose.ISO-8859-8.inc
-rw-r--r-- 1 root root 3617 Apr 6 2017 compose.ISO-8859-9.inc
-rw-r--r-- 1 root root 31 Apr 6 2017 compose.KOI8-R.inc
-rw-r--r-- 1 root root 31 Apr 6 2017 compose.KOI8-U.inc
-rw-r--r-- 1 root root 32 Apr 6 2017 compose.TIS-620.inc
-rw-r--r-- 1 root root 31 Apr 6 2017 compose.VISCII.inc
-rw-r--r-- 1 root root 1359 Apr 6 2017 remap.inc
$

> I'm not quite sure I can reconcile your subject line and text.

Given the baseline option a user chooses as "language" during
installation, the codes sent by the keyboard should be interpreted.

There should be files with the associations of (unicode) numbers and
keys on a keyboard. I don't think that information is "secret" in any
way. Where can I find those files?

> The eventual mapping is then placed in /etc/console-setup/, so mine
> is in /etc/console-setup/cached_UTF-8_del.kmap.gz

$ ls -l /etc/console-setup/cached_UTF-8_del.kmap.gz
-rw-r--r-- 1 root root 4024 Feb 17 06:49
/etc/console-setup/cached_UTF-8_del.kmap.gz

$ cp -v /etc/console-setup/cached_UTF-8_del.kmap.gz /home/$(whoami)/temp
'/etc/console-setup/cached_UTF-8_del.km
ap.gz' -> '/home/debian/temp/cached_UTF-8_del.kmap.gz'

and in this file:

$ ls -l /home/debian/temp/cached_UTF-8_del.kmap/cached_UTF-8_del.kmap
-rw-r--r-- 1 debian debian 122418 Jul 1 11:51
/home/debian/temp/cached_UTF-8_del.kmap/cached_UTF-8_del.kmap

I see no indication whatsoever about the fact that I am using a
standard dvorak keyboard right now and that I am typing in English.

I mean something like:

http://xahlee.info/kbd/russian_keyboard_layout.html

Most probably I can't get why anyone would spend time doing such
things, but Xah Lee seems to be interested on the (IMO nonsensical)
"heat map" produced by typing the first chapter of Dostoevsky's "Crime
and Punishment".

I just need, given the language, the unicode keys on regular keyboards
out there.

I may be wrong, but I don't think that my English is so bad and/or my
question so unhinged.

> the raw information appears to be under /usr/share/X11/xkb/

$ ls -l "/usr/share/X11/xkb/"
total 24
drwxr-xr-x 2 root root 4096 Feb 17 06:30 compat
drwxr-xr-x 4 root root 4096 Feb 17 06:30 geometry
drwxr-xr-x 4 root root 4096 Feb 17 06:30 keycodes
drwxr-xr-x 2 root root 4096 Feb 17 06:30 rules
drwxr-xr-x 14 root root 4096 Feb 17 06:30 symbols
drwxr-xr-x 2 root root 4096 Feb 17 06:30 types
$

find "/usr/share/X11/xkb" -type f -exec grep -il "iso\|lang" {} \;
...

bingo!, but as you said those files look worrying complicated and intratextual.

"The way I understand reality"; given a language, there should be
just tables associating unicodes to keys on a keyboard

>I did just type
>$ apt-file find keymaps
>and it looks as though installing "console-data" might yield
>some interesting files to peruse

# apt-get install console-data
Reading package lists... Done
Building dependency tree
Reading state information... Done
The following additional packages will be installed:
console-common
Suggested packages:
unicode-data
The following NEW packages will be installed:
console-common console-data
0 upgraded, 2 newly installed, 0 to remove and 147 not upgraded.
Need to get 1,270 kB of archives.
After this operation, 2,996 kB of additional disk space will be used.
Do you want to continue? [Y/n] Y
Get:1 http://deb.debian.org/debian stretch/main amd64 console-data all
2:1.12-5 [1,150 kB]
Get:2 http://deb.debian.org/debian stretch/main amd64 console-common
all 0.7.89 [120 kB]
Fetched 1,270 kB in 0s (2,998 kB/s)
Preconfiguring packages ...
Selecting previously unselected package console-data.
(Reading database ... 265142 files and directories currently installed.)
Preparing to unpack .../console-data_2%3a1.12-5_all.deb ...
Unpacking console-data (2:1.12-5) ...
Selecting previously unselected package console-common.
Preparing to unpack .../console-common_0.7.89_all.deb ...
Unpacking console-common (0.7.89) ...
Setting up console-data (2:1.12-5) ...
Looking for keymap to install:
NONE
Setting up console-common (0.7.89) ...
Looking for keymap to install:
NONE
Processing triggers for man-db (2.7.6.1-2) ...
#


$ ls -l "/usr/share/console-setup/"
total 44
-rw-r--r-- 1 root root 272 Apr 6 2017 console-setup
-rwxr-xr-x 1 root root 1793 Oct 9 2016 kbdnames-maker
-rw-r--r-- 1 root root 140 Apr 6 2017 keyboard
-rw-r--r-- 1 root root 31177 Apr 6 2017 KeyboardNames.pl

$ file "/usr/share/console-setup/KeyboardNames.pl"
/usr/share/console-setup/KeyboardNames.pl: Perl script text executable

$

KeyboardNames.pl contains associations of keyboard types with ISO
639-1 language codes

But the kind of lookup tables I need I couldn't find.

# apt-get install unicode-data
Reading package lists... Done
Building dependency tree
Reading state information... Done
The following NEW packages will be installed:
unicode-data
0 upgraded, 1 newly installed, 0 to remove and 147 not upgraded.
Need to get 7,080 kB of archives.
After this operation, 33.6 MB of additional disk space will be used.
Get:1 http://deb.debian.org/debian stretch/main amd64 unicode-data all
9.0-1 [7,080 kB]
Fetched 7,080 kB in 1s (4,260 kB/s)
Selecting previously unselected package unicode-data.
(Reading database ... 265852 files and directories currently installed.)
Preparing to unpack .../unicode-data_9.0-1_all.deb ...
Unpacking unicode-data (9.0-1) ...
Setting up unicode-data (9.0-1) ...
#

and in "/usr/share/unicode/", I found files which look like what I am
looking for, but not quite.

Again, there files with look up tables determined by:
a) user language
b) type of keyboard
c) input method (at times a sequence of key strokes should be entered)

Why is it so hard to find such a thing?

lbrtchx

[Greg Wooledge]

On Thu, Jul 01, 2021 at 12:46:01PM -0400, Albretch Mueller wrote:
> Given the baseline option a user chooses as "language" during
> installation, the codes sent by the keyboard should be interpreted.

> There should be files with the associations of (unicode) numbers and
> keys on a keyboard. I don't think that information is "secret" in any
> way. Where can I find those files?

https://wiki.debian.org/Keyboard

This points to the configuration file /etc/default/keyboard.

That points to the keyboard(5) man page.

Which in turn points to the following resources:

http://www.xfree86.org/current/XKB-Config.html
https://wiki.archlinux.org/index.php/X_KeyBoard_extension
http://pascal.tsu.ru/en/xkb/
http://www.charvolant.org/~doug/xkb/
http://xfree86.org/current/XKBproto.pdf

The Arch wiki page points me to the xkbcomp(1) command, which tells me
that keymaps are stored in a compiled form, and stored within the X
server's memory. You can use xkbcomp to dump this to a file, and then
inspect the file. Of course this does not tell us where the files
normally live on the system, to be slurped up by the X server as it's
beginning its lifespan.

While looking for such files, which I did not quite find, I *did* find
/usr/share/X11/xkb/symbols/us which looks relevant. At least for me,
since my keyboard is configured to have XKBLAYOUT="us".

[David Wright]

On Thu 01 Jul 2021 at 12:46:01 (-0400), Albretch Mueller wrote:
> $ ls -l /etc/console-setup/boottime.kmap.gz
> ls: cannot access '/etc/console-setup/boottime.kmap.gz': No such file
> or directory
> $
>
> Or did you mean in the /etc/console-setup of the installation CD/DVD?
> ~

No idea. That was somebody else's post.

> $ file /etc/console-setup/compose.KOI8-R.inc
> /etc/console-setup/compose.KOI8-R.inc: ASCII text
>
> $ ls -l /etc/console-setup/compose.KOI8-R.inc &
> -rw-r--r-- 1 root root 31 Apr 6 2017 /etc/console-setup/compose.KOI8-R.inc
>
> $ wc -l /etc/console-setup/compose.KOI8-R.inc &
> 1 /etc/console-setup/compose.KOI8-R.inc
>
> $ cat /etc/console-setup/compose.KOI8-R.inc &[2] 8630
> # Compose sequences for KOI8-R
> $
>
> $ ls -l /etc/console-setup/
> total 144
> -rw-r--r-- 1 root root 819 Mar 10 2018 cached_ISO-8859-15.acm.gz
> -rw-r--r-- 1 root root 4736 Mar 10 2018 cached_ISO-8859-15_del.kmap.gz
> -rw-r--r-- 1 root root 2427 Mar 10 2018 cached_Lat15-Fixed16.psf.gz

Those files look like they might date from the original installation.

> -rwxr-xr-x 1 root root 473 Feb 17 06:49 cached_setup_font.sh
> -rwxr-xr-x 1 root root 199 Feb 17 06:49 cached_setup_keyboard.sh
> -rwxr-xr-x 1 root root 73 Feb 17 06:49 cached_setup_terminal.sh
> -rw-r--r-- 1 root root 4024 Feb 17 06:49 cached_UTF-8_del.kmap.gz

Yes, they look rather like mine. The first part of
cached_UTF-8_del.kmap.gz can look really boring unless you're tracking
down some obscure modifier combination, but then it's followed by the
Compose sequences that get dpkg-reconfigure'd into it from remap.inc.

I've never worked out whether they do anything for me, as they appear
to be written in different (perhaps their titular) codesets.

> -rw-r--r-- 1 root root 1359 Apr 6 2017 remap.inc

That's the file I replace. The original version consists only of comments.

> $
>
> > I'm not quite sure I can reconcile your subject line and text.
>
> Given the baseline option a user chooses as "language" during
> installation, the codes sent by the keyboard should be interpreted.
>
> There should be files with the associations of (unicode) numbers and
> keys on a keyboard. I don't think that information is "secret" in any
> way. Where can I find those files?

There's no such table: it cannot exist. Which unicode number would you
assign to CapsLock, or RightShift. There are several layers of
translation which lie between pressing/releasing a key and assigning
a character to the result. Some of these tables are built up out of
component parts, like the basic letter keys, the "shift"s at their
edges, function keys, keypads, multimedia, etc.

I guess the answer is "what useful use would it have?"

For a start, mapping key depressions to unicode text is a many-to-one
mapping. There are keystrokes that I never use—sensible combinations
like ø and ¶, where I find it easier to Compose them by pressing
CapsLock (= Compose) and a sequence of keys¹. No shifting, and the
key sequences make logical sense to me. There's no way of telling
which method was used, or which Shift key I use with which other keys.

Then there are tools to find out "what does this key do?", like
showkey and xev. One approach might be to build you own table by
capturing the output from xev and just striking all the keys in turn.
That gives you the coordinates of the keystrokes to form any
character. Use a character-counter to ascertain the "vocabulary" of
the text, so that you only have to determine the sequence of
keystrokes for characters that exist in the text when you build your
table.

¹ AltGr o yields ø, fair enough,
but
CapsLock /o yields ø
CapsLock 'o yields ó
CapsLock `o yields ò
CapsLock ^o yields ô
CapsLock ~o yields õ
CapsLock -o yields ō
CapsLock "o yields ö
CapsLock !o yields ọ
CapsLock .o yields ȯ
CapsLock #o yields º
CapsLock oo yields °
and there's really no limit, so long as I can recall them:
CapsLock co yields ©
CapsLock ro yields ®
CapsLock so yields §
CapsLock %o yields ‰
and you don't need an AltGr key, and you can configure it
to seamlessly work on both VC and in X.

Cheers,
David.

[Albretch Mueller]

David Chartash at the corpora research mailing list pointed out to me
I could find what I wanted at:

http://kbdlayout.info/

and within Debian using `man 5 keyboard`

> There's no such table: it cannot exist. Which unicode number would you
> assign to CapsLock, or RightShift. There are several layers of
> translation which lie between pressing/releasing a key and assigning
> a character to the result. Some of these tables are built up out of
> component parts, like the basic letter keys, the "shift"s at their
> edges, function keys, keypads, multimedia, etc.

> For a start, mapping key depressions to unicode text is a many-to-one
> mapping.

Well, when I said "look up table" I meant also such sequences of
chars including escape sequences which end up being written as a
character in text files. Non-alphabetical languages use input methods.

> ¹ AltGr o yields ø, fair enough,
> but
> CapsLock /o yields ø
> CapsLock 'o yields ó
> CapsLock `o yields ò
> CapsLock ^o yields ô
> CapsLock ~o yields õ
> CapsLock -o yields ō
> CapsLock "o yields ö
> CapsLock !o yields ọ
> CapsLock .o yields ȯ
> CapsLock #o yields º
> CapsLock oo yields °
> and there's really no limit, so long as I can recall them:
> CapsLock co yields ©
> CapsLock ro yields ®
> CapsLock so yields §
> CapsLock %o yields ‰
> and you don't need an AltGr key, and you can configure it
> to seamlessly work on both VC and in X.

From your examples you included I will only need yielded glyphs if
they are commonly used in a language. Now, defining "commonly used"
would be an entirely different, yet valid question.

I will have to code my way through those files to parse unicode <->
key (or key sequence) "lookup tables" for each language and my effort
will need definitely more than "parsing" for non-alphabetical
languages.

thank you,
lbrtchx

[Greg Wooledge]

On Fri, Jul 02, 2021 at 02:24:20AM -0400, Albretch Mueller wrote:
> From your examples you included I will only need yielded glyphs if
> they are commonly used in a language. Now, defining "commonly used"
> would be an entirely different, yet valid question.
>
> I will have to code my way through those files to parse unicode <->
> key (or key sequence) "lookup tables" for each language and my effort
> will need definitely more than "parsing" for non-alphabetical
> languages.

So, wait, you're NOT just abnormally curious? You actually believe you
have some kind of REASON for wanting to know how the console and/or X
keyboard input mechanism is implemented by Debian?

If this is the case, stop messing around and TELL US what that reason
is. What are you actually trying to do?

As far as I've been able to tell with my rather limited knowledge of
this topic, you're starting from some MASSIVELY incorrect assumptions,
but up until now, correcting all the background noise was never
important, because you were just poking around out of curiosity. Or so
we thought.

[Albretch Mueller]

On 7/2/21, Greg Wooledge <gr...@wooledge.org> wrote:
> you're starting from some MASSIVELY incorrect assumptions,
> but up until now, correcting all the background noise was never
> important, because you were just poking around out of curiosity. Or so
> we thought.

I don't understand why "we" think "I was just poking around out of
curiosity" (provided "we" has some powerful mind reading skills). I
had spent some time finding such files and I thought there should be a
way to somehow get such files out of Debian's installation disks. I
also realized why I couldn't find the link that was suggested to me. I
always exclude Windows results from google searches.

I also included the results here in case anyone else has such needs.

lbrtchx

[Albretch Mueller]

It occurred to me that my use of the term "mapping" may have been a
little confusing. I used it in general and as part of my corpora
research I am moving away from UTF-8. That is all I am doing.

lbrtchx

[David Wright]

On Fri 02 Jul 2021 at 02:24:20 (-0400), Albretch Mueller wrote:
> David Chartash at the corpora research mailing list pointed out to me
> I could find what I wanted at:
>
> http://kbdlayout.info/

That's for Windows, isn't it.

> and within Debian using `man 5 keyboard`

Yes, as it says, this will explain how you use the files I mentioned
to configure a keyboard.

> > There's no such table: it cannot exist. Which unicode number would you
> > assign to CapsLock, or RightShift. There are several layers of
> > translation which lie between pressing/releasing a key and assigning
> > a character to the result. Some of these tables are built up out of
> > component parts, like the basic letter keys, the "shift"s at their
> > edges, function keys, keypads, multimedia, etc.
>
> > For a start, mapping key depressions to unicode text is a many-to-one
> > mapping.
>
> Well, when I said "look up table" I meant also such sequences of
> chars including escape sequences which end up being written as a
> character in text files. Non-alphabetical languages use input methods.

Yes, that's why I brought up Compose sequences. One might think of
Compose as the introducer of printed-characters just as Escape is
of Control-characters.

My point about the mapping of unicode → keys depressed² seems to
have been missed. On this keyboard, I can type ø by
. holding AltGr and typing o
. typing CapsLock / o
. typing CapsLock o /
so if you were to meet ø in your subject text, how would you
determine which keys were struck in order to add them to your
heat map?

> > ¹ AltGr o yields ø, fair enough,
> > but
> > CapsLock /o yields ø
> > CapsLock 'o yields ó
> > CapsLock `o yields ò
> > CapsLock ^o yields ô
> > CapsLock ~o yields õ
> > CapsLock -o yields ō
> > CapsLock "o yields ö
> > CapsLock !o yields ọ
> > CapsLock .o yields ȯ
> > CapsLock #o yields º
> > CapsLock oo yields °
> > and there's really no limit, so long as I can recall them:
> > CapsLock co yields ©
> > CapsLock ro yields ®
> > CapsLock so yields §
> > CapsLock %o yields ‰
> > and you don't need an AltGr key, and you can configure it
> > to seamlessly work on both VC and in X.
>
> From your examples you included I will only need yielded glyphs if
> they are commonly used in a language. Now, defining "commonly used"
> would be an entirely different, yet valid question.

If you look at my email address, you'll realise that I don't type
diacritics very often. So composing one as above, from its appearance,
is much easier than memorising which shifted key³ it's bound to.
And the same emphasis on appearance is how "they" and I choose to
compose symbols, which I type much more frequently.

> I will have to code my way through those files to parse unicode <->
> key (or key sequence) "lookup tables" for each language and my effort
> will need definitely more than "parsing" for non-alphabetical
> languages.

² Most of the documentation I've seen concerns itself with the
direction keys depressed → characters rather than the opposite.
After all, any reverse document presentation is going to be
incomplete, and nobody much cares because all you need to do
in most editors to type a bizarre character is specify its
raw codepoint.

³ None of the British and American keyboards layouts in that
reference above bear much resemblance to the ones I possess,
particularly in their engraved diacritics.

Cheers,
David.

[to...@tuxteam.de]

On Sun, Jul 04, 2021 at 10:25:13AM -0500, David Wright wrote:

[...]

> My point about the mapping of unicode → keys depressed² seems to
> have been missed. On this keyboard, I can type ø by
> . holding AltGr and typing o
> . typing CapsLock / o
> . typing CapsLock o /

Ah, Compose on CapsLock. Great minds think alike, it seems ☺

Cheers
- t

[Albretch Mueller]

On 7/4/21, David Wright <deb...@lionunicorn.co.uk> wrote:
> On Fri 02 Jul 2021 at 02:24:20 (-0400), Albretch Mueller wrote:
>> David Chartash at the corpora research mailing list pointed out to me
>> I could find what I wanted at:
>>
>> http://kbdlayout.info/
>
> That's for Windows, isn't it.

Yes, but at the end of the day the written characters in a text is
what matters to my line of work.

> My point about the mapping of unicode → keys depressed² seems to
> have been missed.

...

> so if you were to meet ø in your subject text, how would you
> determine which keys were struck in order to add them to your
> heat map?

I mentioned "heat patterns" when I talk about his work (which I don't
understand anyway). I used his work because he "mapped" the key
sequences and resulting chars for different languages in a useful way
to me.

lbrtchx

[Greg Wooledge]

On Sun, Jul 04, 2021 at 03:34:22PM -0400, Albretch Mueller wrote:
> On 7/4/21, David Wright <deb...@lionunicorn.co.uk> wrote:
> > On Fri 02 Jul 2021 at 02:24:20 (-0400), Albretch Mueller wrote:
> >> David Chartash at the corpora research mailing list pointed out to me
> >> I could find what I wanted at:
> >>
> >> http://kbdlayout.info/
> >
> > That's for Windows, isn't it.
>
> Yes, but at the end of the day the written characters in a text is
> what matters to my line of work.

Some day he's going to tell us what he's trying to do, and it's going
to turn out not only to be completely off-topic, but also the biggest
X-Y problem of the year.

If the analysis of characters inside a text file is "what matters", then
any investigation into how various operating systems and windowing
systems handle keyboard input is completely irrevelant. Just read the
files.

[David Wright]

—Racecar backwards is racecar, racecar upside down is expensive—

—My wife can type tesseradecades while drinking a cup of tea—

Earlier, the OP wrote “Xah Lee seems to be interested on the (IMO

nonsensical) "heat map" produced by typing the first chapter of

Dostoevsky's "Crime and Punishment".” I have no idea who Xah Lee is,
but heat map is the modern way of saying that you're going to contour
the frequency of keys struck during the typing of said test.

As the OP also mentioned that they used a Dvorak layout keyboard,
then it's pretty obvious that their contour map will likely be
different from someone using the standard layout.

At least on a standard keyboard, every letter key has a dedicated
finger. So when you type racecar, one finger has to move twice
to type c e c, without using any other finger in between. So
another heat map could be generated by considering finger movements
between characters.

One can make this more complex. I looked at the Dvorak wiki, and I see
that they even consider the frequency of inboard stroke flow, not a
term I had come across.

When the OP writes “at the end of the day the written characters in a
text is what matters to my line of work”, they may be attempting to
distance themselves from Xah Lee's work, described here as nonsensical.
People might have said that about Dvorak's work too. I believe the
OP's line of work is corpora research, which I assume only involves
the characters in the text, and not how they're entered into the
computer.

—Most typing is done by the left hand—

—About ten times as many words can be typed
with only the left hand as with the right—

Cheers,
David.