On 2013-10-24 2:12 AM, Hal Lane wrote:

...

Under Fractint For Windows = 0:00:00.12 Are you referring to Richard's Fractint For Windows beta 5 (F4Wb5) Yes, I wuz hoping to find him here. If so, you can post bugs to his code hosting outfit at: https://www.codeplex.com/

Richard has renamed his program: iterated dynamics Search on that to locate his project or go here: https://iteratedynamics.codeplex.com/releases/view/25662 Nothing new since 2009/Apr. Maybe we've lost him. He wrote something about people here being interested in maintaining a DOS program. It didn't make much sense to me since there's also an Xfractint and winfract from which he might still be able to rip code. He said something about standard interface, which to me means menus at the top that take all the fun out of learning to operate something fast.

You can view the program's "Issues" without logging in, but you need to sign up to post.

I haven't even read the bugs.txt in the distro, yet.

...

[pressed] V and that crashed F4W. I use F4Wb5 every day and have learned:

- Don't hit "V", use the existing resolutions, - If you calculate a .PAR file, decrease X-Mag by a factor of 0.75, - If you rotate a zoom box 1) re-set X-Mag to 0.75 & 2) re-set Skew to 0. Even the DOS version haz that problem; infinitesimal skews from error propagation or something. Every time I change the skew to zero on a rotation, it gets changed to something slightly more or less.

- In my WinXP, F4Wb5 hogs the CPU at Normal priority. I adjust it to Below Normal and I can browse the web OK.

I am one to change parameters to something sensible, like maxiter=255. Very few fractals jenerate visible data after sixty-four iterations. I know there are some. I've written one, and it doesn't require high iterations without a deep zoom, and deep zooms jenerate pretty much the same thing az shallow zooms, so I don't bother.

- F4Wb5 clears the screen sometimes when it doesn't need to (at least it does this in the Disk Video modes I use a lot.) I render at about two rezolutions: 600 square, and 3300 by 5100 (11" by 17" @ 300 dpi). The latter goes in a three by three batch file of 1100 by 1700 (or 1700 by 1100), so I can do a tesseral rendition; no waiting for either process, usually. - The formula MandelbrotBC3 does not work in F4Wb5; However, I have found a workaround. There are others. Read the "Issues" for more.

Doing the above makes images the same as DOS Fractint's essentially every time for me (e.g.: Jim Muth's FOTD.) I recall once seeing a difference between the two programs output in a fractal that had large areas that consisted of isolated single pixels at high iteration. The distribution of the isolated pixels was different. At lower iterations isolated pixels seem to be the same. Sometimes, isolated pixels rezult from a periodicity checking value that should be zero. Otherwise, conjugates in the formula, somewhere, tend to do it. If you look at the edges of my screen shot, you will see some black missing at the "outside" (inner set; it's an inversion). There are also some tesseral lines of the wrong colour; compare white with yellow.

IFS seems to work, but I've not tried things like Plasma, Lsystems or Cellular.

- Hal Lane

######################## # hallane@earthlink.net <mailto:hallane%40earthlink.net> ########################

...

-----Original Message----- From: fractdev-bounces@mailman.xmission.com [mailto:fractdev-bounces@mailman.xmission.com]On Behalf Of Jay Litwyn Sent: Thursday, October 24, 2013 12:39 AM To: Fractint developer's list Subject: Re: [Fractdev] Screen Shot Comparing F4W with DOS

Calctime For FRACTINT under DOSBOX = 0:00:17.19 Under Fractint For Windows = 0:00:00.12

17.19 / .12 = 143.25 times az fast.

I cannot run winfract under windows eight. I went to verify that the images were actually the same size (by pressing V), and that crashed F4W. The mouse does not work on palettes under F4W.

--- This email is free from viruses and malware because avast! Antivirus protection is active. http://www.avast.com

-- http://litwyn.comyr.com/ BrewJay's Babble Bin

On 2013-10-29 8:04 PM, Richard wrote:

In article <526D3A63.70002@spamcop.net>, Jay Litwyn <brewjay@spamcop.net> writes:

...

Nothing new since 2009/Apr. Maybe we've lost him. Nope. My time has been sprinkled around other projects, such as my Computer Graphics Museum, and I continue to do some design and research work on id that hasn't resulted in checkins to the repository for a while.

...

He wrote something about people here being interested in maintaining a DOS program. Tim and Jonathan have expressed their interest in maintaining the existing DOS-based fractint. They have an X Window System port, xfractint, and a 16/32-bit MS Windows port, winfract, as well. I seem to recall that chunks of fractint are turned off in these ports, but it's been a few years since I was working on the code.

I contributed to their source repository the source code that is used to build what I called "FractInt for Windows beta 5" as described in this blog post: <http://legalizeadulthood.wordpress.com/2007/04/13/fractint-for-windows-beta-5/>

They are free to take the beta 5 code and enhance/extend/improve it.

Some time after that, we disagreed on the future direction for the code base, so in the tradition of open source projects, I forked the code and created the Iterated Dynamics project. On what basis were the arguments? Compilers? SDK? GTK? DirectX? Whether to rewrite assembler az C? I don't discuss id on this mailing list because I consider that impolite. I withheld a remark on that. I did not think it wuz prudent to rename the product, when it so closely mirrors the DOS version, that a copy *should* be on the FTP site (http://www.fractint.org/ftp/current/). In my arrogant opinion, it duz not really matter what you call it if it's a DOS-version workalike, and some people care about the name of a rose, so to some people "id" wuz not politically sound.

Even the 2048*2048 version of Tesseral renderings requires four MB of address space, though, so even if your jumps are sixteen bit (or eight, plus lots of 8bit data access), it's awkward to be using anything less than a thirty-two bit data segment. I am still not sure if Andrew Schulman got hiz wish for mixed-mode code in Windows. I just know that Windows eight duz not support DOS off the shelf. Intel's design wuz always intended to be compact (az a bonus, they've managed to make *most* 8, 16, and 32 bit register to register operations happen in one cycle for over twenty years). It's a wonder that M$ can put twenty-five gigabytes of OEM installation on my machine that duz *less* than Windows XP, and send me three gigabytes of update (Windows 8.1), for which I see pozitive change, but uh...I would rather see WMP come up with a visualizer when I hit a link to an MP3. The important thing about forks iz that not only are people free to build on one, they are also free to port, and it helps if people *know* where the other forks are, which means pretty much keeping them together on directories (and mailing lists). If UltraFractal were open source, I am sure it would be easy to see where he ripped code from :). A top row of menu options makes it inviting to new users. A keyboard interface makes it inviting to experienced users. Add documentation of the keys, and more space is for that at 16:9, then new users can learn how to use their fingers to write things more beautiful than (H8 windozer).

There is a discussion area on the id project site: <https://iteratedynamics.codeplex.com/discussions>

_______ http://litwyn.comyr.com/Sound/brain_fart.mp3