ftp.xmission.com/2014.06.ftp.xmission.com.tar/ftp.xmission.com/pub/lists/fractdev/archive/fractdev.9804

Internet Message Format | 1998-04-27 | 16KB

From: DeRobertis <[e-mail]>
Subject: Re: (fractdev) DiskVid.c vs. The Prototypes File
Date: 04 Apr 1998 15:39:53 -0500

>DeRobertis wrote:
>
>> For XFRACT, DiskVid.c and the file with all the prototypes do not agree
>> with each other.
>
>First of all, how should I address you? I can't parse your name from
>your email address :-)

My full name is Anthony Stephen DeRobertis.

>
>Second, I'm really asleep at the switch. You are doing a Mac port,
>right? How is it going?  I hope you haven't given up yet, lots of
>folks are asking for this.

Its progressing, albeit slowly. There have been a few peoblems over here
(ie, getting a replacement 32MB DIMM). One does not have fun with a 16MB
PowerPC :(

>
>Thirdly, for XFRACT, diskvid.c is not used. We went to a lot of
>trouble under DOS to make disk video modes, and even wroyte our own
>cache. Since UNIX has plenty of memory, Xfractint just makes a big
>array. All the routines are in diskvidu.c.
>
>I will, however, check that the prototypes match. This afternoon I
>got my Xfract version up to 1961 patch 35 with no difficulties.
>
>I did leave the makefile out of the x1961p28.zip file at my FTP site.
>I'll update this file soon.

I'll worry about that as soon as it compiles :(



-
Thanks for using Fractdev, The Fractint Developer's Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractdev"


-------------------------------------------------------------------------------

From: DeRobertis <[e-mail]>
Subject: (fractdev) CMDFILES.c
Date: 05 Apr 1998 04:07:05 -0400

Question 1: What do all these variables which are initialized do?

Can I make my own printer type for the Macintosh printing? If so, which
value should I use?

   Printer_Type = DEFAULT_PRINTER;      /* assume an IBM/EPSON    */
   Printer_Resolution = PRT_RESOLUTION; /* assume low resolution  */

Is resolution in pixels per inch?


Question 2: What is with:

    int getpower10(LDBL x)
        {
            char string[11]; /* space for "+x.xe-xxxx" */
            int p;

    #ifdef USE_LONG_DOUBLE
             sprintf(string,"%+.1Le", x);
    #else
            sprintf(string,"%+.1le", x);
    #endif
              p = atoi(string+5);
            return p;
        }

Is this just returning the exponent field? Must it work negitive numbers?
    When x is positive, what's wrong with
        return(int(ciel(log10(x))); ?
    is it faster to do the above?



-
Thanks for using Fractdev, The Fractint Developer's Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractdev"


-------------------------------------------------------------------------------

From: DeRobertis <[e-mail]>
Subject: (fractdev) XZoomWaiting
Date: 05 Apr 1998 06:10:14 -0400

Yet Another Question:
    What is "XZoomWaiting"?

It seems to be an extern int out of the blue.


Is this some X thing? Could someone provide me a list of these and what
they do?



And does int main() not get called under X? It has a lot of DOS stuff in it...



-
Thanks for using Fractdev, The Fractint Developer's Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractdev"


-------------------------------------------------------------------------------

From: "Tim Wegner" <[e-mail]>
Subject: Re: (fractdev) CMDFILES.c
Date: 06 Apr 1998 22:39:52 -0600

Anthony asked:

> Question 1: What do all these variables which are initialized do?
>
> Can I make my own printer type for the Macintosh printing? If so, which
> value should I use?
>
>    Printer_Type = DEFAULT_PRINTER;      /* assume an IBM/EPSON    */
>    Printer_Resolution = PRT_RESOLUTION; /* assume low resolution  */
>
> Is resolution in pixels per inch?

Fractint's printing facilities are horrible. While as literal a port
as possible is a good idea at first, in this case I'd say you should
implement a completely separate printing facility using MacIntosh
resources. You should have some Mac-only files for this code, and not
put it in the shared files.


> Question 2: What is with:
>
>     int getpower10(LDBL x)
>         {
>             char string[11]; /* space for "+x.xe-xxxx" */
>             int p;
>
>     #ifdef USE_LONG_DOUBLE
>              sprintf(string,"%+.1Le", x);
>     #else
>             sprintf(string,"%+.1le", x);
>     #endif
>               p = atoi(string+5);
>             return p;
>         }
>
> Is this just returning the exponent field?

Yes.

>Must it work negitive numbers?

Good question! In principle yes, although we've not systematically
considered zooming out.

>     When x is positive, what's wrong with
>         return(int(ciel(log10(x))); ?

We are at present still using standard C functions. You can use C++
functions in Mac-specific files, but please don't in shared modules,
unless you are giving up staying synched with Fractint sources.

>     is it faster to do the above?

Speed is a non-issue for this function. It's just a quick and dirty
way to extract the expoenet for various precision estimates. The
function is never used where speed is important.

This whole precision area needs work. It's not easy.

Tim


-
Thanks for using Fractdev, The Fractint Developer's Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractdev"


-------------------------------------------------------------------------------

From: "Tim Wegner" <[e-mail]>
Subject: Re: (fractdev) XZoomWaiting
Date: 06 Apr 1998 22:39:52 -0600


> Yet Another Question:
>     What is "XZoomWaiting"?
>
> It seems to be an extern int out of the blue.
>
>
> Is this some X thing? Could someone provide me a list of these and what
> they do?
>
>
>
> And does int main() not get called under X? It has a lot of DOS stuff in it...

I need help with your X questions. I'm totally ignorant of X.

I would assume that you would not port any of the X-specific code,
but only the shared DOS/UNIX code.

Tim

-
Thanks for using Fractdev, The Fractint Developer's Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractdev"


-------------------------------------------------------------------------------

From: Sylvie Gallet <[e-mail]>
Subject: (fractdev) Colormap bug: a suggestion
Date: 12 Apr 1998 11:33:14 -0400

Hi Tim,

  Could you add the following lines between lines 1002 and 1003 in
Miscovl.c?

                       if (scanc > 4 && scanc < maxcolor-5)
                          if (abs(2*dacbox[scanc][j] - dacbox[scanc-5][j]=

                              - dacbox[scanc+5][j]) >=3D 2)
                             break;

  (Forgive my ignorance, I don't know the difference between
dacbox[scanc][j]  and (int)dacbox[scanc][j]...)

  The loop should look like that:
  =

               for (k =3D 0; k <=3D i; ++k) {
                  for (j =3D 0; j < 3; ++j) { /* check pattern of chg per=

color */
                     delta =3D (int)dacbox[scanc][j] -
(int)dacbox[scanc-k-1][j];
                     if (k =3D=3D scanc - curc)
                        diff1[k][j] =3D diff2[k][j] =3D delta;
                     else
                        if (delta !=3D diff1[k][j] && delta !=3D diff2[k]=
[j]) {
                           diffmag =3D abs(delta - diff1[k][j]);
                           if (diff1[k][j] !=3D diff2[k][j] || diffmag !=3D=
 1)
                              break;
                           diff2[k][j] =3D delta;
                        if (scanc > 4 && scanc < maxcolor-5)
                           if (abs(2*dacbox[scanc][j] - dacbox[scanc-5][j=
]
                               - dacbox[scanc+5][j]) >=3D 2)
                              break;
                           }
                     }
                  if (j < 3) break; /* must've exited from inner loop abo=
ve
*/
                  }

  Cheers,

        - Sylvie

-
Thanks for using Fractdev, The Fractint Developer's Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractdev"


-------------------------------------------------------------------------------

From: "Tim Wegner" <[e-mail]>
Subject: Re: (fractdev) Colormap bug: a suggestion
Date: 12 Apr 1998 14:25:57 -0600

Sylvie wrote:

>   Could you add the following lines between lines 1002 and 1003 in
> Miscovl.c?

Will do. Can you tell me what you are trying to
accomplish with this? Will get back to you shortly.

I have already learned that changing the the line

>  if (diff1[k][j] != diff2[k][j] || diffmag != 1)

to

>  if (diff1[k][j] != diff2[k][j] || diffmag >= 1)

Causes lossless compression with a worsening of compression
ratio.

Tim


-
Thanks for using Fractdev, The Fractint Developer's Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractdev"


-------------------------------------------------------------------------------

From: "Tim Wegner" <[e-mail]>
Subject: Re: (fractdev) Colormap bug: a suggestion
Date: 12 Apr 1998 14:25:57 -0600

Sylvie:

>   Could you add the following lines between lines 1002 and 1003 in
> Miscovl.c?

The result with lyapunov.map is:

colors=000zk0zk0zk0zk0<250>000

which still has some big errors. If you want to see the executable or
see the results of a different test case let me know.

>   (Forgive my ignorance, I don't know the difference between
> dacbox[scanc][j]  and (int)dacbox[scanc][j]...)

There is no difference. The "(int)" explicity converts dacbox[][]
from unsigned char to integer, but C would make that conversion
anyway.

I appreciate your looking at the color map stuff. I seem to have had
an aversion to diving into this code, but I'm overcoming it <g!>

Tim


-
Thanks for using Fractdev, The Fractint Developer's Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractdev"


-------------------------------------------------------------------------------

From: Sylvie Gallet <[e-mail]>
Subject: Re: (fractdev) Colormap bug: a suggestion
Date: 12 Apr 1998 18:41:52 -0400

Tim,

>> The result with lyapunov.map is:
>>
>> colors=3D000zk0zk0zk0zk0<250>000
>>
>> which still has some big errors.

  Nearly the same errors than before!

  Now, maybe I could explain what I was trying to do...  Consider a map i=
n
which the first 32 colors have the following red components:

 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 2 2 2 2 3 3 3 3 4 4 4 4

  the resulting color string is:

  000<30>400...

  and the map becomes:

 0 0 0 0 1 1 1 1 1 1 1 1 2 2 2 2 2 2 2 2 3 3 3 3 3 3 3 3 4 4 4 4

  My idea was to compare the value of color #c with the average of color
#(c-5) and #(c+5), the second line shows the value of
  a =3D abs[2*color#c - color#(c-5) - #(c+5)]

 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 2 2 2 2 3 3 3 3 4 4 4 4
           0 0 0 0 0 0 1 1 1 1 2 0 0 0 1 1 0 0 1 1 0 0

  and to exit the loop when this value is >=3D 2.

  BTW, the first test should be:

                       if (scanc > (curc+4) && scanc < (maxcolor-5))

>> If you want to see the executable or see the results of a different
>> test case let me know.

  Let me know the result with lyapunov.map and with the example above.

>> I appreciate your looking at the color map stuff. I seem to have had =

>> an aversion to diving into this code, but I'm overcoming it <g!>

  It's a bit obscure... <g!>

  Cheers,

        - Sylvie

-
Thanks for using Fractdev, The Fractint Developer's Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractdev"


-------------------------------------------------------------------------------

From: "Tim Wegner" <[e-mail]>
Subject: Re: (fractdev) Colormap bug: a suggestion
Date: 12 Apr 1998 17:49:01 -0600

Sylvie, your suggested change helps with the default palette. I have
sent you:

1. Modified fractint executable. by using debug= you can select your
change or the "perfect" mode or both. See what's new for debug
values.

2. A utility maperrs.exe that shows the exact differences between two
map files, as well as statistics (total rss difference/colors, and
max error per color.

3. Source code for miscovl.c so you can see exactly what I did.

I think I need to make a little map to PAR and back utility for
optimizing the colors= compression algorithm.

Tim

-
Thanks for using Fractdev, The Fractint Developer's Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractdev"


-------------------------------------------------------------------------------

From: "Tim Wegner" <[e-mail]>
Subject: Re: (fractdev) Colormap bug: a suggestion
Date: 12 Apr 1998 17:54:10 -0600

> Date:          Sun, 12 Apr 1998 18:41:52 -0400
> From:          Sylvie Gallet <[e-mail]>
> Subject:       Re: (fractdev) Colormap bug: a suggestion
> To:            [e-mail]
> Reply-to:      [e-mail]

> Tim,
>
> >> The result with lyapunov.map is:
> >>
> >> colors=000zk0zk0zk0zk0<250>000
> >>
> >> which still has some big errors.
>
>   Nearly the same errors than before!
>
>   Now, maybe I could explain what I was trying to do...  Consider a map in
> which the first 32 colors have the following red components:
>
>  0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 2 2 2 2 3 3 3 3 4 4 4 4
>
>   the resulting color string is:
>
>   000<30>400...
>
>   and the map becomes:
>
>  0 0 0 0 1 1 1 1 1 1 1 1 2 2 2 2 2 2 2 2 3 3 3 3 3 3 3 3 4 4 4 4
>
>   My idea was to compare the value of color #c with the average of color
> #(c-5) and #(c+5), the second line shows the value of
>   a = abs[2*color#c - color#(c-5) - #(c+5)]
>
>  0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 2 2 2 2 3 3 3 3 4 4 4 4
>            0 0 0 0 0 0 1 1 1 1 2 0 0 0 1 1 0 0 1 1 0 0
>
Sylvie wrote:

>   BTW, the first test should be:
>
>                        if (scanc > (curc+4) && scanc < (maxcolor-5))

Our mail crossed! I'll change your proposal accordingly and resend
the file.

Tim



-
Thanks for using Fractdev, The Fractint Developer's Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractdev"


-------------------------------------------------------------------------------

From: RBarn0001 <[e-mail]>
Subject: (fractdev) test
Date: 28 Apr 1998 18:32:17 EDT


I have been unable to use my computer most of this month and my mail box
filled up. I am just checking to see if I was knocked of the list because of
returned mail.
Ron Barnett

-
Thanks for using Fractdev, The Fractint Developer's Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractdev"