lojbanic keyboard: Difference between revisions

From Lojban
Jump to navigation Jump to search
mNo edit summary
 
mNo edit summary
Line 1: Line 1:


A '''syllable''' consists of an '''onset''', a non-[[jbocre: Morphology: vowels]]  '''nucleus''', and a '''coda'''. ''(A y-nucleus is excluded from the definition of '''syllable''' just for convenience, because y-syllables are not allowed in [[jbocre: Morphology: f|fu'ivla]]. y-syllables in [[jbocre: Morphology: b|brivla]] always have an empty '''coda'''.)''
QwertAsdf is when I type "Qwert", followed by "Asdf".


An '''onset''' is an [[jbocre: Morphology: consonants nitial]] consonant or cluster, or a [[jbocre: Morphology: vowels lide]], or an [[jbocre: Morphology: consonants]], or empty. ''(An empty onset can only appear at the beginning of a word, an h onset can never appear at the beginning of a word, a glide onset can never appear immediately after a diphthong.)''
It's also what happens when a clueless newbie stumbles upon a WikiWikiWeb and can't resist trying to add a page, for whatever stupid reason.


A '''nucleus''' is a [[jbocre: Morphology: vowels iphthong]], or a [[jbocre: Morphology: vowels owel]], or a [[jbocre: Morphology: vowels]], not followed directly by another '''nucleus'''.
''Now, that's no way to retain newbies! ;-)''


A '''coda''' is a single [[jbocre: Morphology: consonants onsonant]], or empty, not the start of a '''consonantal-syllable''', nor of an [[jbocre: Morphology: consonants nitial]] consonant or cluster.
Rob Speer came up with the following proposal for a Lojban Dvorak keyboard:


A '''consonantal-syllable''' consists of a [[jbocre: Morphology: consonants onsonant]] and a [[jbocre: Morphology: consonants yllabic]] consonant, and must always be followed by another '''consonantal-syllable''' or by an [[jbocre: Morphology: consonants nitial]] consonant or cluster.
English Dvorak:


A '''final-syllable''' (of a [[jbocre: Morphology: b|brivla]]) is a '''syllable''' whose nucleus is not '''stressed''', followed directly by a [[jbocre: Morphology: special characters ause]] or by a [[jbocre: Morphology: words ojban-word]] other than a [[jbocre: Morphology: c|cmene]].
' , . P Y F G C R L / =


A '''stressed-syllable''', a '''stressed-diphthong''' and a '''stressed-vowel''' are, respectively, a '''syllable''', a [[jbocre: Morphology: vowels iphthong]] and a [[jbocre: Morphology: vowels owel]] with where the vowel is '''stressed''' ('A', 'E', 'I', 'O', 'U') or with '''stress''' being marked by being followed (possibly after intervening y and/or consonantal syllables) by just one '''syllable''' and then by a [[jbocre: Morphology: special characters ause]]. ''(The use of capital 'I' and 'U' in [[jbocre: Morphology: vowels lide]] position, or capital consonants, is not taken as a mark of stress.)''
A O E U I D H T N S -


An '''unstressed-syllable''', an '''unstressed-diphthong''' and an '''unstressed-vowel''' are, respectively, a '''syllable''', a [[jbocre: Morphology: vowels iphthong]] and a [[jbocre: Morphology: vowels owel]] not '''stressed''' as described above.
; Q J K X B M W V Z
 
Lojban Dvorak:
 
; / - Y P J R K C D , =
 
A O E U I L ' T N S .
 
Q H Z G X B M F V W
 
''Is Rob's keyboard just trying to minimally rearrange the Dvorak to suit Lojbanists, or is it a true attempt to implement a layout in the Dvorak style that statistically suits Lojban as Dvorak suits English?''
 
* I'm trying to minimally rearrange the Dvorak. It would take much more effort to actually redo Dvorak's research using Lojban heuristics, and ideally if I typed enough in Lojban to want to use this, I'd want it to be easy to switch from Dvorak.
** If you really wanted to redo Dvorak's research for Lojban, and presumably were willing to relearn how to touch type and come up with a keyboard driver which would implement your keyboard, you could use a genetic algorithm and a very large Lojban corpus: [http://www.visi.com/~pmk/evolved.html] --[[jbocre: Adam|Adam]]
 
''Why are you favoring the slash and semicolon?''
 
* I didn't think of it as favoring at the time, but I suppose the keys up there are easier to get to. Very well then - this lets me change from Dvorak even less.
 
; , . Y P J R K C D / =
 
A O E U I L ' T N S -
 
Q H Z G X B M F V W
 
Incidentally, what are the most frequent letters for Lojban gismu (this is an apparently moot issue for the cmavo, as all combinations seem to have been exhausted)?  They wouldn't be the same as English.  Are the consonants fairly evenly distributed, or is there a most common one that can take the place of the T? ''Furthermore, should we count freq. by words it's in, or by frequency in actual use? The former is easier to measure, the latter makes for better keyboard designs.''
 
* The most common letter in gismu is r.  Oddly, though, it's not the most common letter in any given position, which are: s-a-n-n-i, in that order.  Restricting ourselves to CCVCV gismu, only, that becomes s-r-a-n-i.  --gejyspa
 
So I downloaded the keyboard-evolver code, and seeded it with Dvorak, QWERTY, and my layout above.
 
So much for my big ideas - my layout only slightly better than Dvorak in the first round.
 
The results are at: [[jbocre: Evolved Lojban Keyboard|Evolved Lojban Keyboard]]

Revision as of 17:09, 4 November 2013

QwertAsdf is when I type "Qwert", followed by "Asdf".

It's also what happens when a clueless newbie stumbles upon a WikiWikiWeb and can't resist trying to add a page, for whatever stupid reason.

Now, that's no way to retain newbies! ;-)

Rob Speer came up with the following proposal for a Lojban Dvorak keyboard:

English Dvorak:

' , . P Y F G C R L / =

A O E U I D H T N S -

Q J K X B M W V Z

Lojban Dvorak:

/ - Y P J R K C D , =

A O E U I L ' T N S .

Q H Z G X B M F V W

Is Rob's keyboard just trying to minimally rearrange the Dvorak to suit Lojbanists, or is it a true attempt to implement a layout in the Dvorak style that statistically suits Lojban as Dvorak suits English?

  • I'm trying to minimally rearrange the Dvorak. It would take much more effort to actually redo Dvorak's research using Lojban heuristics, and ideally if I typed enough in Lojban to want to use this, I'd want it to be easy to switch from Dvorak.
    • If you really wanted to redo Dvorak's research for Lojban, and presumably were willing to relearn how to touch type and come up with a keyboard driver which would implement your keyboard, you could use a genetic algorithm and a very large Lojban corpus: [1] --Adam

Why are you favoring the slash and semicolon?

  • I didn't think of it as favoring at the time, but I suppose the keys up there are easier to get to. Very well then - this lets me change from Dvorak even less.
, . Y P J R K C D / =

A O E U I L ' T N S -

Q H Z G X B M F V W

Incidentally, what are the most frequent letters for Lojban gismu (this is an apparently moot issue for the cmavo, as all combinations seem to have been exhausted)? They wouldn't be the same as English. Are the consonants fairly evenly distributed, or is there a most common one that can take the place of the T? Furthermore, should we count freq. by words it's in, or by frequency in actual use? The former is easier to measure, the latter makes for better keyboard designs.

  • The most common letter in gismu is r. Oddly, though, it's not the most common letter in any given position, which are: s-a-n-n-i, in that order. Restricting ourselves to CCVCV gismu, only, that becomes s-r-a-n-i. --gejyspa

So I downloaded the keyboard-evolver code, and seeded it with Dvorak, QWERTY, and my layout above.

So much for my big ideas - my layout only slightly better than Dvorak in the first round.

The results are at: Evolved Lojban Keyboard