[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[lojban-beginners] Re: Orthography
Here is a forward of the other orthography thread that I have, which
concerns an orthography based on a seven-segment display.
-Eppcott
Forwarded conversation
Subject: Re: Orthography
------------------------
From: Matt Arnold <matt.mattarn@gmail.com>
Date: Sep 5, 2005 6:30 PM
To: lojban-list@lojban.org
I'm very enthusiastic about the LCD clock alphabet. John, obviously
you've done a great deal of work in attempting to improve Brandon's
orthography. I don't want to re-do that work, but I don't know binary.
Is there a visual reference somewhere that can show the LCD characters
next to the numbers you are using? If not, I'll figure it out and make
a chart of it.
-epkat
----------
From: John E Clifford <clifford-j@sbcglobal.net>
Date: Sep 5, 2005 8:08 PM
To: lojban-list@lojban.org
Trying to do graphics in e-mail is a pain. Lets
do terminology
top (64)
upper left(16) upper right (32)
middle (8)
lower left (2) lower right (4)
bottom (1)
with obvious abbreviations
1:1_
2:2 |
3:1+2 |_
4:4 |
5:1+4 _|
6:2+4 ||
7:1+2+4 |_|
8:8 -
9:1+8 (and now the problems start) parallel
horizontal lines
10:2+8 upper left corner
11:1+2+8 square C
12:4+8 upper right corner
13:1+4+8 square reversed C
14:2+4+8 square cap
15:1+2+4+8 square
16:16 | floating above the line
17: 1+16 16 with a base line separated below it
18:2+16 long left line (numeral 1 - one
possibility)
19:1+2+16 L
20:4+16 high left line, low right
21:1+4+16 left line floating over lower right
corner
22:2+4+16 long left line, short low right
23:1+2+4+16 square reversed J
24:8+16 floating lower left corner
25:1+8+16 24 floating over a base line
26:2+8+16 left T
27:1+2+8+16 inverted F
28:4+8+16 left-to-right zigzag
29:1+4+8+16 inverted square question mark
30:2+4+8+16 square h
31:1+2+4+8+16 square b (one possible numeral 6)
32:32 floating right line
33:1+32 32 over abase line
34:2+32 reverse of 20: low left, high right
35:1+2+32 line floating off from lower left
corner
36:4+32 long right line (usual numeral 1)
37:1+4+32 reversed L
38:2+4+32 short left line, long right
39:1+2+4+32 square J
40:8+32 floating lower right corner
41:1+8+32 40 over a basee line
42:2+8+32 right to left zigzag
43:1+2+8+32 reversed inverted question mark
44:4+8+32 right T
45:1+4+8+32 inverted reversed F
46:2+4+8+32 reversed h
47:1+2+4+8+32 square d
48:16+32 floating parallel lines
49:1+16+32 floating over a base
50:2+16+32 long left floating short right
51:1+2+16+32 right floating over L
52:4+16+32 short floating left, long right
53:1+4+16+32 left floating over reversed L
54:2+4+16+32 parallel long lines
55:1+2+4+16+32 square U
56:8+16+32 floating cup
and so on
I hope this is enough for visualizing
To unsubscribe from this list, send mail to lojban-list-request@lojban.org
with the subject unsubscribe, or go to http://www.lojban.org/lsg2/, or if
you're really stuck, send mail to secretary@lojban.org for help.
----------
From: Brandon Wirick <brandon@yrick.com>
Date: Sep 5, 2005 7:54 PM
To: lojban-list@lojban.org
All right, you've encouraged me to take this seriously and help to
create a truly great digital/script orthography (I'll call it a DSO
from here on out.) I worked some things out and I have some new ideas.
For reference (in hex),
--40--
20--10
--08--
04--02
--01--
I think for the sake of a clear, unambiguous digital readout, the
first of the 0x7F to go would be those that don't span the entire
space horizontally and vertically (0x24 and 0x12 look identical.)
Also, I really like the idea of 0x40 flagging capital letters. I count
23 continous characters with values less than 0x40 (i.e. without the
top bar.) Three of these are horizontally symetrical, leaving ten
pairs of mirrored characters. Conveniently, there are 23 Roman letters
in the Lojban alphabet. These 23 "lower-case" characters correspond to
23 "capitalized" characters with 0x40 added.
I don't have time to draw it up right now, but I would recommend using
0x3E ('H') for {xy}, becuase it just fits; 0x3F (upside-down 'A') for
{my} because that would fit nicely with {py} and {by}, the other
bilabial consonants being assigned to 0x2F ('b') and 0x1F ('d') in
either order; and 0x37 ('U') for {ny} because that would fit nicely
with {ty} and {dy}, the other dental consonants, being assigned to
0x27 (which looks almost like 't') and 0x17 (which looks almost like
'd'). Conveniently, {ly} and {ry} are closer to dental plosives than
any other plosives, so assigning them 0x25 ('L') and 0x13 (backawards
'L') makes sense.
>From there, all characters greater than 0x30, as well as two
characters that are not, go to the vowels, and all the other
consonants need to find whichever leftover pairs make the most sense.
I'm leaning towards making all the unvoiced consonants be greater than
0x20 and all the unvoiced consonants be less than 0x20, except for
{xy}, although it would make {by} look like 'd' and {py} look like
'b'. As for {xy}, if Aleksej is right, even if it accidentally is
pronounced voiced, it would count as {xy}.
There are seven additional continous characters that would not be
continuous if 0x40 were subtracted. These may be used to encode {denpa
bu}, {slaka bu}, {y'y}, and even {ga'e y'y}.
The script will have to take a more convoluted form of the digital
characters, because some of those characters are hard to write with
the fluid motion of the pen. Oh well; that's a totally creative task
and a subject for another e-mail.
mu'omi'e.uirik.
----------
From: Brandon Wirick <brandon@yrick.com>
Date: Sep 5, 2005 8:17 PM
To: lojban-list@lojban.org
Oops, I think I switched 32 with 16 and 4 with 2 from your convention.
Our ideas are down in text, so anyone can refer to them when making a
graphical model, but it's tough to visualize without seeing something.
I'll get something up in a couple days.
----------
From: Matt Arnold <matt.mattarn@gmail.com>
Date: Sep 5, 2005 9:36 PM
To: lojban-list@lojban.org
No doubt this image has other problems, but I tried to visualize a DSO
in which vowels were horizontally symmetrical. Also, in this version I
eliminated the characters made of two unattached vertical lines
because those were a pain to write with. The red characters represent
the original characters which were replaced. Here's the link:
http://www.nemorathwald.com/lojban/lcd_clock_alphabet.gif
Currently I'm drawing a chart of all possible arrangements of segments
that can be in one character. I believe, but am not sure, that they
are in binary code order. After I upload it, I'd like somebody who
understands binary to number them in The Gimp or some such image
program so we can use John's numbering scheme and reference the chart.
-epkat
----------
From: Brandon Wirick <brandon@yrick.com>
Date: Sep 5, 2005 9:59 PM
To: lojban-list@lojban.org
It certainly doesn't suck! Good idea on the vowels. Too bad it's
impossible to do that with only connected characters.
Maybe I can create an applet or something that generates orthographies
from a hex string. That way we can all play around with orthographies
and communicate along these lines:
"Check out this orthography: 4FA39F6..."
"Ooh, I like what you did with {ty} and {dy}, but I think it would be
better if you did this: 4FA3C96..."
I'll get on that soon, or if anyone else is itching to do it, be my guest.
mu'omi'e.uirik.
----------
From: Matt Arnold <matt.mattarn@gmail.com>
Date: Sep 5, 2005 11:43 PM
To: lojban-list@lojban.org
I have attempted to make a chart in which I count in binary using LCD
clock characters.
http://www.nemorathwald.com/lojban/LCD_binary_chart.jpg
Black represents on, blue represents off. There are a hundred and
twenty eight, if you include zero, so that matches John's count. I'm
fairly sure every possible way for it to light up is represented in
this chart. Each of the seven strokes is to represent a digit in a
binary number, and John described the scheme as follows:
top, upper right, upper left, middle, lower right, lower left, bottom
I would have numbered them but I don't know how to count in binary so
I don't know if I've got them in the wrong order. Can somebody type
out a comma-separated or space-separated list numbering them according
to what each represents? For instance, unless I am mistaken, I think
the first row is:
0
and the second row is:
1 2 3 4 5 6 7
...I think. The last row is:
128
Then I'll remake the chart, numbered, and put it online. Then when we
discuss changes to DSO we could just refer to a character by number
and look it up to know what somebody's talking about.
-epkat
----------
From: Matt Arnold <matt.mattarn@gmail.com>
Date: Sep 6, 2005 12:47 PM
To: clifford-j@sbcglobal.net
John,
where did you get this sequence for interpreting an LCD clock digit as
binary? Top, upper right, upper left, middle, lower right, lower left,
bottom. Is there a reference or source somewhere of a project that has
been done in this area before?
-Matt
----------
From: John E Clifford <clifford-j@sbcglobal.net>
Date: Sep 6, 2005 1:01 PM
To: lojban-list@lojban.org
In my system (we really need to get these three
together), the lines are
0
64 16 32 8 2 4 1
80 96 72 66 68 65
48 24 18 20 17
40 34 36 33
10 12 9
6 3
5
112 88 82 84 81
104 98 100 97
74 76 73
70 67
69
56 50 64 49
26 28 25
22 19
21
120 114 116 113
90 92 89
86 83
85
42
41 44
37 35 38
7 13 11 14
106
105 108
101 99 102
71 77 75 78
58
57 60
53 51 54
23 29 27 30
15 39 45 43 46
122
121 124
117 115 118
87 93 91 94
19 103 109 107 110
47 31 55 61 59 62
63 111 91 119 125 123 126
127
So far, suggestions are (Arnold, Brandon1,
Brandon2 (incomplete), Clifford (many arbitrary))
a 62 45 - 59
A 124 109 - 123
b 31 31 47 31
c 43 43 - 43
d 26 26 39 26
e 15 27 - 61
E 79 97 - 125
f 39 42 - 42
g 27 22 - 83
i 7 39 - 39
I 71 103 - 103
j 29 29 - 29
k 45 38 - 101
l 19 19 19 19
m 64 33 63 27
n 30 17 55 45
o 63 46 - 46
O 127 110 - 110
p 47 47 31 47
r 37 37 37 37
s 42 41 - 108
t 44 44 23 44
u 55 23 - 23
U 119 87 - 87
v 23 28 - 28
x 41 34 62 62
y 57 30 - 30
Y 121 94 - 94
z 28 25 - 40
' 25 20 – 16 (32)
, 21 21 - 2 (4)
. 35 35 – 5 (or 3)
1 - - - 36 (18)
2 - - - 107
3 - - - 109
4 - - - 60
5 - - - 93
6 - - - 95
7 - - - 100
8 - - - 127
9 - - - 125
0 - - - 119
----------
From: John E Clifford <clifford-j@sbcglobal.net>
Date: Sep 6, 2005 1:08 PM
To: matt.mattarn@gmail.com
I am sure there is a standard and what I used is
what I remember it as being, but I have no
reference for it -- and memory is for me not a
vlaid source of knowledge. It is easy to imagine
at lest three other systems that work as well and
as "logically."
----------
From: Matt Arnold <matt.mattarn@gmail.com>
Date: Sep 6, 2005 1:26 PM
To: lojban-list@lojban.org
Awesome, thank you! I'll send a URL tonight with the revised chart.
-epkat
----------
From: Matt Arnold <matt.mattarn@gmail.com>
Date: Sep 6, 2005 7:50 PM
To: lojban-list@lojban.org
Here's the complete chart of LCD characters in binary order. I'll make
further revisions if you find errors. -epkat
----------
From: Christopher Zervic <zervic@gmail.com>
Date: Sep 7, 2005 11:37 AM
To: lojban-list@lojban.org
There already exists a standard notation for the segments of the
seven-segment display (7SD), (which is really what we are talking
about, not necessarily LED, LCD or even anything illuminated or
electronic):
A (top)
B (upper right)
C (lower right)
D (bottom)
E (lower left)
F (upper left)
G (middle)
and the '8th segment' = DP (decimal point)
1 = BC
2 = ABDEG
3 = ABCDG
4 = BCFG
5 = ACDFG
6 = ACDEFG
7 = ABC
8 = ABCDEFG
9 = ABCDFG
0 = ABCDEF
So far I like what I read in this thread, although it remains yet to
see whether it proves to be practical, since Lojban already has
perfect orthographic mapping. Nevertheless the suggested alphabet
could draw inspiration from the Shaw Alphabet (q.v.), and be practical
for English as well.
--
Christopher Zervic, Esq.
----------
From: Matt Arnold <matt.mattarn@gmail.com>
Date: Sep 7, 2005 1:18 PM
To: lojban-list@lojban.org
Thanks Christopher. Can you direct me to reference material about this?
I'm especially appreciative for the term 7SD, since I didn't like any
of the phrases I was using to name it.
However, as long as we have a chart which works for us to communicate
with each other while selecting characters, and the numbers we use
will not be included in the finished orthography, I don't think it's
sufficiently important to remake the chart.
-epkat
----------
From: Christopher Zervic <zervic@gmail.com>
Date: Sep 7, 2005 2:06 PM
To: lojban-list@lojban.org
In Wikipedia under Seven-Segment Display is a good place to start. I
do not know of a standard way of representing all of the possible
characters, so the chart does carry some utility.
--
----------
From: John E Clifford <clifford-j@sbcglobal.net>
Date: Sep 8, 2005 11:22 AM
To: lojban-list@lojban.org
Yeah, there is no real need for yet another way
of writing Lojban, even a another way that is
only for Lojban. People have shown how to write
Lojban in just about every known alphabet (I
don't recall a cuneiform version but I expect it
is out there somewhere) and it at least three
more or less developed novel alphabets. No
advantages over the latin alphabet we use has
been demonstrated. But they are intersting
design projects in their own right.
Sorry about misremembering the order of segments
in 7-segment displays. I think that my ordering
came from the ordering in 11- and 15-segment
displays (roughly contemporay with 7 and 14 and
16 and, even more than they obliterated by 7-pin
dot matrices). Those displays had either xs in
each small box or a diamond in the whole (or in
each box for the 15) and the xs were each either
two segments or four. If you drop those box
fillers out the remaining seven lines were -- I
am pretty damned sure -- ordered as I gave it.
Below are (for the record) the equivalences
between the official and this muddled version (I
put the binarization of the official version in
parentheses).
1 D (8) A (64) 64
2 E (4) AB (96) 96
3 DE (12) ABC (112) 100
4 C (16)
----------
From: Matt Arnold <matt.mattarn@gmail.com>
Date: Sep 8, 2005 3:06 PM
To: lojban-list@lojban.org
My personal reason is that I want to decorate faux-ancient artifacts
from an extinct advanced civilization with a strange and alien
advanced language. For instance, I intend to have a little translation
session at the next Penguicon to entice fans of Stargate into Lojban,
in which we decipher a series of glyphs carved in a plaque.
-epkat
----------
From: John E Clifford <clifford-j@sbcglobal.net>
Date: Sep 8, 2005 3:51 PM
To: lojban-list@lojban.org
OK, "no reason intrinsic to the Lojban project."
There are probably several (and this one is
interesting) reason for doing new alphabets,
including for the Hell of it.
----------
From: Brandon Wirick <brandon@yrick.com>
Date: Sep 8, 2005 6:52 PM
To: lojban-list@lojban.org
That is certainly my reason for doing this. I would admit that it is
just flat out fun if that didn't ostracize me from the world, save the
99th percentile of geekdom.