ftp.xmission.com/2014.06.ftp.xmission.com.tar/ftp.xmission.com/pub/lists/fractint/archive/v01.n100

Internet Message Format | 1998-02-04 | 41KB

From: [e-mail] (fractint-digest)
To: [e-mail]
Subject: fractint-digest V1 #100
Reply-To: fractint-digest
Sender: [e-mail]
Errors-To: [e-mail]
Precedence: bulk


fractint-digest       Thursday, February 5 1998       Volume 01 : Number 100




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

Date: Wed, 04 Feb 1998 11:42:48 -0600
From: "Justin A. Kolodziej" <[e-mail]>
Subject: Re: (fractint) Is this just me, or...?

James Polley wrote:
>
> I hate to disagree, but My copy of netscape is missing some subjects! And it has
> some repeats!
I'm not denying that, but the original complaint was that LookOut
wouldn't read it _at all_!

Enough of this.  I think we know something got munged somewhere.
- --
Justin A. Kolodziej
Why pay for an OS when you can get a clearly superior one for free?
Justin Kolodziej is [e-mail]
Marquette University is www.mu.edu

- -
- ------------------------------------------------------------
Thanks for using Fractint, The Fractals and Fractint Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractint"

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

Date: Wed, 4 Feb 1998 13:54:03 -0700 (MST)
From: Kerry Mitchell <[e-mail]>
Subject: (fractint) comments in formulas and pars

It has been pointed out to me that I was a little verbose in my last
formula submission, so that the important stuff about parameter setting is
not visible in the formula part of the parameters screen (press "z" to
enter parameters and the formula is listed in the lower half of the
screen).  What is the consensus about this?  On the long-winded side, the
comments may help elucidate what's happening, but may be best viewed in a
separate editor.  You can always scroll through the listing with the
control keys and arrow keys.  On the concise side, the formulas are
smaller and the comments don't get in the way of seeing the actual coding.

I have tended to put a short description of the formula, definitions of
the input variables, and suggested parameter values, in the formulas,
along with explanations of what's happening when.  Is this too much?  Just
enough?  Don't care?  Any thoughts or comments?

Kerry

- -------------------------------------------------------------------------------
Kerry Mitchell
[e-mail]
- -------------------------------------------------------------------------------


- -
- ------------------------------------------------------------
Thanks for using Fractint, The Fractals and Fractint Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractint"

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

Date: Wed, 4 Feb 1998 15:59:48 -0500 (EST)
From: [e-mail] (Paul Derbyshire)
Subject: Re: (fractint) Happy Valentine's Day!

Check this out...

HotGoldSpiral      { ; Basedon xlty_julia_sample
                     ; Zoom & colors by PGD
  reset=1960 type=formula formulafile=*.frm formulaname=xlty_jul
  function=round/recip passes=1 center-mag=0.518253/-0.191221/16/1/90
  params=0.2892573747165716/0.01339117893975959/4/1/0.1/0.1 float=y
  maxiter=2000 inside=0 decomp=256 periodicity=0
  colors=000zz0<158>K00K00L00<92>z00 cyclerange=0/255
  }

frm:xlty_jul { ; Kerry Mitchell
        ;
        ; colors by how often x < y, where x & y are the
        ; real & imaginary parts of the fractional part of
        ; the iterate of the Julia set
        ;
        ; p1 = Julia parameter
        ; real(p2) = bailout (try 4)
        ; imag(p2) = color speed (try 1)
        ; p3 = rounding factor (try 1)
        ; fn1 = rounding function, either round() or trunc()
        ; fn2 = arbitrary function of iterate (try log())
        ; calculation uses variable zc, z used for coloring
        ; use decomp=256
        ;
        zc=pixel, c=p1, iter=2, twopi=2*pi, count=0
        maxr=real(p2), speed=imag(p2)*twopi*255/256
        roundfac=p3:
        ;
        ; adjust iterate through arbitrary function and find
        ;   fractional part
        ;
        temp=fn2(zc)/roundfac, frac=roundfac*(temp-fn1(temp))
        ;
        ; update counter if conditions are right
        ;
        if (real(frac)<imag(frac))
          count=count+1
          end if
        ;
        ; bailout
        ;   scale counter into decomposition angle
        ;   set "iteration done" flag (iter = -1)
        ;
        if ((|zc|>maxr)||(iter==maxit))
          angle=speed*count/(iter-1)
          z=cos(angle)+flip(sin(angle))
          iter=-1
        else
          iter=iter+1
          zc=sqr(zc)+c
        endif
        iter>0
        }



- --
    .*.  Friendship, companionship, love, and having fun are the reasons for
 -()  <  life. All else; sex, money, fame, etc.; are just to get/express these.
    `*'  Send any and all mail with attachments to the hotmail address please.
Paul Derbyshire [e-mail] [e-mail]

- -
- ------------------------------------------------------------
Thanks for using Fractint, The Fractals and Fractint Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractint"

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

Date: Wed, 04 Feb 1998 15:14:49 -0600
From: "Damien M. Jones" <[e-mail]>
Subject: Re: (fractint) comments in formulas and pars

Kerry,

 - It has been pointed out to me that I was a little verbose in my last
 - formula submission, so that the important stuff about parameter setting is
 - not visible in the formula part of the parameters screen (press "z" to
 - enter parameters and the formula is listed in the lower half of the
 - screen).  What is the consensus about this?

I don't know about consensus, but how about putting brief descriptions of
the parameters in the first section (which appears on the z screen) and
then longer descriptions and explanations of what goes on further down?
Interested parties will scroll; folks just wanting to plug numbers in will
get terse help on what numbers do what.

Damien M. Jones   \\
[e-mail]  \\  http://www.icd.com/tsd/  (temporary sanity designs)
                    \\  http://www.fractalus.com/ (fractals are my hobby)


- -
- ------------------------------------------------------------
Thanks for using Fractint, The Fractals and Fractint Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractint"

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

Date: Wed, 04 Feb 1998 16:30:02 -0800
From: Peter Jakubowicz <[e-mail]>
Subject: Re: (fractint) comments in formulas and pars

Kerry,

>I have tended to put a short description of the formula, definitions of
>the input variables, and suggested parameter values, in the formulas,
>along with explanations of what's happening when.  Is this too much?

No. I find your commenting your formulas extremely helpful, as well as your
accompanying explanations. Figuring out what some of these uncommented
formulas mean is about as fun as debugging someone's old COBOL code --
altho I suppose one could look at it as a challenge. So, the more verbose
the better I would say, at least in your case, for what that is worth.


- -
- ------------------------------------------------------------
Thanks for using Fractint, The Fractals and Fractint Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractint"

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

Date: Wed, 4 Feb 1998 13:27:08 -0800
From: "Jay Hill"<[e-mail]>
Subject: Re: (fractint) comments in formulas and pars

Hi Kerry,

> On the long-winded side, the comments may help elucidate what's
> happening, but may be best viewed in a separate editor.
> You can always scroll through the listing with the control keys
> and arrow keys.  On the concise side, the formulas are smaller
> and the comments don't get in the way of seeing the actual coding.

You volumn of comments is fine. The only issue I would have is that
the controlkey/arrorkey method of viewing is not well know. I just
looked at the online help in Fractint and did not see it.   I'm
pushing for tighter format in the first 10 lines.  I would like to
see at least a line each for each parameter (that gets us
down to line 7) and one each for the functions (that gets us to line
10). And there are ten available.  The first line should identify the
author
and date.   Next in priority to show up in the ten lines would be items
relating to what can be set in other windows, like X, which must be
known to use the formula correctly.

The really well documented formulas coming to the list have a style
which includes early blank lines, rows of ************ and so on which
I feel are fine down in the body. But up in the quick look window they
only push the important info off the bottom.  Yes, you can scroll, but
this is not well known.  It is after all, only in v19.6.

Jay




- -
- ------------------------------------------------------------
Thanks for using Fractint, The Fractals and Fractint Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractint"

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

Date: Wed, 4 Feb 1998 17:07:20 EST
From: [e-mail]
Subject: Re: (fractint) list volume, pars, attachments, a solution?

In a message dated 2-3-98 11:33:05 PM Central Standard Time,
[e-mail] writes:

<< I actually like the discussion posts. It's a whole lot better to have a
 high list volume and discussion than a cold, impersonal list with two or
 three messages a month. >>

  Yah, I agree with this. I really dislike discussion places with just a bunch
of robots talking. (::Robot voice:: "Here is a for-mu-la I just made." "The i-
ter-a-tion count has too ma-ny whip-pet grom-mets in it." "The prob-lem needs
to be fixed by sum-mon-ing an ex-or-cist for Fract-int." "010011001101101."
:-P) I really prefer to read mail by real people with real emotions and real
thoughts, and not just some random collection of letters, dots, and little @
thingies who might just be computer programs for all I know. :-P

- -
- ------------------------------------------------------------
Thanks for using Fractint, The Fractals and Fractint Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractint"

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

Date: Wed, 4 Feb 1998 14:11:55 -0800
From: "Jay Hill"<[e-mail]>
Subject: Re: (fractint) M-set area: an Idea

Mrvn wrote,

>By the way, what are the upper and lower bounds? And how are they
>calculated?
> I know that it can be done by brute force calculating every pixel and
> counting,

See http://home.earthlink.net/~mrob/-muency-/pixelcounting.html

> but some sort of speedup might be used also.

See the state of the art here (the lower bound is 1.5063036).
http://www.geocities.com/CapeCanaveral/Lab/3825/Period-Area-16.html

> Of course Guessing, tesseral and boundary are out of question,
> because they don't give exact results. But what about distance
> estimation and cyclic checks?

See http://inls.ucsd.edu/y/Complex/area.ps.Z

In 1993 I made this formula for the period 1 component for

z := z^n + c

The component is a cycloid, so the area is an exact lower bound for
the total area.   For large n, the area approaches pi.

Area of period 1 is (n+1)pi/n^((n+1)/(n-1))

An upper bound would be pi.

Jay




- -
- ------------------------------------------------------------
Thanks for using Fractint, The Fractals and Fractint Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractint"

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

Date: Wed, 4 Feb 1998 17:18:18 -0500 (EST)
From: [e-mail] (Paul Derbyshire)
Subject: (fractint) Fractint 19.6 bug!

I found a bug in Fractint 19.6. It looks like it suffers some kind of
internal error when given the following syntactically correct formula:


Explode_M_2 {
  c=pixel
  z=0
  ; Find critical point using Newton's method.
  c2=2*c
  z2=z+2
  s=sqr(z2)
  t=z2*s
  ez=exp(z)
  func=t*ez-c2
  der=3*s*ez+t*ez
  z=z-func/der
  z2=z+2
  s=sqr(z2)
  t=z2*s
  ez=exp(z)
  func=t*ez-c2
  der=3*s*ez+t*ez
  z=z-func/der
  z2=z+2
  s=sqr(z2)
  t=z2*s
  ez=exp(z)
  func=t*ez-c2
  der=3*s*ez+t*ez
  z=z-func/der
  z2=z+2
  s=sqr(z2)
  t=z2*s
  ez=exp(z)
  func=t*ez-c2
  der=3*s*ez+t*ez
  z=z-func/der
  z2=z+2
  s=sqr(z2)
  t=z2*s
  ez=exp(z)
  func=t*ez-c2
  der=3*s*ez+t*ez
  z=z-func/der
  z2=z+2
  s=sqr(z2)
  t=z2*s
  ez=exp(z)
  func=t*ez-c2
  der=3*s*ez+t*ez
  z=z-func/der
  z2=z+2
  s=sqr(z2)
  t=z2*s
  ez=exp(z)
  func=t*ez-c2
  der=3*s*ez+t*ez
  z=z-func/der
  z2=z+2
  s=sqr(z2)
  t=z2*s
  ez=exp(z)
  func=t*ez-c2
  der=3*s*ez+t*ez
  z=z-func/der
  z2=z+2
  s=sqr(z2)
  t=z2*s
  ez=exp(z)
  func=t*ez-c2
  der=3*s*ez+t*ez
  z=z-func/der
  z2=z+2
  s=sqr(z2)
  t=z2*s
  ez=exp(z)
  func=t*ez-c2
  der=3*s*ez+t*ez
  z=z-func/der
  z2=z+2
  s=sqr(z2)
  t=z2*s
  ez=exp(z)
  func=t*ez-c2
  der=3*s*ez+t*ez
  z=z-func/der
  z2=z+2
  s=sqr(z2)
  t=z2*s
  ez=exp(z)
  func=t*ez-c2
  der=3*s*ez+t*ez
  z=z-func/der
  z2=z+2
  s=sqr(z2)
  t=z2*s
  ez=exp(z)
  func=t*ez-c2
  der=3*s*ez+t*ez
  z=z-func/der
  z2=z+2
  s=sqr(z2)
  t=z2*s
  ez=exp(z)
  func=t*ez-c2
  der=3*s*ez+t*ez
  z=z-func/der
  z2=z+2
  s=sqr(z2)
  t=z2*s
  ez=exp(z)
  func=t*ez-c2
  der=3*s*ez+t*ez
  z=z-func/der
  z2=z+2
  s=sqr(z2)
  t=z2*s
  ez=exp(z)
  func=t*ez-c2
  der=3*s*ez+t*ez
  z=z-func/der
  z2=z+2
  s=sqr(z2)
  t=z2*s
  ez=exp(z)
  func=t*ez-c2
  der=3*s*ez+t*ez
  z=z-func/der
  z2=z+2
  s=sqr(z2)
  t=z2*s
  ez=exp(z)
  func=t*ez-c2
  der=3*s*ez+t*ez
  z=z-func/der
  z2=z+2
  s=sqr(z2)
  t=z2*s
  ez=exp(z)
  func=t*ez-c2
  der=3*s*ez+t*ez
  z=z-func/der
  z2=z+2
  s=sqr(z2)
  t=z2*s
  ez=exp(z)
  func=t*ez-c2
  der=3*s*ez+t*ez
  z=z-func/der
  z2=z+2
  s=sqr(z2)
  t=z2*s
  ez=exp(z)
  func=t*ez-c2
  der=3*s*ez+t*ez
  z=z-func/der
  z2=z+2
  s=sqr(z2)
  t=z2*s
  ez=exp(z)
  func=t*ez-c2
  der=3*s*ez+t*ez
  z=z-func/der
  z2=z+2
  s=sqr(z2)
  t=z2*s
  ez=exp(z)
  func=t*ez-c2
  der=3*s*ez+t*ez
  z=z-func/der
  z2=z+2
  s=sqr(z2)
  t=z2*s
  ez=exp(z)
  func=t*ez-c2
  der=3*s*ez+t*ez
  z=z-func/der
  z2=z+2
  s=sqr(z2)
  t=z2*s
  ez=exp(z)
  func=t*ez-c2
  der=3*s*ez+t*ez
  z=z-func/der
  z2=z+2
  s=sqr(z2)
  t=z2*s
  ez=exp(z)
  func=t*ez-c2
  der=3*s*ez+t*ez
  z=z-func/der
  z2=z+2
  s=sqr(z2)
  t=z2*s
  ez=exp(z)
  func=t*ez-c2
  der=3*s*ez+t*ez
  z=z-func/der
  z2=z+2
  s=sqr(z2)
  t=z2*s
  ez=exp(z)
  func=t*ez-c2
  der=3*s*ez+t*ez
  z=z-func/der
  z2=z+2
  s=sqr(z2)
  t=z2*s
  ez=exp(z)
  func=t*ez-c2
  der=3*s*ez+t*ez
  z=z-func/der
  z2=z+2
  s=sqr(z2)
  t=z2*s
  ez=exp(z)
  func=t*ez-c2
  der=3*s*ez+t*ez
  z=z-func/der
  z2=z+2
  s=sqr(z2)
  t=z2*s
  ez=exp(z)
  func=t*ez-c2
  der=3*s*ez+t*ez
  z=z-func/der
  z2=z+2
  s=sqr(z2)
  t=z2*s
  ez=exp(z)
  func=t*ez-c2
  der=3*s*ez+t*ez
  z=z-func/der
  z2=z+2
  s=sqr(z2)
  t=z2*s
  ez=exp(z)
  func=t*ez-c2
  der=3*s*ez+t*ez
  z=z-func/der
  z2=z+2
  s=sqr(z2)
  t=z2*s
  ez=exp(z)
  func=t*ez-c2
  der=3*s*ez+t*ez
  z=z-func/der
  z2=z+2
  s=sqr(z2)
  t=z2*s
  ez=exp(z)
  func=t*ez-c2
  der=3*s*ez+t*ez
  z=z-func/der
  z2=z+2
  s=sqr(z2)
  t=z2*s
  ez=exp(z)
  func=t*ez-c2
  der=3*s*ez+t*ez
  z=z-func/der
  z2=z+2
  s=sqr(z2)
  t=z2*s
  ez=exp(z)
  func=t*ez-c2
  der=3*s*ez+t*ez
  z=z-func/der
  z2=z+2
  s=sqr(z2)
  t=z2*s
  ez=exp(z)
  func=t*ez-c2
  der=3*s*ez+t*ez
  z=z-func/der
  z2=z+2
  s=sqr(z2)
  t=z2*s
  ez=exp(z)
  func=t*ez-c2
  der=3*s*ez+t*ez
  z=z-func/der
  z2=z+2
  s=sqr(z2)
  t=z2*s
  ez=exp(z)
  func=t*ez-c2
  der=3*s*ez+t*ez
  z=z-func/der
  z2=z+2
  s=sqr(z2)
  t=z2*s
  ez=exp(z)
  func=t*ez-c2
  der=3*s*ez+t*ez
  z=z-func/der
  z2=z+2
  s=sqr(z2)
  t=z2*s
  ez=exp(z)
  func=t*ez-c2
  der=3*s*ez+t*ez
  z=z-func/der
  z2=z+2
  s=sqr(z2)
  t=z2*s
  ez=exp(z)
  func=t*ez-c2
  der=3*s*ez+t*ez
  z=z-func/der
  z2=z+2
  s=sqr(z2)
  t=z2*s
  ez=exp(z)
  func=t*ez-c2
  der=3*s*ez+t*ez
  z=z-func/der
  z2=z+2
  s=sqr(z2)
  t=z2*s
  ez=exp(z)
  func=t*ez-c2
  der=3*s*ez+t*ez
  z=z-func/der
  z2=z+2
  s=sqr(z2)
  t=z2*s
  ez=exp(z)
  func=t*ez-c2
  der=3*s*ez+t*ez
  z=z-func/der
  z2=z+2
  s=sqr(z2)
  t=z2*s
  ez=exp(z)
  func=t*ez-c2
  der=3*s*ez+t*ez
  z=z-func/der
  z2=z+2
  s=sqr(z2)
  t=z2*s
  ez=exp(z)
  func=t*ez-c2
  der=3*s*ez+t*ez
  z=z-func/der
  z2=z+2
  s=sqr(z2)
  t=z2*s
  ez=exp(z)
  func=t*ez-c2
  der=3*s*ez+t*ez
  z=z-func/der
  z2=z+2
  s=sqr(z2)
  t=z2*s
  ez=exp(z)
  func=t*ez-c2
  der=3*s*ez+t*ez
  z=z-func/der
  :
  z=exp(z)+c/sqr(z+2),
  real(z)<=900000
}


When this formula is selected from the formula screen, it buzzes and
prints a red error box containing this garbage, which makes no sense to
me, does not seem to refer to any actual error in the formula, and which
must be the manifestation of some kind of a bug.

  Error(7):  Table overflow
    ...(z),func=t*ez-c2,der=3*s*ez+t*ez,z=z-func/der,z2=z+2,s=sqr(z2),t=z...
                               ^

  Any key to continue...


Please issue a bugfix release (19.7? 19.61?) ASDAP correcting this, as
well as the video mode screen bug where setting a new key combo corrupts
fractint.cfg, and the problem with occasional hangs rendering formula
types at 1024x768 and lerger. (I don't really want to wait until 20.0
towork on this formula!)


- --
    .*.  Friendship, companionship, love, and having fun are the reasons for
 -()  <  life. All else; sex, money, fame, etc.; are just to get/express these.
    `*'  Send any and all mail with attachments to the hotmail address please.
Paul Derbyshire [e-mail] [e-mail]

- -
- ------------------------------------------------------------
Thanks for using Fractint, The Fractals and Fractint Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractint"

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

Date: Wed, 4 Feb 1998 17:15:58 EST
From: [e-mail]
Subject: Re: (fractint) comments in formulas and pars

Hi Kerry-

In a message dated 98-02-04 15:56:58 EST, you write:

<< I have tended to put a short description of the formula, definitions of
 the input variables, and suggested parameter values, in the formulas,
 along with explanations of what's happening when.  >>

count me among those who appreciate and value your comments.  I also agree
with the idea that any parameters necessary to make the formula work are
better placed early so they are visible without scrolling.  That will keep
those who haven't spotted the "control - arrow keys" instruction yet from
getting frustrated.  As for the rest, I like having the formula explained in
your step by step fashion.  It is a great help to the less numerate (such as
myself) among us.   Thank you.  Tom

- -
- ------------------------------------------------------------
Thanks for using Fractint, The Fractals and Fractint Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractint"

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

Date: Wed, 4 Feb 1998 17:35:33 EST
From: [e-mail]
Subject: Re: (fractint) Fractint 19.6 bug!

In a message dated 2-4-98 4:19:35 PM Central Standard Time,
[e-mail] writes:

<< When this formula is selected from the formula screen, it buzzes and
 prints a red error box containing this garbage, which makes no sense to
 me, does not seem to refer to any actual error in the formula, and which
 must be the manifestation of some kind of a bug.

   Error(7):  Table overflow
     ...(z),func=t*ez-c2,der=3*s*ez+t*ez,z=z-func/der,z2=z+2,s=sqr(z2),t=z...
                                ^

   Any key to continue... >>

  I think that means it just ran out of memory to use the formula or
something. That formula's kind of big. :-P

- -
- ------------------------------------------------------------
Thanks for using Fractint, The Fractals and Fractint Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractint"

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

Date: Thu, 05 Feb 1998 12:20:07 +1300
From: "Morgan L. Owens" <[e-mail]>
Subject: Re: (fractint) list volume, pars, attachments, a solution?

At 10:17 04/02/98 -0700, Rich Thomson wrote:
>
>In article <[e-mail]> ,
>    "Morgan L. Owens" <[e-mail]>  writes:
>
>> [moderation software accepting archive query requests]
>
>I don't want the moderation software to become too complicated; I'd
>rather use existing tools wherever I can.
>
I'll agree with this. One of the Twelve Networking Truths [RFC1925] states:

(12) In protocol design, perfection has been reached not when there
     is nothing left to add, but when there is nothing left to take
     away.

MLO


- -
- ------------------------------------------------------------
Thanks for using Fractint, The Fractals and Fractint Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractint"

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

Date: Thu, 05 Feb 1998 12:36:30 +1300
From: "Morgan L. Owens" <[e-mail]>
Subject: Re: (fractint) Fractint 19.6 bug!

At 17:35 04/02/98 EST, [e-mail] wrote:
>In a message dated 2-4-98 4:19:35 PM Central Standard Time,
>[e-mail] writes:
>
><<
>
>   Error(7):  Table overflow
>     ...(z),func=t*ez-c2,der=3*s*ez+t*ez,z=z-func/der,z2=z+2,s=sqr(z2),t=z...
>                                ^
>
>   Any key to continue... >>
>
>  I think that means it just ran out of memory to use the formula or
>something. That formula's kind of big. :-P
>
I used to see this error a lot before the formula parser was rewritten for
- - what version was it? 19.3? It was always because I was trying something
too big; trying to rewrite things to make them much shorter helped, though
offhand I can't see much room for abbreviation in your example: perhaps
fewer preiterations? Or how about bundling the preiteration code into the
loop itself, with a flag to switch from the prep code to the loop after a
certain number of iterations? Something that smells like "if
(iterationcount<=preplimit [prep code] else [loop code] endif
iterationcount=iterationcount+1".

I think the error message could be better worded, however. "Memory
overflow" or "Formula too large" perhaps.

MLO



- -
- ------------------------------------------------------------
Thanks for using Fractint, The Fractals and Fractint Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractint"

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

Date: Wed, 4 Feb 1998 16:19:55 -0800
From: "Jay Hill"<[e-mail]>
Subject: Re: (fractint) Fractint 19.6 bug!

Hi Paul,

Try this (while they debug Fractint  :-]  )   which solves for the
critical point another way.  The .2 and .8 are a damping factor
which seems to work well. Convergence is about as fast
as Newton. My formula gets you to about 8 digits in 8 steps.
Newton works better as a polisher, since it does great in the
finish.  It finishes with full accurage in my C++ tests (except
when it blows away my DOS window  :-(.  I have included a
sample view into this space.  Have fun!

Jay

frm:Explode_M_2j   { ; by Jay Hill, 1998
                     ; after Paul Derbyshire
  c=pixel
  z=0  ; eight steps of Jay's formula
z=.2*z+.8*(-2+exp((log(2*c)-z)/3))
z=.2*z+.8*(-2+exp((log(2*c)-z)/3))
z=.2*z+.8*(-2+exp((log(2*c)-z)/3))
z=.2*z+.8*(-2+exp((log(2*c)-z)/3))
z=.2*z+.8*(-2+exp((log(2*c)-z)/3))
z=.2*z+.8*(-2+exp((log(2*c)-z)/3))
z=.2*z+.8*(-2+exp((log(2*c)-z)/3))
z=.2*z+.8*(-2+exp((log(2*c)-z)/3))
; finish with one of Derbyshire's Newton steps
  z2=z+2
  s=sqr(z2)
  t=z2*s
  ez=exp(z)
  func=t*ez-2*c
  der=3*s*ez+t*ez
  z=z-func/der
 :   ; Derbyshire's formula
  z=exp(z)+c/sqr(z+2),
  real(z)<=900000
}

explos             { ; by Jay Hill, 1998
  reset=1960 type=formula formulafile=explode.frm
  formulaname=Explode_M_2j
  center-mag=-8.88178e-015/5.32907e-015/0.05827815 float=y
  maxiter=2000 inside=bof61
  periodicity=0
  colors=@default
  }
explos2            { ; by Jay Hill, 1998
  reset=1960 type=formula formulafile=explode.frm
  formulaname=explode_m_2j center-mag=-7.45891/-0.344906/0.1655629
  float=y maxiter=2000 inside=bof61 periodicity=0 colors=@default
  }



- -
- ------------------------------------------------------------
Thanks for using Fractint, The Fractals and Fractint Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractint"

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

Date: Wed, 04 Feb 1998 16:30:31 -0800
From: Mark Christenson <[e-mail]>
Subject: Re: (fractint) Fractint 19.6 bug!

At 12:36 PM 2/5/98 +1300, Morgan L. Owens wrote:
>At 17:35 04/02/98 EST, [e-mail] wrote:
>>In a message dated 2-4-98 4:19:35 PM Central Standard Time,
>>[e-mail] writes:
>> ...
>I used to see this error a lot before the formula parser was rewritten for
>- what version was it? 19.3? It was always because I was trying something
>too big; trying to rewrite things to make them much shorter helped, though
>offhand I can't see much room for abbreviation in your example: perhaps
>fewer preiterations?
> ...

Maybe it's already been addressed (I'm still relatively new to the list),
but I have to ask an obvious question: is iteration loop support being
planned for the next version of Fractint?


Bud


- -
- ------------------------------------------------------------
Thanks for using Fractint, The Fractals and Fractint Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractint"

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

Date: Wed, 04 Feb 1998 16:39:23 -0800
From: Mark Christenson <[e-mail]>
Subject: Re: (fractint) Fractint 19.6 bug!

I meant *nested* iteration loops. Obviously, Fractint *already*
supports iteration loops!  ;o)

Bud



- -
- ------------------------------------------------------------
Thanks for using Fractint, The Fractals and Fractint Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractint"

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

Date: Wed, 04 Feb 1998 17:53:34 -0700
From: Rich Thomson <[e-mail]>
Subject: Re: (fractint) Fractint 19.6 bug!

In article <[e-mail]> ,
    Mark Christenson <[e-mail]>  writes:
> is iteration loop support being
> planned for the next version of Fractint?

Its generally been decided that making fractint's iteration counter
visible inside formulas is a Good Thing.  Also, there has been talk of
adding do/while for/next style looping constructs to the formula
parser for looping.  Is that the kind of thing you want?
- --
                                               Rich Thomson
                                               [e-mail]

- -
- ------------------------------------------------------------
Thanks for using Fractint, The Fractals and Fractint Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractint"

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

Date: Wed, 4 Feb 1998 17:08:38 -0800
From: "Jay Hill"<[e-mail]>
Subject: Re: (fractint) Fractint 19.6 bug!

Hi Paul,

I wrote:
> Try this  which solves for the critical point another way.

I have run a few more tests. There are multiple roots for
some values of c, and my algorithm does find different
roots sometimes. But that should not be a problem. The
formulation gets a dozen digits or more, your milage
may vary.

Oh, and there is an optimization which should be stuck
back into your formula,  putting c2=2*c.   Note the formula
name change.   I wonder how much optimization Fractint
does like this?

Jay

Explode_M_2J (xaxis){ ; by Jay Hill 1998
                     ; after Paul Derbyshire
  c=pixel, c2=2*c,   z=0
; eight steps of Jay's formula for critical point
z=.2*z+.8*(-2+exp((log(c2)-z)/3))
z=.2*z+.8*(-2+exp((log(c2)-z)/3))
z=.2*z+.8*(-2+exp((log(c2)-z)/3))
z=.2*z+.8*(-2+exp((log(c2)-z)/3))
z=.2*z+.8*(-2+exp((log(c2)-z)/3))
z=.2*z+.8*(-2+exp((log(c2)-z)/3))
z=.2*z+.8*(-2+exp((log(c2)-z)/3))
z=.2*z+.8*(-2+exp((log(c2)-z)/3))
; finish with one of Derbyshire's Newton steps
  z2=z+2
  s=sqr(z2)
  t=z2*s
  ez=exp(z)
  func=t*ez-c2
  der=3*s*ez+t*ez
  z=z-func/der
 :   ; Paul Derbyshire's formula
  z=exp(z)+c/sqr(z+2),
  real(z)<=900000
}

explos3            { ; by Jay Hill 1998
                     ; after Paul Derbyshire
  reset=1960 type=formula formulafile=explode.frm
  formulaname=explode_m_2J center-mag=-2.60858/5.77006/2.143265
  float=y maxiter=2000 inside=bof61 decomp=256 periodicity=0
  colors=000000<28>uISwJTwJT<45>wv1ww0vw1<45\
  >2du0cw0cw<62>mAwmAwlAv<58>111\
  000000000
  }



- -
- ------------------------------------------------------------
Thanks for using Fractint, The Fractals and Fractint Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractint"

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

Date: Thu, 5 Feb 1998 00:20:20 -0500 (EST)
From: [e-mail] (Paul Derbyshire)
Subject: Re: (fractint) Fractint 19.6 bug!

[Odd: I got this reply before I got my copy of the original!]

>  I think that means it just ran out of memory to use the formula or
>something. That formula's kind of big. :-P

Out of memory, my left parietal lobe. I have 32 megabytes EDO in this
sucker, and then there's my swap.

- --
    .*.  Friendship, companionship, love, and having fun are the reasons for
 -()  <  life. All else; sex, money, fame, etc.; are just to get/express these.
    `*'  Send any and all mail with attachments to the hotmail address please.
Paul Derbyshire [e-mail] [e-mail]

- -
- ------------------------------------------------------------
Thanks for using Fractint, The Fractals and Fractint Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractint"

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

Date: Thu, 5 Feb 1998 00:22:04 -0500 (EST)
From: [e-mail] (Paul Derbyshire)
Subject: (fractint) Simplgif...

Well, I finally ran across the infamous simplgif bug myself: a certain
4-image fractmig turned into a bogus GIF89a under simplgif; the file was a
valid gif89a but everything from about line 1000 down was missing.

I yanked the fixed simplgif from Tim's FTP directory on phoenix, and it
worked perfectly on the same image... way to go, Tim!... :-)


- --
    .*.  Friendship, companionship, love, and having fun are the reasons for
 -()  <  life. All else; sex, money, fame, etc.; are just to get/express these.
    `*'  Send any and all mail with attachments to the hotmail address please.
Paul Derbyshire [e-mail] [e-mail]

- -
- ------------------------------------------------------------
Thanks for using Fractint, The Fractals and Fractint Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractint"

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

Date: Thu, 5 Feb 1998 00:33:02 -0500 (EST)
From: [e-mail] (Paul Derbyshire)
Subject: Re: (fractint) Fractint 19.6 bug!

>Try this (while they debug Fractint  :-]  )   which solves for the
>critical point another way.  The .2 and .8 are a damping factor
>which seems to work well. Convergence is about as fast
>as Newton. My formula gets you to about 8 digits in 8 steps.
>Newton works better as a polisher, since it does great in the
>finish.  It finishes with full accurage in my C++ tests (except
>when it blows away my DOS window  :-(.  I have included a
>sample view into this space.  Have fun!

[deletia]

Thanks a bunch! But...how the freaking hell does it work? :-)

- --
    .*.  Friendship, companionship, love, and having fun are the reasons for
 -()  <  life. All else; sex, money, fame, etc.; are just to get/express these.
    `*'  Send any and all mail with attachments to the hotmail address please.
Paul Derbyshire [e-mail] [e-mail]

- -
- ------------------------------------------------------------
Thanks for using Fractint, The Fractals and Fractint Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractint"

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

Date: Thu, 5 Feb 1998 00:41:38 -0500 (EST)
From: [e-mail] (Paul Derbyshire)
Subject: Re: (fractint) Fractint 19.6 bug!

>I wrote:
>> Try this  which solves for the critical point another way.
>
>I have run a few more tests. There are multiple roots for
>some values of c, and my algorithm does find different
>roots sometimes. But that should not be a problem. The
>formulation gets a dozen digits or more, your milage
>may vary.

Multiple roots??
All I did was take the exponential graph and add a bump somewhere west of
the origin. That should be the only critical point; there's also the
asymptotic value 0. Are you 100% sure your computation isn't either
finding preimages (not a problem) or converging incorrectly? (Newton can
sometimes converge to a point or a cycle that's not a root, for degree >=3
polynomials and other functions...)

I don't suppose you can reel off a particular value of c and two or more
of the multiple roots? :-)

>I wonder how much optimization Fractint does like this?

I don't know, and I don't think Tim is telling <vbg>



- --
    .*.  Friendship, companionship, love, and having fun are the reasons for
 -()  <  life. All else; sex, money, fame, etc.; are just to get/express these.
    `*'  Send any and all mail with attachments to the hotmail address please.
Paul Derbyshire [e-mail] [e-mail]

- -
- ------------------------------------------------------------
Thanks for using Fractint, The Fractals and Fractint Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractint"

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

Date: Thu, 5 Feb 1998 01:47:18 EST
From: [e-mail]
Subject: (fractint) Fractint Wish-  mousedriven parameter entry+ fractal preview on 1 screen

Dear Fractinters,

I've been thinking this over for a while - let me know what you think about
it!

                            So many formulas - so little time!

As a new Fractint user working with formulas, I find that a major bottleneck
in obtaining interesting fractals involves finding the optimal parameter
settings (such as real and imaginary p1 values) to enter for a new formula.
Since this is impossible for most people to predict beforehand, much time is
spent typing in "best guess" values on the parameter screen, watching the
fractal render, and shuttling back to the parameter screen (z key) to repeat
the process again and again, an exercise in real patience even with a small
viewing window!  Moreover, being methodical is no guarantee that you'll find
the best parameter settings for neat images. For example, if you enter real p1
values in increments of 0.2, how do you know that you're not missing an abrupt
increase in fractal complexity (ie, an interesting image) between, say, 0.88
and 0.92?  And you can't depend on zero default values to yield a decent image
- - some listed formulas give an empty screen  at p1=0 (eg, REB004A in
fractint.frm).

So, what I propose for a future version of Fractint is a mouse-driven,
streamlined interface with parameter entry and a small fractal preview window
all on the same screen, much like the orbits window (<o> key) mode in the
present version. In orbits mode, moving the cursor with the mouse to a given
point on an escape-time fractal image produces the <orbit> for that point in
the inset orbits window.

In my proposed interface, the monitor screen will be a real vs. imaginary
coordinate grid for parameter p1. Moving the cursor by mouse to a given point
on the screen will yield an on-the-fly rendering of the <fractal> for those p1
values, in a thumbnail window.  Since long formulae can slow down rendering,
this thumbnail window may have to be quite small, but even a 50 x 50 pixel
window should reveal the presence or absence of meaningful fractal detail.
Contrasty color maps will also help in this respect.

The main idea is that by <simply moving the mouse> one could see realtime
changes in fractal structure across a <whole range> of real and imaginary p1
parameter values. All on 1 screen with no keyboard entry!

Additional refinements are possible. Instead of p1, the screen grid could be
changed to represent p2 or p3 values.  p2,p3 parameters could also be changed
by mouse using separate slider controls. Switching between functions (f1(n),
etc.) could still be done in the present way with the left and right cursor
keys. And if one has moved the mouse to a point on the screen which yields
interesting fractal detail, a click on the right mouse button could save the
setting as a par file.

This type of interface would allow a thurough and systematic screening of
parameter settings for a given formula more quickly and efficiently than with
the present setup.  After surveying the fractal "forest" in this manner, one
would have more time to zoom in on fractal "trees" of particular
interest...which is (at least to me) the "fun" part of fractal exploration!

- -
- ------------------------------------------------------------
Thanks for using Fractint, The Fractals and Fractint Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractint"

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

Date: Wed, 4 Feb 1998 23:08:10 -0800
From: "Jay Hill" <[e-mail]>
Subject: Re: (fractint) M-set area: an Idea

Hi group,
Paul D. seems to be off the list, so here is a forward...................

>The component is a cycloid, so the area is an exact lower bound for
>the total area.   For large n, the area approaches pi.
>
>Area of period 1 is (n+1)pi/n^((n+1)/(n-1))
>
>An upper bound would be pi.

Son of a bitch! Because all the components except period 1 shrink with
increasing n, the area for the limit z^infinity+c is pi. (This suggests the
limit of the M-set as n goes to infinity is the closed unit disk.) Now if
the areas of the entire sets follow some scaling law, this could be
extrapolated from pi backwards to get the area of M to >perfect< precision.
One can use the area of the set for z+c, which is 0 (the M-set is the
point c=0), and the estimates for 2, 3, 4, 5, and pi, and match to a
function f(x) that is 0 at 0, approaches pi as x tends to +infinity,and
matches the estimated areas for the appropriate sets at x=2, 3, 4, 5.

- --
    .*.  Friendship, companionship, love, and having fun are the reasons for
 -()  <  life. All else; sex, money, fame, etc.; are just to get/express these.
    `*'  Send any and all mail with attachments to the hotmail address please.
Paul Derbyshire [e-mail] [e-mail]


- -
- ------------------------------------------------------------
Thanks for using Fractint, The Fractals and Fractint Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractint"

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

End of fractint-digest V1 #100
******************************