ftp.xmission.com/2014.06.ftp.xmission.com.tar/ftp.xmission.com/pub/lists/fractint/archive/v01.n036

Internet Message Format | 1997-10-18 | 42KB

From: [e-mail] (fractint Digest)
To: [e-mail]
Subject: fractint Digest V1 #36
Reply-To: [e-mail]
Sender: [e-mail]
Errors-To: [e-mail]
Precedence:


fractint Digest        Sunday, October 19 1997        Volume 01 : Number 036



In this issue:

    Re: (fractint) 80-bit long double floating point
    Re: (fractint) Fractal Geometry of Nature -- $ 7.95 !
    Re: (fractint) DOS vs. Windows (vs. X-Windows?), revisited
    Re: (fractint) DOS vs. Windows (vs. X-Windows?), revisited
    (fractint) RE:  Removing Builtin types...? (was: Integer math)
    (fractint) Version 20
    Re: (fractint) Copyright law
    (fractint) How to pronounce Mandelbrot's name?
    Re: (fractint) How to pronounce Mandelbrot's name?
    Re: (fractint) Why not Java?
    Re: (fractint) DOS vs. Windows (vs. X-Windows?), revisited
    Re: (fractint) Integer math is history, and another modest proposal
    Re: (fractint) Integer math is history, and another modest proposal
    Re: (fractint) How to pronounce Mandelbrot's name?
    Re: (fractint) How to pronounce Mandelbrot's name?
    Re: (fractint) Why not Java?
    Re: (fractint) DOS vs. Windows (vs. X-Windows?), revisited
    (fractint) test - ignore
    Re: (fractint) test - ignore
    Re: (fractint) Copyright law
    (fractint) Integer math is history, and another modest proposal
    Re: (fractint) Integer math is history, and another modest propo
    Re: (fractint) Integer math is history, and another modest propo
    (fractint) if you are removed from the list
    Re: (fractint) Integer math is history, and another modest propo
    Re: (fractint) Copyright law

See the end of the digest for information on subscribing to the fractint
or fractint-digest mailing lists and on how to retrieve back issues.

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

Date: Sat, 18 Oct 1997 01:11:17 -0400 (EDT)
From: [e-mail] (Paul Derbyshire)
Subject: Re: (fractint) 80-bit long double floating point

>
>OK fair enough.
>So 80 bit numbers are acceptable.
>Does DJGPP allow them?

It does indeed. This will compile without problems:

int decorator;
long shot;
double or_nothing;
long double barreled_shotgun_for_blowing_up_Imps;
...


And the long double does have the full 80 bits of precision in
calculating, storing and retrieving.


- --
    .*.  Where feelings are concerned, answers are rarely simple [GeneDeWeese]
 -()  <  When I go to the theater, I always go straight to the "bag and mix"
    `*'  bulk candy section...because variety is the spice of life... [me]
Paul Derbyshire [e-mail], http://chat.carleton.ca/~pderbysh

- ------------------------------------------------------------
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: Sat, 18 Oct 1997 01:19:35 -0400 (EDT)
From: [e-mail] (Paul Derbyshire)
Subject: Re: (fractint) Fractal Geometry of Nature -- $ 7.95 !

>
>Is the the book by Mandelbrot?
>I would Love to have a copy.
>Is there a way to ge tit instead of off the net?
>Alas. I'm not hooked up to the internet.

You're not?? Then my attempts to receive messages through telepathy has
succeeded!

:-)

- --
    .*.  Where feelings are concerned, answers are rarely simple [GeneDeWeese]
 -()  <  When I go to the theater, I always go straight to the "bag and mix"
    `*'  bulk candy section...because variety is the spice of life... [me]
Paul Derbyshire [e-mail], http://chat.carleton.ca/~pderbysh

- ------------------------------------------------------------
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: Sat, 18 Oct 1997 01:23:49 -0400 (EDT)
From: [e-mail] (Paul Derbyshire)
Subject: Re: (fractint) DOS vs. Windows (vs. X-Windows?), revisited

Try running that fucker in pure DOS (command prompt only from boot menu)
and see how Windows stacks up then!

Do you have any IDEA what happens to a DOS app running on Winblows, when
it doesn't have the focus, and idle sensitivity is at max and a Windows
app is also using a lot of cpu?


- --
    .*.  Where feelings are concerned, answers are rarely simple [GeneDeWeese]
 -()  <  When I go to the theater, I always go straight to the "bag and mix"
    `*'  bulk candy section...because variety is the spice of life... [me]
Paul Derbyshire [e-mail], http://chat.carleton.ca/~pderbysh

- ------------------------------------------------------------
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: Sat, 18 Oct 1997 01:44:29 -0500
From: "Justin A. Kolodziej" <[e-mail]>
Subject: Re: (fractint) DOS vs. Windows (vs. X-Windows?), revisited

Paul Derbyshire wrote:

> Try running that **** in pure DOS (command prompt only from boot menu)
> and see how Windows stacks up then!

Been there, done that, got the results (in case you missed them):

Fractint Integer under Windows:  26 min 40.47 sec
Fractint Float under Windows:  26 min 43.87sec
WinFrac Integer: 12 min 17.59 sec
WinFrac Float:  12 min 79.98 sec
FractInt Integer mode when booted "Safe Mode Command Prompt Only":  8 min 17.68
sec
FractInt Float mode when booted "Safe Mode Command Prompt Only": 8 min 5.43 sec
FractInt Integer mode when booted "Command Prompt Only":  11 min 16.03 sec
FractInt Float mode when booted "Command Prompt Only": 11 min 17.89 sec

And watch your language from now on.  As it says on the Fractint Wish List home
page, "Fractals are a family occupation," and I'm sure no one on this mailing list
appreciates your comments in that letter.  You can use that kind of language when
you E-mail Microsoft, but please don't use it here.  I've had more than my share
of that in my time,  and my use of that A-word was very questionable at best as
well.

Justin K.

Bad(?) Linux slogans:
- -- Linux: The OS for truly independent spirits! --
- -- Linux: The OS of champions! --
- -- Linux: Marquette's Web server (www.mu.edu) uses it, shouldn't you? --
- -- Linux: Because it's free --
- -- Linux: Because you can --
- -- Install Linux: Tell Microsoft where to stick it! --
Justin Kolodziej 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: Sat, 18 Oct 1997 03:33:17 -0400
From: "Robin Y. Millette" <[e-mail]>
Subject: (fractint) RE:  Removing Builtin types...? (was: Integer math)

Hello!
    I hope this one will not generate mail only to prove different =
remailers act differently... Now with the topic of the day:

- -----Original Message-----
From:    Lee Skinner [SMTP:[e-mail]]
Sent:    vendredi, 17 octobre 1997 23:35
To:    [e-mail]
Subject:    (fractint) Integer math

<<Tim>>Now another modest proposal. While we are cutting and slashing =
code, what
would you think about cutting out all built-in fractal types that can be
implemented with a formula in the formula parser? We could include a
special formula file with all the removed built-in types. <<Tim>>

<<Lee>>Modest??<g> I have many thousands of images implemented with the =
built-in
types. Sometimes I like to revisit some of them and zoom in on them. =
Could
some mechanism be devised to somehow declare equivalence of the deleted
built-in types with their formulas so that the images would not have to =
be
regenerated?<<Lee>>

I don't see how that would be so hard to implement as suggested, =
although I haven't been thru the code... Maybe I should do that, and get =
involved in the coding while I've got djgpp sitting on my hd :)

But <g>, how much space/time are we talking about here? Is it worth the =
trouble? Ints? yeah since their a great deal of code replication without =
much (if any) speed improvements... Is "applying" a parsed formula =
slower then a hard coded one? If so, could that be improved as well? =
Maybe a "specific" gccfractint? (of gccfract..?) to work within? I'm =
just guessing here, maybe getting ahead of myself ..?

What are your thoughts on this?
(And please, try to make your subject meaningful!)

CIAO!

'   '         '       '     '   '     '       '         '   '
Robin Y. Millette                            ICQ uin: 1266281
Waglo Institution         http://www.generation.net/~millette
Answer the Bovine Call!   DIFT: http://imail.org/Digidome-rc5


- ------------------------------------------------------------
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: Sat, 18 Oct 1997 03:42:21 -0400
From: "Robin Y. Millette" <[e-mail]>
Subject: (fractint) Version 20

Hello again :)

    Two quickies this time: when are we hoping to see the new version, and are they plans for a new name?


- ------------------------------------------------------------
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: Sat, 18 Oct 1997 06:44:53 -0700 (PDT)
From: Jon Noring <[e-mail]>
Subject: Re: (fractint) Copyright law

[e-mail] (Ian J Kaplan) wrote:
>[e-mail] wrote:

>> It applies. Several of us fractal artists have been involved in a suit with
>> someone using images without permission. The case was settled out of court,
>> but the take home lesson is that intellectual properties laws, including
>> copyright, do apply. That's why many of us include copyright notices in our
>> par files.
>> Ron Barnett

>   Oh, I never intended to suggest that copyright didn't apply. I'm just
>looking for clarification of exactly what is copyrightable and what isn't...

As I said before, IANAL, but I think common sense with some lay knowledge of
IP law tells me the current thinking that using somebody else's PAR file to
regenerate and use without modification an image without permission (implicit
or explicit) is a copyright violation.  And the PAR file itself can, I
believe, be protected by copyright so if I reproduce somebody else's PAR file
without permission, for example in a book, it would probably be considered by
the courts to be a copyright violation totally separate from the image it
represents.  However, the information contained in the PAR file *cannot* be
copyrighted -- it is only the tangible expression of the information which
can be copyrighted.  (The same goes for formula files -- the information is
not copyrightable -- mathematical formulas cannot be copyrighted (I believe,
but they might be patentable!), but the tangible expression of the formulas
is -- everything in copyright law boils down to "tangible expression".)

But taking somebody else's PAR file, and then modifying it so as to make a
"new work" image which is different enough from the original (what constitutes
a "new work" is something for the courts to decide), would be considered Fair
Use.  After all, it is recognized in the philosophy of copyright that
no work truly stands alone, but has been influenced by and stands atop other
works, which is one of the reasons for the Fair Use allowance since the
purpose of copyright is to "promote the arts and sciences" and having too
draconian of a copyright law would not meet this fundamental Constitutional
and simply common sense requirement.

In my perspective, substantially changing the color map so as to make the image
*clearly look different* than the original may be allowed by the courts (but
editing one color to improve an image's appearance but otherwise leaving the
general image look similar would probably be a copyright violation.)  And
the same could be said of zooming and coordinates -- small changes to enhance
the image may be a violation, but significantly changing the zoom/coords to
create a clearly new work, even with the same color map, may be allowable.
The bottom line is -- is it a "new work", or is it "Memorex".  :^)

I don't think the courts have tackled fractal or similar images yet but when
they do, it is important that we of the fractal art community make sure that
we educate the court as to the process and nuances of creating fractal images,
otherwise the courts may set incorrect precedents -- for example, the
Communications Decency Act (CDA) was thrown out partly because those opposing
the CDA *made sure* the courts fully understood the many nuances of the
Internet which make it unique, and from this understanding the courts were
able to correctly apply Constitutional principles.  As it stood, the CDA was
drafted by those who had *no idea* of what the Internet was, how it worked,
and its various nuances.  They do now.

Just a lay perspective.  If you have real questions on Copyright law, go find
a good IP attorney and don't rely on what I wrote.

Jon Noring

_____________________________________________________________________________
 OmniMedia Digital Publishing | E-Books: http://www.awa.com/library/omnimedia
 9671 S. 1600 West St.        | Digital/Fractal Art: (coming soon!)
 South Jordan, UT 84095       |
 801-253-4037                 | 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: Sat, 18 Oct 1997 11:06:39 -0500
From: "Justin A. Kolodziej" <[e-mail]>
Subject: (fractint) How to pronounce Mandelbrot's name?

Just a very quick (and hopefully not controversial) question:  I'm doing
a speech on fractals for my Public Speaking class in college, and I'm
not sure how to pronounce Benoit Mandelbrot's first name.  Is it Ben-wah
like it probably would be in French, or as it is spelled in English?
(probably a dumb question, since he's French) <g>

Justin K.

Bad(?) Linux slogans:
- -- Linux: The OS for truly independent spirits! --
- -- Linux: The OS of champions! --
- -- Linux: Marquette's Web server (www.mu.edu) uses it, shouldn't you? --

- -- Linux: Because it's free --
- -- Linux: Because you can --
- -- Install Linux: Tell Microsoft where to stick it! --
Justin Kolodziej 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: Sat, 18 Oct 1997 14:17:06 -0700
From: Peter Jakubowicz <[e-mail]>
Subject: Re: (fractint) How to pronounce Mandelbrot's name?

> (probably a dumb question, since he's French) <g>
>
Can't help you with the pronunciation (tho I've only heard it pronounced
Ben-wah), but if you're giving a talk on him you might be interested to
know he was born in Warsaw in 1924 and didn't move to Paris until 1936
(see Gleick's Chaos, p. 87); but I have no idea what country he
considers himself to be of, France or Poland.

                    Peter

- ------------------------------------------------------------
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: Sat, 18 Oct 1997 14:26:21 -0600
From: "Tim Wegner" <[e-mail]>
Subject: Re: (fractint) Why not Java?

Paul opined:

> No offense but this JAVA/C++ debate is really spamming up the mailing
> list. I have what appears to be roughly 150K of "Why not Java?" in my
> mbox, making the mail system slow as hell and causing disk shortage. Gak.

I agree. Discussion of languages for Fractint or fractal programs is
an appropriate subject, but this has gone overboard. I'm not saying
the subject is entirely off topic, but I'd like to ask list members
to show restraint on this subject so we don't overload the list.

Tim





- ------------------------------------------------------------
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: Sat, 18 Oct 97 16:56:50
From: [e-mail] (Dr I D Entwistle)
Subject: Re: (fractint) DOS vs. Windows (vs. X-Windows?), revisited

On Fri 17 Oct 97 (21:17:23), [e-mail] wrote:
>Justin wrote:
>
>> I hate to bring this subject up again, but I noticed that in the
>whole
>> discussion on killing off the DOS version of Fractint, no real
>numbers
>> indicating speed were mentioned.
  Some real data for the discussion on the future demise of Integer Fractint.
The default 640x480 pixel Fractint  M set image generated on my Cyrix 100 Mhz
5x86 cpu(32 Mb RAM) which is used as a second processor on an Acorn RISC PC
with mode SF4  under DOS 6.22 is displayed in 1.48 secs(Integer) and 1.82 secs
(Float). When Fractint is run from a Windows 95 window the times are
2.09(Integer) and 2.41(Float). As I mainly use Fractint for following the
Fractint field and running FOTD I would continue to use v 19.6 for integer
until I change processors . For generating  my own M set fractals I use the
main cpu of my computer, a Dec Strongarm 200 Mhz SA110 32 Bit RISC processor.
The 64 bit integer application I use gives the Fractint default image in 0.25
secs and can generate all the M  sets upto the 64 bit limit in less than 3
minutes. I certainly support the idea that the loss of the integer facility in
Fractint is worth it to add new facilities and I look forward to downloading v
20.
      Quite the worst piece of fractint  data I have come across is my default
time for the 640x480 M set image using Winfract 18.2 under Windows 95. Both the
Integer and Float images took 120 seconds!!!
  Ian

- --
 __  __  __  __      __ ___   _____________________________________________
|__||__)/ __/  \|\ ||_   |   /
|  ||  \\__/\__/| \||__  |  /...Internet access for all Acorn RISC machines
___________________________/ [e-mail] URL
http://www.argonet.co.uk/users/ian.ent





- ------------------------------------------------------------
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: Sat, 18 Oct 1997 19:18:35 -0300 (ADT)
From: mctaylor <[e-mail]>
Subject: Re: (fractint) Integer math is history, and another modest proposal

On Fri, 17 Oct 1997, Tim Wegner wrote:

> what would you think about cutting out all built-in fractal types
> that can be implemented with a formula in the formula parser? We

1) Speed? Wouldn't it be a massive slowdown
2) Lee's concerns are valid
3) Why bother? Just to make it smaller? Any other benefits?
4) Mandelbrot would be left I assume. ;)

- --
Michael C. Taylor <[e-mail]> <http://www.mta.ca/~mctaylor/>
 Programmer, Computing Services, Mount Allison University, Canada
 sci.fractals FAQ <http://www.mta.ca/~mctaylor/sci.fractals-faq/>
    fractal and cryptography archive <http://fractal.mta.ca/>



- ------------------------------------------------------------
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: Sat, 18 Oct 1997 20:46:03 -0400 (EDT)
From: [e-mail]
Subject: Re: (fractint) Integer math is history, and another modest proposal

In a message dated 97-10-17 22:27:38 EDT, you write:

<< Now another modest proposal. While we are cutting and slashing code,
 what would you think about cutting out all built-in fractal types
 that can be implemented with a formula in the formula parser? We
 could include a special formula file with all the removed built-in
 types.
  >>
Tim,
The built-in types have some bailout parameter options. If the built-in types
were removed, could the bailout options (e.g. manh, manr) be included?
Ron Barnett

- ------------------------------------------------------------
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: Sat, 18 Oct 1997 21:04:00 -0400
From: "Brock Kevin Nambo" <[e-mail]>
Subject: Re: (fractint) How to pronounce Mandelbrot's name?

- -----Original Message-----
From: Justin A. Kolodziej <[e-mail]>
Date: Saturday, October 18, 1997 12:16 PM


>Just a very quick (and hopefully not controversial) question:  I'm doing
>a speech on fractals for my Public Speaking class in college, and I'm
>not sure how to pronounce Benoit Mandelbrot's first name.  Is it Ben-wah
>like it probably would be in French, or as it is spelled in English?


While you're answering that, could someone tell me how to pronounce his last
name too?
You miss things when you only have the Internet to do them over :(

>>BKNambo
- --
- -----------------------------------------------------------------
| H     [e-mail] (:P McQ)|  UIN: 1936556   ____    __|
|=@==== http://come.to/brocks.place |afw?*********** /_   \  / /|
| H H H "World Domination Through Trivia" -S3Kitties  / /\ \/ / |
| H H H Marcher -- Just my imagination -- Nightwatch /_/  \__/  |
- -----------------------------------------------------------------



- ------------------------------------------------------------
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: Sat, 18 Oct 1997 20:18:52 -0500
From: "Justin A. Kolodziej" <[e-mail]>
Subject: Re: (fractint) How to pronounce Mandelbrot's name?

> While you're answering that, could someone tell me how to pronounce his last
> name too?
> You miss things when you only have the Internet to do them over :(
>
> >>BKNambo

Mandelbrot is just as it looks: Man-del-brot.

Bad(?) Linux slogans:
- -- Linux: The OS for truly independent spirits! --
- -- Linux: The OS of champions! --
- -- Linux: Marquette's Web server (www.mu.edu) uses it, shouldn't you? --
- -- Linux: Because it's free --
- -- Linux: Because you can --
- -- Install Linux: Tell Microsoft where to stick it! --
Justin Kolodziej 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: Sat, 18 Oct 1997 20:19:27 -0500
From: "Justin A. Kolodziej" <[e-mail]>
Subject: Re: (fractint) Why not Java?

Tim Wegner wrote:

> Paul opined:
>
> > No offense but this JAVA/C++ debate is really spamming up the mailing
> > list. I have what appears to be roughly 150K of "Why not Java?" in my
> > mbox, making the mail system slow as hell and causing disk shortage. Gak.
>
> I agree. Discussion of languages for Fractint or fractal programs is
> an appropriate subject, but this has gone overboard. I'm not saying
> the subject is entirely off topic, but I'd like to ask list members
> to show restraint on this subject so we don't overload the list.
>
> Tim

It's my fault.  I posted the original message, and expected one or two
replies.  Instead, I got tons.  We've determined that Java sucks for fractals
(at least for now), so I hope we can give the topic a rest for a while.

Bad(?) Linux slogans:
- -- Linux: The OS for truly independent spirits! --
- -- Linux: The OS of champions! --
- -- Linux: Marquette's Web server (www.mu.edu) uses it, shouldn't you? --

- -- Linux: Because it's free --
- -- Linux: Because you can --
- -- Install Linux: Tell Microsoft where to stick it! --
Justin Kolodziej 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: Sat, 18 Oct 1997 20:36:18 -0500
From: "Justin A. Kolodziej" <[e-mail]>
Subject: Re: (fractint) DOS vs. Windows (vs. X-Windows?), revisited

> <lengthy data description snipped>
>       Quite the worst piece of fractint  data I have come across is my default
> time for the 640x480 M set image using Winfract 18.2 under Windows 95. Both the
> Integer and Float images took 120 seconds!!!
>   Ian

120 seconds????? Did you say 120???? At 640x480????  For the entire Mandelbrot at
150 iterations that you get when you start Fractint?
On my Pentium Pro 180, integer takes .66 sec and float takes .83 sec. Wait... let's
turn on pixel-by-pixel update... now integer takes 36.03 sec and float takes 36.30
sec.  Well, this _might_ explain it, but it should take about 1 minute, not 2,
unless Cyrix chips are broken or something.  Pixel-by-pixel update DOES slow things
down, but not that severely!

Bad(?) Linux slogans:
- -- Linux: The OS for truly independent spirits! --
- -- Linux: The OS of champions! --
- -- Linux: Marquette's Web server (www.mu.edu) uses it, shouldn't you? --
- -- Linux: Because it's free --
- -- Linux: Because you can --
- -- Install Linux: Tell Microsoft where to stick it! --
Justin Kolodziej 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: 19 Oct 1997 12:13:49 +1000
From: [e-mail] (Graham Daveney)
Subject: (fractint) test - ignore

This is a test please ignore

- ------------------------------------------------------------
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: Sun, 19 Oct 1997 00:33:52 -0400 (EDT)
From: [e-mail] (Paul Derbyshire)
Subject: Re: (fractint) test - ignore

>This is a test please ignore

This is a test follow-up, please ignore.


- --
    .*.  Where feelings are concerned, answers are rarely simple [GeneDeWeese]
 -()  <  When I go to the theater, I always go straight to the "bag and mix"
    `*'  bulk candy section...because variety is the spice of life... [me]
Paul Derbyshire [e-mail], http://chat.carleton.ca/~pderbysh

- ------------------------------------------------------------
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: Sat, 18 Oct 1997 22:05:09 -0700
From: Andrew Schoonmaker <[e-mail]>
Subject: Re: (fractint) Copyright law

At 10:53 10/16/97 -0600, you wrote:
>
>In article <[e-mail]> ,
>    Andrew Schoonmaker <[e-mail]>  writes:
>> A couple of questions...  One, what do you mean by "period 16" and "period
>> 64"?
>
>I'll take a stab at it... each "bud" in M is a basin of attraction.
>Inside each "bud" is a point c, where z := z^2 + c is a cycle.  The
>length of the cycle is the period referred to; we can label the entire
>basin of attraction with the period of this point c.  The main cardiod
>is period 1, the bud along the real axis to the left of the main
>cardiod is period 2, the next bud to the left of that along the real
>axis is period 4, and so on.  The period doubling as you progress down
>the real line can be viewed in fractint by looking at the bifurcation
>fractal type.  The bifurcation fractal type is a visualization of the
>dynamics of M along the real line.
>
>Clearer or muddier now? :)

Oh, that's about what I thought you meant.  Doesn't Fractint have some
internal coloring scheme which displays this?

 -Andrew (who's started reading Metamagical Themas...)
- --
Andrew Schoonmaker ([e-mail])
"This is not a complete.  Sentence.  This either."
 -David Moser

- ------------------------------------------------------------
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: Sun, 19 Oct 1997 04:29:17 -0400
From: Thierry Boudet <[e-mail]>
Subject: (fractint) Integer math is history, and another modest proposal

        Hello Fractal World !

> Version 20 will NOT support integer math,

        are beta available for bug hunting ?

> what would you think about cutting out all built-in fractal types
> that can be implemented with a formula in the formula parser?

        what about speed in "Built-in Vs. Formulas" ?

        Best regards.   Thierry.



- ------------------------------------------------------------
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: Sun, 19 Oct 1997 11:09:17 -0600
From: "Tim Wegner" <[e-mail]>
Subject: Re: (fractint) Integer math is history, and another modest propo

Thierry asked:

>         are beta available for bug hunting ?

At the moment we don't distribute betas in a possibly vain
attempt to keep the developers' lives simple <g!>

>         what about speed in "Built-in Vs. Formulas" ?

This might be a good project for the benchmark lovers on the list.
Generally, formula types are faster than built-in types. The
mandelbrot may be an exception because it is implemented in
assembler. If you use the debug=90 you will see that the mandelbrot
slows up - this makes the mandelbrot use C code typical of most of
the built-in types rather than assembler.

However, we're not going to jump in and remove built-in types without
a lot more thought., if ever. I was just throwing out the idea to see
what the reaction would be, and got some good feedback. When you
implement a built-in type using a formula, the formula has to reflect
the same bailout option as the builtin type. Perhaps we need a
standard way to embed the bailout options in formulas. We could
define a function bailout() that would be determined by the current
bailout function, and called at the bottom of a formula where the
current bailout test goes. If we did this, then it would be easier
to make a formula behave like a  built-in type.

The bottom line is that we're not seriously considering removing
built-in types now, but we are actively experimenting with removing
integer math.

Tim


- ------------------------------------------------------------
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: Sun, 19 Oct 1997 18:37:51 +0200
From: Dirk Meyer <[e-mail]>
Subject: Re: (fractint) Integer math is history, and another modest propo

Tim Wegner wrote:

> Generally, formula types are faster than built-in types.

Really? Faster? Then you should improve the formula parser to
compile C code and compile Fractint with it :-)

- ------------------------------------------------------------
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: Sun, 19 Oct 1997 11:37:48 -0600
From: "Tim Wegner" <[e-mail]>
Subject: (fractint) if you are removed from the list

From time to time a list member's email address might not work, and
all the mail being sent by the list bounces back to me. There are a
variety of reasons why this could happen; your local machine might
be down or there might be some problem on the net.

I have learned not to panic when this happens; when the list first
started I unsubscribed a couple of people right away, then realized
the problem was very temporary.

What I do now is wait a couple of days. If mail is still not going
through after a few days, I unsubscribe the non-working email address
from the list.

If you discover you are off the list, and want to be on it, just
resubscribe. Hopefully whatever caused the earlier problem will be
over. Unfortunately the same problem that prevents list mail
from getting to you also prevents me from sending you a friendly
email describing the problem. If I do unsubscribe you I assure you
it's not personal. It just means that your email address appears not
to be working. It may someday happen that I have to unsubscribe
people for other reasons, but that hasn't happened yet; we seem to be
a civil bunch <grin!>

Folks can help by unsubscribing when you know you email address is
going away. You can only unsubscribe from the same address that you
are subscribed under, but if your email address has changed, try
unsubscribing the old address anyway - your request comes to my
attention and is easily accomplished by hand.

If any folks are experienced running lists, I'd be interested to hear
how you handle the situation of apparently bad email addresses of
subscribers (how long you wait etc. etc.) (Send me email please,
don't post to the list because the information won't interest most
list members.) So far the list administration task is not too bad.

Thanks. The list seems to me to be going well. I'd encourage shy
beginners to ask questions. List members have been good about
answering questions. I've intentionally invited more technical
topics to complement the fractal-art list, so much list traffic is
too technical for some people's taste. But do remember that less
techical discussion and beginner questions are also welcome.

Tim



- ------------------------------------------------------------
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: Sun, 19 Oct 1997 12:02:15 -0600
From: "Tim Wegner" <[e-mail]>
Subject: Re: (fractint) Integer math is history, and another modest propo

Dirk wrote:

> Really? Faster? Then you should improve the formula parser to
> compile C code and compile Fractint with it :-)

You joke, but as I mentioned in another message, the formula parser
literally CAN compile formula code, though we don't (yet) let you use
it that way.

The fact that the parser is so fast (if you have a 387 or
equivalent) is a tribute to Chuck Ebbert's expert programming. Your
fractal is generated with variables that are kept almost entirely
in the coprocessor stack registers where they can be rapidly
accessed. Whenever possible, computations are done directly with
built-in coprocessor functions. His optimizer avoids whenever
possible moving numbers between memory and the coprocessor.

C code optimizers are better than they were, but they haven't caught
up with Chuck yet <grin!>

Tim

- ------------------------------------------------------------
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: Sun, 19 Oct 1997 12:39:22 -0500
From: "Justin A. Kolodziej" <[e-mail]>
Subject: Re: (fractint) Copyright law

Andrew Schoonmaker wrote:

> At 10:53 10/16/97 -0600, you wrote:
> >
> >In article <[e-mail]> ,
> >    Andrew Schoonmaker <[e-mail]>  writes:
> >> A couple of questions...  One, what do you mean by "period 16" and "period
> >> 64"?
> >
> >I'll take a stab at it... each "bud" in M is a basin of attraction.
> >Inside each "bud" is a point c, where z := z^2 + c is a cycle.  The
> >length of the cycle is the period referred to; we can label the entire
> >basin of attraction with the period of this point c.  The main cardiod
> >is period 1, the bud along the real axis to the left of the main
> >cardiod is period 2, the next bud to the left of that along the real
> >axis is period 4, and so on.  The period doubling as you progress down
> >the real line can be viewed in fractint by looking at the bifurcation
> >fractal type.  The bifurcation fractal type is a visualization of the
> >dynamics of M along the real line.
> >
> >Clearer or muddier now? :)
>
> Oh, that's about what I thought you meant.  Doesn't Fractint have some
> internal coloring scheme which displays this?
>

Yes it does.  Just set the inside coloring scheme to per.  Then the inside of the
M-set is colored according to what period the point settles down to.  However,
unless you use a lot of iterations, the results aren't very good, and it tends to
slow down the calculation immensely.

Justin K.

Bad(?) Linux slogans:
- -- Linux: The OS for truly independent spirits! --
- -- Linux: The OS of champions! --
- -- Linux: Marquette's Web server (www.mu.edu) uses it, shouldn't you? --
- -- Linux: Because it's free --
- -- Linux: Because you can --
- -- Install Linux: Tell Microsoft where to stick it! --
Justin Kolodziej 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"

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

End of fractint Digest V1 #36
*****************************

To subscribe to fractint Digest, send the command:

    subscribe fractint-digest

in the body of a message to "[e-mail]".  If you want to
subscribe something other than the account the mail is coming from, such
as a local redistribution list, then append that address to the
"subscribe" command; for example, to subscribe "local-fractint":

    subscribe fractint-digest [e-mail]

A non-digest (direct mail) version of this list is also available; to
subscribe to that instead, replace all instances of "fractint-digest"
in the commands above with "fractint".

Back issues are available for anonymous FTP from ftp.xmission.com, in
pub/lists/fractint/archive.  These are organized by date.