black word/white word encodings

[George]

C with 2 sprained hands.

I would nlike to emulate the t4 encoding scheme using the following

graphic:

* * *
* * *

***** o O o o
hello isn't working; let's ace the 'ello'.





*****
*****
*****
*****
*****
*****
*****
*****
************
************
************
***** ***
***** ***
***** ***
***** ***
***** ***
***** ***

Given that I can suck this in using a twenty by twenty char array, what

would be a sufficient black word/white word encoding scheme.
--
George

The action we take and the decisions we make in this decade will have
consequences far into this century. If America shows weakness and
uncertainty, the world will drift toward tragedy. That will not happen on
my watch.
George W. Bush

Picture of the Day http://apod.nasa.gov/apod/

 

[George]

I don't understand your question. Still, there's some code on my
website that you might find useful, from my chapter of the book "C
Unleashed". There's a source file there that contains a black and
white character set using 16 unsigned shorts for each character.

Is the width of an unsigned short fixed.

See http://jk-technology.com/C_Unleashed/code_list.html

I think I have the identical materials on the disc that comes with the
book. Of course the disc doesn't have the errata, but I don't think there
were any significant ones with this material.

There are also links there to source and header files for making
binary files from text files, using the character set. And for T4
encoding and decoding.

This is as far as I get tonight:

Using the encoding formats on pp. 766-770, let's see what happens to our h,
which is a 20 x 20 that contains:




*****
*****
*****
*****
*****
*****
*****
*****
************
************
************
***** ***
***** ***
***** ***
***** ***
***** ***
***** ***


The first line consists of spaces and maybe a cr and a line feed. Any such
characters are simply not asterisks, ie, they are white. The first line is
empty and therefore contains 20 whites. Following the development, they
would contain 0xFF. We can represent this as 20W.

When we hit the top of the h, we have 3W, 5B, 12W.
Halfway down we have 3W, 12B, 5W.
Then 3W, 5B,4W,3B,5W

Ultimately, we have rows of 20W.

In the development, the blamk lines are translated as
1,728WM, 0WT, EOL

I think the analog is:
20 WT, EOL

and:

0001000 000000000001

I don't see how these words would differ except where that 1 bit is.
Aren't they both padded out with zeroes?


--
George

It's going to be the year of the sharp elbow and the quick tongue.
George W. Bush

Picture of the Day http://apod.nasa.gov/apod/

 

[James Kuyper]

George wrote:
....

Is the width of an unsigned short fixed.

For any given implementation, the width is a fixed value, of at least 16
bits. It can be different on different implementations.

 

[George]

George wrote:
...

For any given implementation, the width is a fixed value, of at least 16
bits. It can be different on different implementations.

And the words for 1000 and 1 would be the same size and have the same
number of zeroes and ones?

Is the length of all these code words the same? If yes, does that mean
they get padded out with zeroes to the left?

Example:

Run length white code word black code word

2 0111 011

Are these not identical except for the third bit, counting from the right.

If there are 16 bits to a word, does not tte former have 13 zeroes while
the latter has 14. If yes, this contradicts claims about EOL on page 771.

--
George

Every nation in every region now has a decision to make. Either you are
with us, or you are with the terrorists.
George W. Bush

Picture of the Day http://apod.nasa.gov/apod/

 

[James Kuyper]

[A lot of stuff I didn't understand, but one question that seemed quite
clear:]
Your question was about 'unsigned short', and so was my answer. Your
further questions either reflect a severe lack of elementary knowledge
about binary representations, or a failure to clearly communicate the
fact that you're actually referring to something other than unsigned
short. I'm favoring the second explanation, and if I'm right I won't be
able to answer your question until you make it clearer.

However, on the off chance that your questions are actually about the
binary representation used in unsigned short, I will answer them in that
context.

And the words for 1000 and 1 would be the same size and have the same
number of zeroes and ones?

unsigned short objects representing 1000 and 1 will take up the same
exact amount of memory.

A unsigned short object representing 1 will have one value bit set to 1,
and the rest of the value bits set to 0; padding bits, if any, might be
set to either 0 or 1.

A unsigned short object representing 1000 will have six value bits set
to 1, and the rest of the value bits set to 0; padding bits, if any,
might be set to either 0 or 1.

Is the length of all these code words the same? If yes, does that mean
they get padded out with zeroes to the left?

All unsigned short objects have the same size. All of the value bits
that don't need to be set to determine the value of the object must be
cleared (0). There could be padding bits, that are not value bits,
though this is rather uncommon; if they are present, they can be either
0 or 1.

 

[Nick Keighley]

And the words for 1000 and 1 would be the same size and have the same
number of zeroes and ones?

Is the length of all these code words the same?  If yes, does that mean
they get padded out with zeroes to the left?

Example:

Run length  white code word black code word

   2          0111               011

Are these not identical except for the third bit, counting from the right..

If there are 16 bits to a word, does not tte former have 13 zeroes while
the latter has 14.  If yes, this contradicts claims about EOL on page 771.

your questions need to become much clearer. For instance

"And the words for 1000 and 1 would be the same size and have the same
number of zeroes and ones?"

what is a "word". What base is 1000 and 1 in? What representation
are you using? Are they strings? Do you mean "the same number of
bits"?

 

[George]

[A lot of stuff I didn't understand, but one question that seemed quite
clear:]
Your question was about 'unsigned short', and so was my answer. Your
further questions either reflect a severe lack of elementary knowledge
about binary representations, or a failure to clearly communicate the
fact that you're actually referring to something other than unsigned
short. I'm favoring the second explanation, and if I'm right I won't be
able to answer your question until you make it clearer.

I've seen this material discussed in mathematics, where we never really get
down to the zeroes and ones.

However, on the off chance that your questions are actually about the
binary representation used in unsigned short, I will answer them in that
context.


unsigned short objects representing 1000 and 1 will take up the same
exact amount of memory.

A unsigned short object representing 1 will have one value bit set to 1,
and the rest of the value bits set to 0; padding bits, if any, might be
set to either 0 or 1.

A unsigned short object representing 1000 will have six value bits set
to 1, and the rest of the value bits set to 0; padding bits, if any,
might be set to either 0 or 1.

The text here is critical, from pp 767-771 of unleashed.

The code words are variable length and strung together one after another
without regard for higher-level boundaries.

Ultimately lines are padded out on the eol word to make--they pad them out
to the left with zeroes--octets, which are unsigned shorts.

All unsigned short objects have the same size. All of the value bits
that don't need to be set to determine the value of the object must be
cleared (0). There could be padding bits, that are not value bits,
though this is rather uncommon; if they are present, they can be either
0 or 1.

Padding with ones?


--
George

When I take action, I'm not going to fire a $2 million missile at a $10
empty tent and hit a camel in the butt. It's going to be decisive.
George W. Bush

Picture of the Day http://apod.nasa.gov/apod/

 

[George]

your questions need to become much clearer. For instance

"And the words for 1000 and 1 would be the same size and have the same
number of zeroes and ones?"

what is a "word". What base is 1000 and 1 in? What representation
are you using? Are they strings? Do you mean "the same number of
bits"?

Clarity isn't my strong suit now. This is new material for me.

By words I mean "code words" of the T.4 protocol. There are 64 terminating
white code words, a like number of black terminating code words, for a
total of 128 terminating code words. Since 1728/64 = 27, there are 27 white
make-up code words and a like number of black, for a total of 54 make-up
code words.

The final code word is the eol word, which has eleven leading zroes, unlike
any other code word. The eol you can pad out, but no other.

For our code to reproduce "h," we have:

%- characters are simply not asterisks, ie, they are white. The first line
is
%- empty and therefore contains 20 whites. Following the development, they
%- would contain 0xFF. We can represent this as 20W.
%-
%- When we hit the top of the h, we have 3W, 5B, 12W.
%- Halfway down we have 3W, 12B, 5W.
%- Then 3W, 5B,4W,3B,5W
%-
%- Ultimately, we have rows of 20W.
%-
%- In the development, the blamk lines are translated as
%- 1,728WM, 0WT, EOL
%-
%- I think the analog is:
%- 20 WT, EOL

This is 0001000 000000000001

3W, 5B, 12 W , eol is
1000 0011 001000 000000000001

3W, 5B,4W,3B,5W, eol is

1000 0011 1011 011 1100 000000000001

20 WT eol is again:
0001000 000000000001

What would be a good tool to use to string these together without a
prohibibitively long line or cr's or lf's?

--
George

To those of you who received honours, awards and distinctions, I say well
done. And to the C students, I say you, too, can be president of the United
States.
George W. Bush

Picture of the Day http://apod.nasa.gov/apod/

 

[CBFalconer]

.... snip ...

Well, this gives me a chance to trot out my latest sig:

Sometimes, when I look at my children, I say to myself:
"Lillian, you should have remained a virgin." -- Lillian Carter
I had a rose named after me, and I was very flattered. But I was
not pleased to read the description in the catalog:
"No good in bed, but fine against a wall.' -- Eleanor Roosevelt

 

[George]

Well, this gives me a chance to trot out my latest sig:

Sometimes, when I look at my children, I say to myself:
"Lillian, you should have remained a virgin." -- Lillian Carter

Lillian is what the South wants to be: better people. Jimmy's stubborn
refusal to not give in to the Bush family ethics makes him unpopular with
most people. Not so with me. The carpenter union card I keep in my pocket
is from local 1 in Chicago. He's unpopular with me for the reasom that
most other folks like him.

His "habitat for humanity," in my book, were better "lawyers for humanity,"
or "data miners for numanity," or "coffee for humanity," butplease not
something that depresses wages for the workers in the vineyards.

§%- By words I mean "code words" of the T.4 protocol. There are 64
terminating
%- white code words, a like number of black terminating code words, for a
%- total of 128 terminating code words. Since 1728/64 = 27, there are 27
white
%- make-up code words and a like number of black, for a total of 54 make-up
%- code words.
%-
%- The final code word is the eol word, which has eleven leading zroes,
unlike
%- any other code word. The eol you can pad out, but no other.
%-
%- For our code to reproduce "h," we have:
%-
%- %- characters are simply not asterisks, ie, they are white. The first
line
%- is
%- %- empty and therefore contains 20 whites. Following the development,
they
%- %- would contain 0xFF. We can represent this as 20W.
%- %-
%- %- When we hit the top of the h, we have 3W, 5B, 12W.
%- %- Halfway down we have 3W, 12B, 5W.
%- %- Then 3W, 5B,4W,3B,5W
%- %-
%- %- Ultimately, we have rows of 20W.
%- %-
%- %- In the development, the blamk lines are translated as
%- %- 1,728WM, 0WT, EOL
%- %-
%- %- I think the analog is:
%- %- 20 WT, EOL
%-
%- This is 0001000 000000000001
%-
%- 3W, 5B, 12 W , eol is
%- 1000 0011 001000 000000000001
%-
%- 3W, 5B,4W,3B,5W, eol is
%-
%- 1000 0011 1011 011 1100 000000000001
%-
%- 20 WT eol is again:
%- 0001000 00000000000


§
Typing with 2 sprained hands and no vicadin

I've decided that I like pain better than being flabby and undone, so I
have forsaken the hydrocodone/apap, and will type until my hands stop
complaining to me. A smallish amount of marijuana sits on the book I
received today, _The Fortran 2003 Handbook_, and a largeish amount of beer
contributes to the heat in my kitchen.


*****
*****
*****
*****
*****
*****
*****
*****
************
************
************
***** ***
***** ***
***** ***
***** ***
***** ***
***** ***

The target is to have one blank space on top and one below. This line is:
0001000 000000000001
§

I think we can economize on such lines but that would be another day.
Today we have:
123456781234567812345678
0001000000000000001

I think Jack wants to say CHAR_BITS, where he now says octets. I don't
mean that to say octets isn't "right." There's a lot of C between pp.
765-771, and *not equating* charbits to eight would have been the bigger of
the weevils.

§

I had a rose named after me, and I was very flattered. But I was
not pleased to read the description in the catalog:
"No good in bed, but fine against a wall.' -- Eleanor Roosevelt

If it were 1938 and the first lady were talking to me about what is fine
against a wall, I'd say she were a sexy bitch.

--
George

We've climbed the mighty mountain. I see the valley below, and it's a
valley of peace.
George W. Bush

Picture of the Day http://apod.nasa.gov/apod/