ftp.xmission.com/2014.06.ftp.xmission.com.tar/ftp.xmission.com/pub/lists/fractint/archive/v01.n075

Internet Message Format | 1998-01-14 | 42KB

From: [e-mail] (fractint-digest)
To: [e-mail]
Subject: fractint-digest V1 #75
Reply-To: fractint-digest
Sender: [e-mail]
Errors-To: [e-mail]
Precedence: bulk


fractint-digest       Thursday, January 15 1998       Volume 01 : Number 075




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

Date: Wed, 14 Jan 1998 17:12:45 -0000
From: Edward Avis <[e-mail]>
Subject: (fractint) RE: Client-server Fractint

>The other way, the cray takes 10 secs for the top corner, which
>is just above its limit and gives the i386 one square, then there is
>no idel comp left and it does the other two squares. Several houhrs
l>ater the i386 is finished with its square and returns its data.

I'm not going to mount a spirited defence of my suggestion - after all it's
only an idea for others to comment on.  But I would point out that the 386
would not take hours - after all it would split up the square it was given
into subsquares, and send some of those to other servers, etc...

Perhaps the Cray would know how fast (or otherwise) the 386 was, and not
send it very much data to process.

Also there's no reason to say that if we're finished sooner than expected,
we can't pre-empt calculation on other machines and do it ourselves.

- --
Ed Avis
[e-mail]
http://members.tripod.com/~mave/index.html




- -
- ------------------------------------------------------------
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: 14 Jan 1998 18:40:54 +0100
From: Brederlow <[e-mail]>
Subject: Re: (fractint) RE: Client-server Fractint

"Morgan L. Owens" <[e-mail]> writes:

> At 14:55 12/01/98 -0000, Edward Avis wrote:
> >>Each server should do a 'scan' to determine the parts of the image it is
> >assigned which are slow. >The sizes of next generation chunks could be
> >determined for each portion using a quad division >algorithm.
> >
> >Yes, that would be a good idea.  It seems to me that a lot of this may just
> >be idle speculation, but it's interesting anyway, so why not:
> >
> >For each square, calculate the top left-hand corner.  If it takes longer
> >than say ten seconds, and depending on how "busy" other servers are, then
> >send off the other three pieces to be calculated separately.  Otherwise,
> >calculate them ourselves.
> >
> >How do we calculate the top left-hand corner?  Easy! Just apply the
> >algorithm above, recursively!  Of course at one point we have got to stop
> >and actually do some calculation, the point at which we do this could be
> >determined according to the complexity of the formula.
> >
> I'm just wondering if there might be some more sophisticated way of
> breaking the task down that will distribute the server workload more
> fairly. I thought this because the approaches that divide the image up into
> chunks ignores any defined symmetries of the fractal. I know that
> multitasking makes up for much of the difference, but we may still be
> calculating up to four times as much as we need to. This could be resolved
> by having the server check for symmetries in the (sub)image it's given to
> compute, and delegating tasks accordingly but I still wonder.

Of cause the symetries should be taken into account before breaking up
the image. Only the unique part of an image is taken and every patch
that is calculated is then mirrored to its symetries and send back to
the client.

> I thought perhaps instead of contiguous blocks an interlaced approach could
> be used, so that the points being calculated by each server are spread
> evenly over the entire image. Kind of subidividing the task upwards instead
> of down. That's just off the top of my head and not really what I'm getting
> at.

Passes, guessing or tesseral could also be used. With passes its clear
what pixel should be calculated, with guessing one would have to do
one complete line of a patch and then start a second server on that
line to do the next guessing pass. The first one could calculate the
second line inbetween. Also the image should be divided into several
smaller patches first so the work can be distributed evenly and one
line doesn't take too long.

My favourite division algorithm (of those done by fractint) is
tesseral. For a distributed algorithm the image should be divided into
smaller patches (depending on its probable time of computation). For
each patch 8 requests (two for each line) should be generated. When
those are computed the patch is eigther subdivided, filled or (if it
gets to small) computed by some fallback method (normaly
guessing). From the time taken for each half line a good aproximation
for the complexity of the sub patches can be derived. For each patch
(except the first) two of the 4 lines will be done already, which give
a good aproximation of the time for the full outline of a patch.

> A more sophisticated multitasking model would require more sophisticated
> protocols: the present idea could be done simply by passing pars/formulas
> and gifs back and forth, while something else may need to carry information
> on orbits back and forth. "I'm up to here... it still needs this done..."
>
> Someone more familiar with the Fractint engine may be able to provide more
> constructive suggestions along these lines than myself.

Have you heard of SOI (Simultanious Orbital Itteration)?
The algorithm computes 9 points for a patch and interpolates the pixel
inbetween. When the interpolation gets worse above a certain limit it
divides the area into 4 patches. The new 9 points for that patch are
taken from the original 9 points (4 of them) and are interpolated (5
of them). When the distortion gets to big (too many subdivision in a
short time) a fallback algorithm (guessing) is used to calculate the
full patch.

If 1-Pass is used as the fallback algorithm every pixel is computed
and one might say that nothing is gained. With guessing as a fallback
some pixels are saved, but normal guessing would have saved those too.

The trick here is that the interpolated pixels are used as the start
of the fallback algorithm. Lets say the SOI did 100 iterations before
a patch must be calculated by the fallback. It then has already done
100 itterations of every single pixel in that patch and the fallback
only has to do the remaining iterations until escape or until
maxiter. Of cause SOI hasen't realy done 100 iterations for every
pixel, but interpolated most of them. At the start several iterations
might be done before subdivision. For each iteration it does there
with the 9 starting points, one iteration for every pixel in the image
(except the 9) is saved. For a 1280x1024 image thats over one million
itterations saved for each one done.

Another nice feature of the algorithm is that it gets better while
zooming in. For the full M-Set the first itteration distorts the image
so much that the patch has to be calculated by the fallback. When
zooming in it can do more and more itterations before subdivision and
it can do more and more subdivisions. For the full M-Set it doesn't
matter that the algorithm does the fallback without saveing anything,
since that only takes seconds, but when the speed is realy needed,
i.e. when deep zooming, it realy kicks in and a great increase in
speed is gained. As an example, xfractint took 4 hours on one image
where Almonbread (using SOI) only took 12 minutes and that was with
doubles only.

The SOI algorithm could be used nicely for distributing patches to
several computers. The amount of calculation until subdivision is
small and in a short time many patches can be created. Also the number
of iterations until division is a good rule to guess the complexity of
a patch.

May the Source be with you.
            Mrvn

- -
- ------------------------------------------------------------
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, 14 Jan 1998 11:02:39 -0600
From: "Damien M. Jones" <[e-mail]>
Subject: Re: (fractint) Re: Tim's_Fract

Angela,

 - Which raises yet another question.....if one needs to do this....reset the
 - decomp for a value other than 0....is there a way to do that in the formula
 - itself? I got the formula from Linda then had to send out an agonized cry
 - for help because I didn't know to reset that option.....can hapless users
 - such as myself be protected??

What I did with my FRM collection was to include a PAR file, with default
settings for each formula.  So rather than press "T", choose "formula", and
pick the formula type, you press "@" and select the formula type from the
list.  This lets me set important parameters like outside=real, float=yes,
periodicity=no, and so on.  And it lets me give a one-line description for
each formula.

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, 14 Jan 1998 11:05:15 -0600
From: "Damien M. Jones" <[e-mail]>
Subject: Re: (fractint) Simplgif update

Guy,

 - PS: Partobat.zip (ver. 3.3) is attached to this mail..

Attachments, even small ones like this, have been discouraged on this list.
 If you want to make a file available to everyone, it is better to put it
on a web page somewhere and provide a URL for it.

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, 14 Jan 1998 11:28:11 -0700
From: Rich Thomson <[e-mail]>
Subject: Re: (fractint) Simplgif update

In article <[e-mail]> ,
    [e-mail] (Guy Marson)  writes:
> But I got one suggestion:
> Is it possible to program SIMPLGIF.EXE to produce a similar output-file (the
> makemig.bat file) than that of Michael Peters PARTOBAT.EXE?
> It will be great to get the possibility to resume the generation of long
> lasting giant fractals..

If there is interest, I can supply a perl script that generates a
batch file to generate an image for every par description in a par
file.  Then you can use this with the par file entries generated by
fractint's b command; the generated batch file supports resuming
exactly where the batch process left off, even saving partially
completed images for later restart.  The basic structure the batch
file uses is something like this:

    if haven't done this image yet
    if we have a partial image
        resume fractint with partial image
    else
        start fractint with par definition
    endif
    endif

I had tried to use partobat, but it just didn't suit my tastes.  So,
I wrote a perl script to generate the batch file that I use.  Perhaps
there are some batch file wizards that could improve the generated
script even more.  I also could have written a perl script to do the
processing directly, but writing out a batch file seemed simpler at the
time.

Perl is available for free from your nearest cpan archive site; see
<ftp://ftp.digital.com/pub/plan/perl/CPAN/ROADMAP.html> for a roadmap
and links to other mirror sites.  Perl is very handy for doing large
batch manipulations.  Using perl and my script, I generated an image
for every single PAR entry in the fractxtra package.
- --
  ``Between stimulus and response is the will to choose.''  -- Steven Covey
 =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
     3D Paint: The Power to Create in 3D;        Rich Thomson
     email me for more info                [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, 14 Jan 1998 11:31:56 -0700
From: Rich Thomson <[e-mail]>
Subject: Re: (fractint) Simplgif update

Oh, another thing I didn't like about partobat is that I couldn't tell
what par was being computed by consulting the bat file.  So my perl
script includes the par file as comments in the batch file.  It
helps!
- --
  ``Between stimulus and response is the will to choose.''  -- Steven Covey
 =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
     3D Paint: The Power to Create in 3D;        Rich Thomson
     email me for more info                [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, 14 Jan 1998 22:38:45 +0100
From: [e-mail] (Guy Marson)
Subject: Re: (fractint) Simplgif update

Hi Damien,

You're right, but I got no homepage, sorry to bother you.. (again)


Guy


>
> - PS: Partobat.zip (ver. 3.3) is attached to this mail..
>
>Attachments, even small ones like this, have been discouraged on this list.
> If you want to make a file available to everyone, it is better to put it
>on a web page somewhere and provide a URL for it.
>
>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"
>


Fungi for fun = FunGUY... @:-]

(Guy Marson, 45b, rue de Bettembourg, L-5810 Hesperange)
(Tel./Fax : [phone])  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, 14 Jan 1998 14:43:23 -0700
From: Rich Thomson <[e-mail]>
Subject: Re: (fractint) Simplgif update

In article <[e-mail]> ,
    [e-mail] (Guy Marson)  writes:
> You're right, but I got no homepage, sorry to bother you.. (again)

You can get a free home page from a variety of places (geocities seems
to be popular), and also there are plenty of people who can make the
file available on their web space for you if you ask.
- --
  ``Between stimulus and response is the will to choose.''  -- Steven Covey
 =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
     3D Paint: The Power to Create in 3D;        Rich Thomson
     email me for more info                [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, 14 Jan 1998 22:47:18 +0100
From: Guy Marson <[e-mail]>
Subject: Re: (fractint) Simplgif update

Oh, thanks for the info!
Will see what I can do!

good night!

Guy

At 14:43 14.01.1998 -0700, you wrote:
>
>In article <[e-mail]> ,
>    [e-mail] (Guy Marson)  writes:
>> You're right, but I got no homepage, sorry to bother you.. (again)
>
>You can get a free home page from a variety of places (geocities seems
>to be popular), and also there are plenty of people who can make the
>file available on their web space for you if you ask.
>--
>  ``Between stimulus and response is the will to choose.''  -- Steven Covey
> =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
>     3D Paint: The Power to Create in 3D;        Rich Thomson
>     email me for more info                [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"
>


- -
- ------------------------------------------------------------
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, 15 Jan 1998 12:39:14 +1300
From: "Morgan L. Owens" <[e-mail]>
Subject: Re: (fractint) RE: Client-server Fractint

At 18:40 14/01/98 +0100, Brederlow wrote:
>"Morgan L. Owens" <[e-mail]> writes:
>
>> At 14:55 12/01/98 -0000, Edward Avis wrote:
>> >>Each server should do a 'scan' to determine the parts of the image it is
>> >assigned which are slow. >The sizes of next generation chunks could be
>> >determined for each portion using a quad division >algorithm.
>> >
>> >
>> I'm just wondering if there might be some more sophisticated way of
>> breaking the task down that will distribute the server workload more
>> fairly.
>
>> I thought perhaps instead of contiguous blocks an interlaced approach could
>> be used, so that the points being calculated by each server are spread
>> evenly over the entire image. Kind of subidividing the task upwards instead
>> of down. That's just off the top of my head and not really what I'm getting
>> at.
>

>My favourite division algorithm (of those done by fractint) is
>tesseral. For a distributed algorithm the image should be divided into
>smaller patches (depending on its probable time of computation). For
>each patch 8 requests (two for each line) should be generated. When
>those are computed the patch is eigther subdivided, filled or (if it
>gets to small) computed by some fallback method (normaly
>guessing). From the time taken for each half line a good aproximation
>for the complexity of the sub patches can be derived. For each patch
>(except the first) two of the 4 lines will be done already, which give
>a good aproximation of the time for the full outline of a patch.
>
One point I was thinking of is when, while one is drawing an image with
tesseral or boundary tracing options with "Fill colour" something other
than normal. With image subdivision regulated by tesseral as suggested
above, this is not a problem, but boundary-traced images could be seriously
fragged. (Yes, I've been known to play Doom, but I think the word is
appropriate in this context!)
Also, boundary traced images are likely to be different anyway, when
computed on a client-server network than when computed as at present,
because islands have a chance of being hit by a subdivision boundary and
hence found by the algorithm. This is not a bad thing in my opinion.

In general, I wonder how useful "accelerated" drawing methods would be in
this environment, except when fill colour is nonnormal.

>> A more sophisticated multitasking model would require more sophisticated
>> protocols: the present idea could be done simply by passing pars/formulas
>> and gifs back and forth, while something else may need to carry information
>> on orbits back and forth. "I'm up to here... it still needs this done..."
>>
>
>Have you heard of SOI (Simultanious Orbital Itteration)?
>
...
>
>The SOI algorithm could be used nicely for distributing patches to
>several computers. The amount of calculation until subdivision is
>small and in a short time many patches can be created. Also the number
>of iterations until division is a good rule to guess the complexity of
>a patch.
>
It would be nice if, instead of computing only subimages, if individual
points or sets of points could be transfered back and forth between
machines in little packets containing coordinates (real and screen), colour
(if known) and any other relevant information. Then, for example with solid
guessing, the server could send out the first-pass points to be computed,
and as they come back (in whatever order) decide for each block whether to
do another pass or not. This would also apply to the SOI algorithm
described and no doubt to others. Communications overhead would become a
serious consideration, but it would smooth traffic out so that it's not a
big lump (subimage) every now and then, while the machine waiting for it
sits idle.

I've been wondering how to distribute the boundary tracing algorithm so
that the result remains faithful to the one-processor case? One idea I had
- - roughly speaking, to start and delegate a trace from every point along
the top boundary; subsequently start a trace from the leftmost uppermost
uncomputed pixel; and ignore repeated patches - resulted in a huge
redundancy of calculation. It would be to some extent alleviated if
individual points were to be transferred, with an "abort" signal (which
would have to be included anyway) whenever two regions were recognised as
being equal.

Obviously, servers will need to multitask. Otherwise someone might choke
the network with a big job such that every server will be waiting for
another server to be freed up so that it can have something to delegate
tasks to. After all, calculating a 1280x1024 image could easily result in
several hundred thousand tasks arranged in a big tree.

It would also be good to take advantage of various properties of the
formula itself. As well as pure symmetry, time could be saved in the case
of images containing FracText-type regions or funny bailout conditions, or
PTC/PHC images. In the first case, when the image is subdivided into
chunks, the formula sent out could be optimised to eliminate cases which
cannot arise in the given subimage. In the second, the subdivision would be
the "interlaced" method I mentioned above, with only the relevant formula
sent to each "subinterlace".

This is getting toward what I mean when I suggested a more sophisticated
subdivision technique; one that is versatile enough to recognise and take
advantage of some of a fractal's peculiarities.

Morgan L. Owens


- -
- ------------------------------------------------------------
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, 15 Jan 1998 12:50:36 +1300
From: "Morgan L. Owens" <[e-mail]>
Subject: (fractint) Client-server Fractint

>
>This is getting toward what I mean when I suggested a more sophisticated
>subdivision technique; one that is versatile enough to recognise and take
>advantage of some of a fractal's peculiarities.
>

Ant, gingerbread, chip, bifurcation, henon, diffusion, cellular, dynamic,
lorenz, ifs, Lsystem, julia_inverse, kamtorus...

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, 14 Jan 1998 17:46:54 -0800
From: Wizzle <[e-mail]>
Subject: Re: (fractint) Re: Tim's_Fract

Damien...

Your formula file is very well organized...and I used your pars as
examples.  But I was looking of a way to do this within the formula....if
it can't be done already it would be a useful feature.....a sort of combo
par/formula. We users have simple needs....we want to make basic things work.

Angela


At 11:02 AM 1/14/98 -0600, you wrote:
>Angela,
>
> - Which raises yet another question.....if one needs to do this....reset the
> - decomp for a value other than 0....is there a way to do that in the
formula
> - itself? I got the formula from Linda then had to send out an agonized cry
> - for help because I didn't know to reset that option.....can hapless users
> - such as myself be protected??
>
>What I did with my FRM collection was to include a PAR file, with default
>settings for each formula.  So rather than press "T", choose "formula", and
>pick the formula type, you press "@" and select the formula type from the
>list.  This lets me set important parameters like outside=real, float=yes,
>periodicity=no, and so on.  And it lets me give a one-line description for
>each formula.
>
>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"
>
>

- -
- ------------------------------------------------------------
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, 14 Jan 1998 17:59:33 -0800
From: Wizzle <[e-mail]>
Subject: Re: (fractint) Simplgif update

Guy....

No homepage??? whyever not??? they are free at geocities and other places

Angela aka wizzle

At 10:38 PM 1/14/98 +0100, you wrote:
>Hi Damien,
>
>You're right, but I got no homepage, sorry to bother you.. (again)
>
>
>Guy
>
>
>>
>> - PS: Partobat.zip (ver. 3.3) is attached to this mail..
>>
>>Attachments, even small ones like this, have been discouraged on this list.
>> If you want to make a file available to everyone, it is better to put it
>>on a web page somewhere and provide a URL for it.
>>
>>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"
>>
>
>
>Fungi for fun = FunGUY... @:-]
>
>(Guy Marson, 45b, rue de Bettembourg, L-5810 Hesperange)
>(Tel./Fax : [phone])  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"
>
>

- -
- ------------------------------------------------------------
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, 14 Jan 1998 22:58:23 -0600
From: "Justin A. Kolodziej" <[e-mail]>
Subject: Re: (fractint) Simplgif update

Wizzle wrote:
>
> Guy....
>
> No homepage??? whyever not??? they are free at geocities and other places
>
> Angela aka wizzle
>
I'm not speaking for Guy, but here's my answer:  Maybe I just don't feel
I have anything worth saying to half a billion people yet.  But that's
just me.

This is getting off topic, so I'll leave it at that.
- --
Justin Kolodziej
10 Base-T + 64 Bytes per Second = 1 Frustrated User
Justin Kolodziej is [e-mail]
Marquette University is [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, 15 Jan 1998 12:37:58 +1
From: "J.P. Louvet" <[e-mail]>
Subject: Re: (fractint) Simplgif update

le 14 Jan 98 a 22:58, Justin A. Kolodziej ecrivait (Justin A. Kolodziej
wrote) :

> I'm not speaking for Guy, but here's my answer:  Maybe I just don't feel
> I have anything worth saying to half a billion people yet.  But that's
> just me.

First before all Guy,
is your program in the Spanky database ? It is the best place.

And a second remark : do most the users of this list have a mailer made by
a Neenderthal programmer ? What are your problems with attached files, and
with the =3D ? Have you no mime decoding ?

Sorry if it is a stupid question...

Jean-Pierre louvet : [e-mail]
Fractal album :
http://graffiti.cribx1.u-bordeaux.fr/MAPBX/louvet/jpl0.html

- -
- ------------------------------------------------------------
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, 15 Jan 1998 07:41:35 -0500
From: Jack Valero <[e-mail]>
Subject: (fractint) Attached Neanderthals with 3D disease

At 12:37 PM 15/01/98 +1, Jean-Pierre louvet wrote:

>And a second remark : do most the users of this list have a mailer made by
>a Neenderthal programmer ? What are your problems with attached files, and
>with the =3D ? Have you no mime decoding ?

Although I've never experienced these problems it appears to be a concern
to many on the list. I believe some time ago it became a general consensus
to avoid attachments- a case of submitting to the lowest common denominator.

But truly, your comment was rather unfair to the Neanderthals. Evidently they
were far too clever to inflict email upon themselves!
Regards - Jack

visit our fractal gallery: http://www.globalserve.net/~jval/phractal.html

- -
- ------------------------------------------------------------
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, 15 Jan 1998 16:04:26 +1
From: "J.P. Louvet" <[e-mail]>
Subject: Re: (fractint) Attached Neanderthals with 3D disease

le 15 Jan 98 a 7:41, Jack Valero ecrivait (Jack Valero wrote) :

> But truly, your comment was rather unfair to the Neanderthals. Evidently
> they were far too clever to inflict email upon themselves! Regards - Jack

No incivility against Neanderthals, but I think that they programmed with
some language machine using a processor made with a piece of silex (or
perhaps they discovered some tools used now on Unix workstations ?). Later
Homo sapiens got Photoshop and began to paint the walls of their caverns.

Clearly out of topic...

Penalty par maybe with 3D disease(my mailer is Mime compliant) :

Nothing            { ; J.P.Louvet
  reset=1960 type=tim's_error function=cosh passes=t
  center-mag=2.18686/-0.820367/4.306632 params=0/0 float=y maxiter=250
  inside=0 potential=255/220/256
  colors=0M2<19>0`40a40`4<29>0D0011<31>5xz<29>054032120552<8>UUJXWL_ZNa`Pc\
  bR<2>jiWlkYml_<4>ushvtjwulxvnywp<3>zxw<2>ywrywpxvnwul<2>trfsqdqobona<2>j\
  iWhgUedS<2>ZZMXWKUUJRRH<7>552B00<11>Q10R10R10<8>Z10Z10_10_20_20<7>`20`20\
  _20<2>Z10Z10Y10X10<7>R10Q10O10<9>C000000E0<8>0L2 cyclerange=0/255 }


Jean-Pierre louvet : [e-mail]
Fractal album :
http://graffiti.cribx1.u-bordeaux.fr/MAPBX/louvet/jpl0.html

- -
- ------------------------------------------------------------
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, 15 Jan 1998 16:07:50 +0100
From: [e-mail] (Guy Marson)
Subject: (fractint) Simplgif update

Hi Tim,

Some more tests showed new problems: My P90 is hanging after loading c. 75%
of the SIMPLGIF.GIF.. into a viewer or into Fractint.
It is a (800x600) (x=5x, y=5x) = 4000x3000 pixel fractal..

This fault is probably *not* comming from bad HD-clusters, I^we been trying
3x to generate the 'simplgif.gif'. The corresponding 'fractmig.gif' is ok ..
no problems to view the fractal with a viewer or in Fractint.

If someone of the other fractinters would like to make the same test, I can
send you the corresponding .ZIP-file (7.1kB)..


cheers,

Guy


Fungi for fun = FunGUY... @:-]

(Guy Marson, 45b, rue de Bettembourg, L-5810 Hesperange)
(Tel./Fax : [phone])  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, 15 Jan 1998 16:20:00 +0100
From: [e-mail] (Guy Marson)
Subject: Re: (fractint) Simplgif update

>le 14 Jan 98 a 22:58, Justin A. Kolodziej ecrivait (Justin A. Kolodziej
>wrote) :
>
>> I'm not speaking for Guy, but here's my answer:  Maybe I just don't feel
>> I have anything worth saying to half a billion people yet.  But that's
>> just me.
>
>First before all Guy,
>is your program in the Spanky database ? It is the best place.

Dear Jean-Pierre, it is not "my" program, it is freeware from Michael
Peters, and, in my opinion, a very usable tool...

>
>And a second remark : do most the users of this list have a mailer made by
>a Neenderthal programmer ?

Hi, hi,hi...

>What are your problems with attached files, and
>with the =3D ? Have you no mime decoding ?

of course, I can switch Eudora to MIME..

>
>Sorry if it is a stupid question...
>
>Jean-Pierre louvet : [e-mail]
>Fractal album :
>http://graffiti.cribx1.u-bordeaux.fr/MAPBX/louvet/jpl0.html
>
>-
>------------------------------------------------------------
>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"
>


Fungi for fun = FunGUY... @:-]

(Guy Marson, 45b, rue de Bettembourg, L-5810 Hesperange)
(Tel./Fax : [phone])  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, 15 Jan 1998 16:20:29 +0100
From: [e-mail] (Guy Marson)
Subject: Re: (fractint) Simplgif update

>le 14 Jan 98 a 22:58, Justin A. Kolodziej ecrivait (Justin A. Kolodziej
>wrote) :
>
>> I'm not speaking for Guy, but here's my answer:  Maybe I just don't feel
>> I have anything worth saying to half a billion people yet.  But that's
>> just me.
>
>First before all Guy,
>is your program in the Spanky database ? It is the best place.

Dear Jean-Pierre, it is not "my" program, it is freeware from Michael
Peters, and, in my opinion, a very usable tool...

>
>And a second remark : do most the users of this list have a mailer made by
>a Neenderthal programmer ?

Hi, hi,hi...

>What are your problems with attached files, and
>with the =3D ? Have you no mime decoding ?

of course, I can switch Eudora to MIME..

>
>Sorry if it is a stupid question...
>
>Jean-Pierre louvet : [e-mail]
>Fractal album :
>http://graffiti.cribx1.u-bordeaux.fr/MAPBX/louvet/jpl0.html
>
>-
>------------------------------------------------------------
>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"
>


Fungi for fun = FunGUY... @:-]

(Guy Marson, 45b, rue de Bettembourg, L-5810 Hesperange)
(Tel./Fax : [phone])  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, 15 Jan 1998 08:55:58 -0800
From: "Jay Hill" <[e-mail]>
Subject: Re: (fractint) Attached Neanderthals with 3D disease

Hi Fractintiers,

I'm trying to locate zeros of a function (an MSet) using Newton's method.

f = a(z^2+c)^2 / (z^2 + b + c)  + c

f' = 2az(z^2+c)(z^2 + 2b + c)/(z^2 + b + c)^2

But z appears in the derivative, and so it is zero everywhere. But I know f
has zeros!   Any ideas Paul?

Jay

Something          { ; Jay Hill (was J. P. Louvet's Nothing)
  reset=1960 type=tim's_error function=cosh passes=t
  center-mag=0.0767767/0/1.0472/1/-90 params=0/0 float=y maxiter=250
  inside=0 potential=255/220/256
  colors=0M2<19>0`40a40`4<29>0D0011<31>5xz<29>054032120552<8>UUJXWL_ZNa`Pc\
  bR<2>jiWlkYml_<3>srfushvtjwulxvn<4>zxw<2>ywrywpxvnwul<2>trfsqdqobona<2>j\
  iWhgUedS<2>ZZMXWKUUJRRH<7>552B00<11>Q10R10R10<8>Z10Z10_10_20_20<7>`20`20\
  _20<2>Z10Z10Y10X10<7>R10Q10O10<9>C000000E0<8>0L2 cyclerange=0/255
  savename=somethin
  }


- ----------
> From: J.P. Louvet <[e-mail]>
> To: [e-mail]
> Subject: Re: (fractint) Attached Neanderthals with 3D disease
> Date: Thursday, January 15, 1998 7:04 AM
>
> le 15 Jan 98 a 7:41, Jack Valero ecrivait (Jack Valero wrote) :
>
> > But truly, your comment was rather unfair to the Neanderthals. Evidently
> > they were far too clever to inflict email upon themselves! Regards - Jack
>
> No incivility against Neanderthals, but I think that they programmed with
> some language machine using a processor made with a piece of silex (or
> perhaps they discovered some tools used now on Unix workstations ?). Later
> Homo sapiens got Photoshop and began to paint the walls of their caverns.
>
> Clearly out of topic...
>
> Penalty par maybe with 3D disease(my mailer is Mime compliant) :
>
> Nothing            { ; J.P.Louvet
>   reset=1960 type=tim's_error function=cosh passes=t
>   center-mag=2.18686/-0.820367/4.306632 params=0/0 float=y maxiter=250
>   inside=0 potential=255/220/256
>   colors=0M2<19>0`40a40`4<29>0D0011<31>5xz<29>054032120552<8>UUJXWL_ZNa`Pc\
>   bR<2>jiWlkYml_<4>ushvtjwulxvnywp<3>zxw<2>ywrywpxvnwul<2>trfsqdqobona<2>j\
>   iWhgUedS<2>ZZMXWKUUJRRH<7>552B00<11>Q10R10R10<8>Z10Z10_10_20_20<7>`20`20\
>   _20<2>Z10Z10Y10X10<7>R10Q10O10<9>C000000E0<8>0L2 cyclerange=0/255 }
>
>
> Jean-Pierre louvet : [e-mail]
> Fractal album :
> http://graffiti.cribx1.u-bordeaux.fr/MAPBX/louvet/jpl0.html
>
> -
> ------------------------------------------------------------
> 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"

- -
- ------------------------------------------------------------
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 #75
*****************************