• You are not logged in.

Locale Colemak variants for several countries (the eDreymaR way)!

  • Started by DreymaR
  • 32 Replies:
  • Reputation: 23
  • From: Belgium
  • Registered: 26-Feb-2008
  • Posts: 482

Similarly, the modern Georgian keyboard layout is, more or less phonetically, based on Qwerty as well (the traditional typewriter layout is no longer used).  Although the Georgian Mkhedruli alphabet has 33 letters, it has no capital letters, so by putting extra letters on Shift and AltGr, they were able to occupy only the standard letter block, and keep all numbers and symbols in original Qwerty position.  This allows for another straight-forward Colemak variant:

ქ წ ფ პ გ ჯ ლ უ ყ ; [ ]
 ა რ ს ტ დ ჰ ნ ე ი ო '
  ზ ხ ც ვ ბ კ მ , . /

I don't know the Georgian language at all, so this is just a theoretical exercise, but in xkb it would look like this:

partial alphanumeric_keys
xkb_symbols "colemak" {
    include "ge(basic)"

    name[Group1]= "Georgian (Colemak)";

    key <AD01> { [ Georgian_khar, Q                            ] };
    key <AD02> { [ Georgian_cil,  Georgian_char                ] };
    key <AD03> { [ Georgian_phar, F,             Georgian_fi   ] };
    key <AD04> { [ Georgian_par,  P                            ] };
    key <AD05> { [ Georgian_gan,  G,             0x010010f9    ] };
    key <AD06> { [ Georgian_jhan, Georgian_zhar, 0x010010f7    ] };
    key <AD07> { [ Georgian_las,  L                            ] };
    key <AD08> { [ Georgian_un,   U                            ] };
    key <AD09> { [ Georgian_qar,  Y,             0x010010f8    ] };
    key <AD10> { [ semicolon,     colon                        ] };

    key <AC01> { [ Georgian_an,   A,             0x010010fa    ] };
    key <AC02> { [ Georgian_rae,  Georgian_ghan, 0x010000ae    ] };
    key <AC03> { [ Georgian_san,  Georgian_shin                ] };
    key <AC04> { [ Georgian_tar,  Georgian_tan                 ] };
    key <AC05> { [ Georgian_don,  D                            ] };
    key <AC06> { [ Georgian_hae,  H,             Georgian_hoe  ] };
    key <AC07> { [ Georgian_nar,  N,             0x010010fc    ] };
    key <AC08> { [ Georgian_en,   E,             Georgian_he   ] };
    key <AC09> { [ Georgian_in,   I,             Georgian_hie  ] };
    key <AC10> { [ Georgian_on,   O                            ] };

    key <AB01> { [ Georgian_zen,  Georgian_jil                 ] };
    key <AB02> { [ Georgian_xan,  X,             Georgian_har  ] };
    key <AB03> { [ Georgian_can,  Georgian_chin, 0x010000a9    ] };
    key <AB04> { [ Georgian_vin,  V,             Georgian_we   ] };
    key <AB05> { [ Georgian_ban,  B                            ] };
    key <AB06> { [ Georgian_kan,  K                            ] };
    key <AB07> { [ Georgian_man,  M                            ] };

    include "capslock(backspace)"
    include "level3(ralt_switch)"
    include "space(level4)"

};
Last edited by ghen (10-Mar-2018 16:36:55)
Offline
  • 1
  • Reputation: 23
  • From: Belgium
  • Registered: 26-Feb-2008
  • Posts: 482

Any Qwerty-based layout can systemetically be remapped to Colemak in the same way.
The following sed script does so for xkb layouts:

#!/bin/sed -f
  
s/AC04/AD03/;t  # E>F
s/AD05/AC04/;t  # F>T
s/AC05/AD05/;t  # T>G
s/AC03/AC05/;t  # G>D
s/AC02/AC03/;t  # D>S
s/AD04/AC02/;t  # S>R
s/AD10/AD04/;t  # R>P
s/AC10/AD10/;t  # P>;
s/AD09/AC10/;t  # ;>O
s/AD06/AD09/;t  # O>Y
s/AC07/AD06/;t  # Y>J
s/AB06/AC07/;t  # J>N
s/AC08/AB06/;t  # N>K
s/AD03/AC08/;t  # K>E

s/AC09/AD07/;t  # U>L
s/AD08/AC09/;t  # L>I
s/AD07/AD08/;t  # I>Y

This can be applied to the layout to be shuffled from within vim: use "v" to visually select the layout section and pipe it through the script: "! sed -f colemakize.sed | sort -k 2r,2.4", and save the result as a new colemak layout.  Voila. :-)

Offline
  • 1
  • Reputation: 1
  • Registered: 26-Feb-2019
  • Posts: 5

I did create a layout meant to be a semi colemak pheonetic layout
For example ו is a very common letter and it can mean ouvw. So I made it o
And ב is a common costant that usually means b, but sometimes v. So I placed that on v
unknown.png
It actually ends up quite close to my other layout
I might actually use this if no actual good layout for hebrew emerges from my efforts
Some other swaps: E evolved from the same letter as ה. Same for ח and h
ש means both s and sh as known. It's also a lot more useful,  Appearing almost 3 times as much in the text I used.
ף should be moved somewhere. IDK where. Or ץ but that is the lesser used letter.

I call this layout "colemak but it destroys your left pinky"

Last edited by Kasher_CS (27-Feb-2019 07:13:01)
Offline
  • 1
  • Reputation: 2
  • Registered: 04-Jul-2019
  • Posts: 5

I have created a variation for Spanish in Latin American version for windows.

Here is my post


1.jpg
2-Shft.jpg
3-AltGr.jpg

Offline
  • 0
  • Reputation: 214
  • From: Viken, Norway
  • Registered: 13-Dec-2006
  • Posts: 5,368

This is nothing more than a "Keep Local Symbols" variant, is it? So please state that in your posts, please.

*** Learn Colemak in 2–5 steps with Tarmak! ***
*** Check out my Big Bag of Keyboard Tricks for Win/Linux/TMK... ***

Offline
  • 0
  • Reputation: 0
  • From: USA
  • Registered: 23-Dec-2024
  • Posts: 6

After looking at the current Colemak-Hebrew layout, I've been thinking that we can perhaps do a little better. Specifically, by taking into account the common Phoenician ancestry between Hebrew and Latin letters, instead of phonetics alone, it's possible to move several common letters to more ergonomic positions while remaining logically close to Latin Colemak. However, I don’t actually know any Hebrew, so I am curious what speakers think.

My proposed mappings, along with justification for each, can be found here: https://pastebin.com/sAg0jGGJ. (Unfortunately, it seems that the forum eats directional formatting characters, so I can’t include it inline.) In total, seven letters are moved from their current positions. Sources used are:

Letter frequency table
Wikipedia: History of the Hebrew alphabet
Wikipedia: Phoenician alphabet

Last edited by Jules-Bertholet (29-Dec-2024 21:35:30)
Attachments:
  • Size: 20.13 KiB
  • Downloads: 16
Offline
  • 0
  • Reputation: 214
  • From: Viken, Norway
  • Registered: 13-Dec-2006
  • Posts: 5,368

The Hebrew layout was developed in conjunction with Hebrew speakers. Background history is largely irrelevant I believe? It's meant to be an aid for Colemak users in writing Hebrew; most of these will want the current language but there are provisions for other needs too.

*** Learn Colemak in 2–5 steps with Tarmak! ***
*** Check out my Big Bag of Keyboard Tricks for Win/Linux/TMK... ***

Offline
  • 0
  • Reputation: 0
  • From: USA
  • Registered: 23-Dec-2024
  • Posts: 6

My proposed Hebrew mappings are not really a downgrade in terms of phonetic match between modern English and Hebrew pronunciation. In fact, in some ways the phonetic correspondence is actually improved. Considering each proposed change in turn:

• ו V → F: a phonetic downgrade, as Waw usually represents a V sound But a small one, as the F sound is just the unvoiced form of V. This proposed change is motivated by ergonomics, as V seems an uncomfortable position for the #2 most common letter. One alternate possibility is to place Waw on both V and F.
• פ ף F → P: Pe can make either an F or a P sound. When Hebrew is written with dagesh marks, P-Pe takes a dagesh, but most of the time these marks are not used. So this change is phonetically neutral.
• ש W → S: Shin makes an S or Sh sound, so phonetically this is an improvement.
• ה H → E: He makes an H sound, but it can also represent the E vowel as a mater lectionis. So phonetically, this is a wash.
• ע E → O: Ayin has nothing to do with either E or O phonetically. So, this is also a wash.
• ס S → X: Samekh makes an S sound, so this is a clear phonetic downgrade, the worst of my proposal. Hence my alternate suggestion of putting Samekh on C (soft C souds like S), displacing the rarer tsadi (which sounds like neither C nor X) to X.
• ח X → H: Heth makes a sound that is much closer to English H than English X. So a phonetic upgrade.

So, to summarize, from a phonetic perspective, my proposal makes two improvements, three neutral changes, and two downgrades (both with suggested mitigations). Overall, I think the changes to the right hand are a clear win, but the ones to the left hand are more dubious/debatable.

———

Also, a note on the dagesh. There are two ways of representing Hebrew letters with dagesh in Unicode: letter + combining dagesh diacritic, or precomposed character. However, unlike most precomposed characters, the Hebrew letters w/ dagesh are considered by Unicode to be compatibility characters, not in NFKC-normal form and intended primarily for compatibility with older software and character sets. The Unicode recommendation for new texts is to use the combining character form. (This also applies to the two dotted Shins.) Unicode spec

Last edited by Jules-Bertholet (30-Dec-2024 23:02:28)
Offline
  • 0