Letters with diacritic available on wrong letter

[matsbla]

Some letters, like Ɛ, C and Ŋ have their letters with diacritic available on other letters:

I think it is a bit confusing. I guess it would make sense to only have them available on the main letters.

[andjc]

As indicated before, this is temporary. The intention is to 1) test and make fixes to layout 2) copy fixed kmn and layout file to a template that will form the basis of quickly generating language specific layouts 3) strip the additional diacritics from these keys before finalising the keyboard.

[matsbla]

Okay I understand!

Could you point me to where in the code mappings between letters and long-press is? Is it in this file? https://github.com/globalbility/keyman-togo/blob/68cceca2133308741b8482f9b8e5f566b4e27567/gl_togo/source/gl_togo.kmn

Maybe on E you should only get E with diacritic + similar letters to E without diacritic. Then if you will use the similar letters with diacritic (which I guess will be quite rare) you will need to do a second long-press on that letter to get it. Because as I see it it becomes quite crowded and chaotic to get 28 alternatives on long-press.

[andjc]

IN the final version all the open-e combinations will be removed, so TogoKey will be a bit more manageable. Only two langauge specific keyboard layouts do not have open-e so those will be over crowded, but since those langauges do not use any of the E like vwoels we could pare pack longpress for E in those two cases.

It isn't technically possible to have longpres on a character available via longpress

So the two options would be to 1) reduce diacritic inventory for those langauge specific layouts that are over burdened (not making them available), or 2) redesigning the layout to remove all diacritics form long press, and add a special key to access all the cmbining diacritics via longpress, maybe doing a pass in the kmn file to generate precombinded where posisble. Although i think this would be more complex to type with, but could be implemented fairly easily with the current kmn file, would just require the tedious work of redesigning the touch layout.

The kmn file contains the rules ofr generating characters from keys

the gl_togo.keyman-touch-layout file describes the touch layout.

[andjc]

Rules relating to touch layout in kmn file:

• look for stores with "Touch" in the name
• rules with platform("touch") directive
• rules with no platform directive, and their associated stores

Most of the complexity in the keyboard layout results frm two factors: 1) they layout uses deadkeys 2) NFC data is generated where posisble

If we were just using combining diacritics most of the rules in the keyboard layout would disappear.