Post by Admin on May 13, 2020 19:47:17 GMT -5
It seems like I have pushed the ADKATT idea about as far as I can push it. There's a number of reasons for that.
1. A very thorough analysis of the 119 Cyrillic languages on Omniglot shows that they do some unusual things, which I didn't expect at first. Some use combining modifiers that you would expect to see in Latin languages, not Cyrillic ones. For instance, there are Cyrillic letters that get Circumflex, Caron, Tilde, Ring Above, Inverted Breve and Dot Above. To me, that is strange, but, that's the way it is. Can't really argue about how people spell their languages. They do what they do.
2. A number of such "specialty-accented" letters are, for all intents and purposes, Latin. For instance, one is C with Acute accent. Now, the people that USE such a letter *might* actually put an Acute on a Cyrillic C, but in practice that doesn't always work so well. The problem is that Cyrillic C doesn't normally take ANY accent (other than the Cyrillic equivalent of C with Cedilla). So, if you try to add an accent to them, a lot of the time it gets misplaced and looks ugly. As long as the font looks right, the most attractive way to do it is to actually insert accented Latin into the otherwise-Cyrillic words. This means inserting Latin into Cyrillic, or adding many different kinds of combining modifiers if you go that way, has to be REALLY easy. The ADKATT design handled *some* of this, but not all of it.
3. It is desirable to have two keyboards that are as similar as possible while having both Latin and Cyrillic that are easy to type. That is because Cyrillic typists may need to type Latin languages too, some have languages (the Serbo-Croatian type of languages is one example) that are commonly typed in both alphabets, and some are going through transitions from Cyrillic to Latin (Kazakh is an example). Such typists may need to insert letters from the "other" alphabet for tutorial purposes, to show "examples" and so on. This very forum shows examples of where I did just that.
4. The ADKATT approach assumed that typists might want ALL of the dead keys to be "hot" to make them easier to type. But to do that, it makes typing ordinary punctuation harder and more error prone.
5. To implement the ADKATT approach, it required the keyboard to utilize what is called a Kana Lock. This is a function normally present in Japanese keyboards. However, when the keyboard is engaged in Kana Lock mode, it seems to make the typing just "disappear" for some reason, when using certain applications in Windows 10. That problem does not appear when Kana is used as a simple "modifier" rather than a "lock". For instance, the Q Keyboard uses a Kana modifier as the CC key, and that usage does not cause problems in Windows 10.
6. To actually HAVE a Kana Lock it requires that some key not used for some other purpose be used for that, without taking away something that someone else was already using in some software or other. In practice, the only key you can do that with and get away with it is the Caps Lock. However, when that is done, it's necessary to type it as a modifier + Caps Lock, such as ALT Caps Lock. If you do that, and if you don't time your keystrokes just right, the keyboard can get "stuck" in Kana Lock or in Caps Lock, and you have to struggle with the keyboard to get it "unstuck". That is just too much to ask of an average user, who would likely see such behavior as "tricky" and "temperamental". Those are undesirable qualities for a keyboard.
Thus, what we are left with now is a keyboard with a lot of conventional dead keys, one that is closer to the Q Keyboard than I originally planned on.
Because these two new keyboards (Latin-based and Cyrillic-based) make it easy to "cross over" to use the "other" alphabet, they effectively "fuse" together two keyboards into one. There will still actually be TWO keyboards, based on which alphabet you need the most, but you can get a great deal of the "other" alphabet when you need it, without having to always switch keyboard layouts in Windows.
For that reason, the keyboards will be called "FUSION" keyboards. There will be a FUSION Latin-Cyrillic and a FUSION Cyrillic-Latin version. Because there is no "Kana Lock" any more, each keyboard operates in a single "mode" with a single set of dead keys.
To accomplish this, each dead key may control up to two alphabets. For instance, the Acute dead key is used both for G with Acute, and for Ѓ with Acute. That requires all dead keys to be "composite" dead keys.
A side effect of having composite dead keys is that adding a combining modifier to a letter is the SAME, because it's the SAME dead key. So, to add a combining Acute, you type AA + apostrophe, then type the Space Bar. That is true REGARDLESS of what kind of letter you want to add it to. It also means that EVERY combining modifier available on the FUSION Latin keyboard is on the FUSION Cyrillic keyboard, because the dead keys and the dead key tables are the SAME. So, "everybody gets everything".
Because there is only a single "mode", we can set aside the CC modifier (CL and CU) for the "other" alphabet. This helps with the Cyrillic version, since that takes up the numeric row for additional letters. By always having the "other" alphabet on CC, we can immediately access the QWERTY digit row and its associated punctuation like @ # $ % etc. That is a convenience regular Cyrillic keyboards don't have, and they force users to type with a keyboard having a numeric keypad. In contrast, they FUSION keyboards give you full access to the complete QWERTY character set even on laptop keyboards that don't have a numeric pad.
So, this seems to be a promising approach. Works continues. I'll keep you informed as things develop.
1. A very thorough analysis of the 119 Cyrillic languages on Omniglot shows that they do some unusual things, which I didn't expect at first. Some use combining modifiers that you would expect to see in Latin languages, not Cyrillic ones. For instance, there are Cyrillic letters that get Circumflex, Caron, Tilde, Ring Above, Inverted Breve and Dot Above. To me, that is strange, but, that's the way it is. Can't really argue about how people spell their languages. They do what they do.
2. A number of such "specialty-accented" letters are, for all intents and purposes, Latin. For instance, one is C with Acute accent. Now, the people that USE such a letter *might* actually put an Acute on a Cyrillic C, but in practice that doesn't always work so well. The problem is that Cyrillic C doesn't normally take ANY accent (other than the Cyrillic equivalent of C with Cedilla). So, if you try to add an accent to them, a lot of the time it gets misplaced and looks ugly. As long as the font looks right, the most attractive way to do it is to actually insert accented Latin into the otherwise-Cyrillic words. This means inserting Latin into Cyrillic, or adding many different kinds of combining modifiers if you go that way, has to be REALLY easy. The ADKATT design handled *some* of this, but not all of it.
3. It is desirable to have two keyboards that are as similar as possible while having both Latin and Cyrillic that are easy to type. That is because Cyrillic typists may need to type Latin languages too, some have languages (the Serbo-Croatian type of languages is one example) that are commonly typed in both alphabets, and some are going through transitions from Cyrillic to Latin (Kazakh is an example). Such typists may need to insert letters from the "other" alphabet for tutorial purposes, to show "examples" and so on. This very forum shows examples of where I did just that.
4. The ADKATT approach assumed that typists might want ALL of the dead keys to be "hot" to make them easier to type. But to do that, it makes typing ordinary punctuation harder and more error prone.
5. To implement the ADKATT approach, it required the keyboard to utilize what is called a Kana Lock. This is a function normally present in Japanese keyboards. However, when the keyboard is engaged in Kana Lock mode, it seems to make the typing just "disappear" for some reason, when using certain applications in Windows 10. That problem does not appear when Kana is used as a simple "modifier" rather than a "lock". For instance, the Q Keyboard uses a Kana modifier as the CC key, and that usage does not cause problems in Windows 10.
6. To actually HAVE a Kana Lock it requires that some key not used for some other purpose be used for that, without taking away something that someone else was already using in some software or other. In practice, the only key you can do that with and get away with it is the Caps Lock. However, when that is done, it's necessary to type it as a modifier + Caps Lock, such as ALT Caps Lock. If you do that, and if you don't time your keystrokes just right, the keyboard can get "stuck" in Kana Lock or in Caps Lock, and you have to struggle with the keyboard to get it "unstuck". That is just too much to ask of an average user, who would likely see such behavior as "tricky" and "temperamental". Those are undesirable qualities for a keyboard.
Thus, what we are left with now is a keyboard with a lot of conventional dead keys, one that is closer to the Q Keyboard than I originally planned on.
Because these two new keyboards (Latin-based and Cyrillic-based) make it easy to "cross over" to use the "other" alphabet, they effectively "fuse" together two keyboards into one. There will still actually be TWO keyboards, based on which alphabet you need the most, but you can get a great deal of the "other" alphabet when you need it, without having to always switch keyboard layouts in Windows.
For that reason, the keyboards will be called "FUSION" keyboards. There will be a FUSION Latin-Cyrillic and a FUSION Cyrillic-Latin version. Because there is no "Kana Lock" any more, each keyboard operates in a single "mode" with a single set of dead keys.
To accomplish this, each dead key may control up to two alphabets. For instance, the Acute dead key is used both for G with Acute, and for Ѓ with Acute. That requires all dead keys to be "composite" dead keys.
A side effect of having composite dead keys is that adding a combining modifier to a letter is the SAME, because it's the SAME dead key. So, to add a combining Acute, you type AA + apostrophe, then type the Space Bar. That is true REGARDLESS of what kind of letter you want to add it to. It also means that EVERY combining modifier available on the FUSION Latin keyboard is on the FUSION Cyrillic keyboard, because the dead keys and the dead key tables are the SAME. So, "everybody gets everything".
Because there is only a single "mode", we can set aside the CC modifier (CL and CU) for the "other" alphabet. This helps with the Cyrillic version, since that takes up the numeric row for additional letters. By always having the "other" alphabet on CC, we can immediately access the QWERTY digit row and its associated punctuation like @ # $ % etc. That is a convenience regular Cyrillic keyboards don't have, and they force users to type with a keyboard having a numeric keypad. In contrast, they FUSION keyboards give you full access to the complete QWERTY character set even on laptop keyboards that don't have a numeric pad.
So, this seems to be a promising approach. Works continues. I'll keep you informed as things develop.