[UFBeta] Building Block Beta Formula Test- Updated version 9.8.4
Jayce Cruel
I've updated the building block formula once again. There aren't any real code changes, per say...mostly aesthetic stuff...
I removed a lot of the checkboxes, particularly for the block modifications and the mandras tweaks (they are now all hidable/unhideable as headings). The checkbozes you see left are the ones that turn on and off the code, which helps the formula run faster (why run the code if you don't actually use it, right?)
I also removed the extra "P1" from the block choice list, and renamed the other "user variable" to "Number". I hope that change doesn't confuse anyone
I also removed a couple of other "choices" that would never be used. Specifically, I removed the "z in tweak" and "z in prior tweak"...there is no "tweak" section, so it makes sense that there should be no "z in tweak section"
Oh, I also fixed the defaults so the blocks that are being used by default show, rather than just the first block of each section.
Finally, I cleaned up the interface of each block. With the exception of the first block of each section (which is always shown), as long as the operation is "unused" (which means the block is not being used in the section), the rest of the block's parameters will not be shown in the interface. This is the same with the modifications. I hid some of the parameters for modification, as long as the "mod operation" is "unused". As long as it remains "unused", you most likely wouldn't use it anyway...
I am still working on the help text. This may take a little while (it is currently about 25% done), which is why I wanted to dish this out now, rather than wait who knows how long for the help text...I know the interface is still confusing, and people don't know exactly what's what...please bear with me while I get the help text typed up...
Also, Toby, I want to answer your latest questions:
Also, I'm unclear as to how to use multiple custom variables. It appears
that they nest within each other, but the naming is very confusing. They are
not blocks, right? They are simply choices in any int param that chooses
variables in the equation and init blocks, as far as I can see. As it stands
you have ten equation blocks and only five variable "blocks" with five
custom variables. But these don't appear to actually be blocks, but choices
within other blocks. Is my confusion confusing, or do you get what I'm
trying to say? At first glance I wonder why there are custom variables for
only the first half of the equation blocks...
Well, the custom variables are blocks, but they are also "choices" that you can use within a block. The "Show custom Variables" checkbox displays the 5 sets of custom variable blocks, which allow you to "define" any one of the 5 custom variables available (there are 5 blocks per custom variable that you can use to "define" each of the custom variables). However, in any block of any section of the formula, you can "choose" (with the "choice" parameter) any of the 5 custom variables you previously defined as your "choice" for that block or modification (you can also choose a custom variable as a "modification choice" as well). These 5 custom variables you defined in the custom variable section can be used anywhere throughout the formula, in any block or modification, even block 10 of the equation section. You could even use custom variable 1 as a "choice" for a block in custom variable 5. The point of the custom variables was to allow you to define
your own choice- something beyond a simple standard element (for example, you can define custom variable 1 as "z ^ z", and custom variable 2 as "z * z + pixel"), and have a block choice be custom variable 1, and the modification choice as custom variable 2)
Toby: Of course if you use bools to circumvent code then leave them. It just
seemed odd, for instance, to check ""Show Block 1 Modifications" and find
the only thing that happens is the appearance of a second bool asking if I
want to modify block 1. It seems logical (to me at least) that if I want
block 1 mods shown I intend to tweak them...
It's been fixed. The checkboxes for the modifications is gone, replaced by a (hopefully cleaner) header. The only thing I don't like is that, if you do decide to modify a block, and later close it up, the modifier header still remains. I guess that could be an indication that you modified the block, but I was hoping it would completely go away...and if you uncheck the box, the modifying code won't run at all...
JC: As far as the bool params being on top of the group they open- I could try
it that way. The top 5 bool parameters are a throwback to the former JC
Dabbler series. I just left them at that, figuring that, since there were
so many parameters per section that the window probably wouldn't hold them
all, if I left them at the top, they would be easier to find. I continued
that trend for the custom variables as well, leaving them at the top of all
the custom parameters so they would be easier to find. I don't see any
reason why I couldn't try it the other way, and see how it works
Toby: It's up to you. But consider this: If I check a bool param that opens
a group, isn't it easier not to have to go searching for where that group
starts down a bunch of headers for other groups? The bools with their little
boxes are quite recognizable and actually serve quite well to delineate
groups--much better than a huge number of headers which look pretty much
alike, and which have to be read and digested to find where, for instance,
Init blocks end and Equation blocks begin.
Sorry, I haven't fixed this yet...I honestly think it's cleaner to leave them at the top, but I will try it the other way...I promise!
Toby: On a different topic, I really think that the next step for this formula is
a help file. I am totally unclear about how what relates to what. I open the
Init blocks and find that, for instance, if I choose current z for for Init
it gives me the same effect as choosing #pixel, and everything else gives me
a black screen. What exactly is happening here? I don't have the slightest
idea...
I'm working on the help file. That will come soon.
As far as choosing current z for the init, I think that the pixel value must be in the init section somewhere. If it's not there, it has to be in the equation section somewhere. Otherwise, z won't change from pixel to pixel, and you'll wind up getting a single color image. I'll have to say something about that in the help text...
I really need a block diagram so that I have some idea of the order of all
these things. Without this you are going to get people playing with a few
variables, finding that either nothing happens, or something happens but
they have no idea why, and just simply going back to the Mandelbrot in the
Standard folder, where at least they have a fighting chance of coming up
with some usable image.
Well...it's like any other formula:
Z is set to it's inital value in the "init section". This is done once per pixel. The default is a number (0,0)
(the custom variables are defined afterwards, in the loop section)
It is then that z is changed and manipulated by the equation defined in the "equation section". The default is "z * z + pixel", which is what you get if you write each parameter, block by block, together in a line
The "mandras tweaks" from that section are also applied to z at this time.
a bailout test is performed next in the "bailout section". The LEFTSIDE is tested against the RIGHTSIDE, using the "bailout test" parameter (>, <, >=. <=, ==, !=, etc.)
the default is "z >= 4", where "z" is the Left Side. and the number (4,0) is the Right Side, and ">=" is the "bailout test"
If it passes the bailout test, the formula stops iterating the for current pixel, and moves on to the next one. Otherwise, z is changed again, starting with the "equation section"
And each "section" that I mentioned above (with the exception of the mandras tweaks) consists of 5 or more "blocks" combined together, with each "block" containing a basic element of a fractal formula (such as current z.pixel, etc.), and modifications to that block
In other words:
1. The Init blocks are combined and calculated from the "init section". Z is set to that initial value
2. z is changed through the 10 blocks that are calculated and combined in the "equation section". z is also changed from the various available "mandras tweaks"
3. the blocks for each side of the Bailout test are calculated (the right side of the bailout test, and the left side). They are then tested to see if the formula should continue iterating, or stop and move on to the next pixel in the image
Sorry to be so tough. I am all the more frustrated because your formula
looks so very interesting but I have no idea how to drive it.
Don't be sorry. That's what this beta test group is for- to give comments, suggestions, hints and tips on how to improve formulas, as well as ask questions from those "in the know" . I appreciate the "tough love" you're giving me. It helps me perfect the formula I have now, and will ultimately make me a better fractal formula programmer in the long run
I look forward to hearing what you have to say, as well as answering any more of your questions. Like I said, I'm working on the help text. As soon as it's done, I'll put it up here for you.
Thank you again for all your testing efforts!
-Jayce Cruel
____________________________________________________________________________________
Yahoo! oneSearch: Finally, mobile search
that gives answers, not web links.
http://mobile.yahoo.com/mobileweb/onesearch?refer=1ONXIC
Jayce Cruel
I accidentally changed the default for the init section, block 1, from "pixel" to a number. If you want a julia style formula, you need to set the init block 1 back to "pixel"
I fixed this for you, and attached the fixed formula here. Not that anything was wrong with the other one- the only difference between this one and the one I just posted is that the default for the init section, block 1 has been changed from a number to the pixel value...this just makes it easier to go from mandelbrot to julia (for those who like to do that sort of thing)
Sorry about that...
-Jayce Cruel
____________________________________________________________________________________
Shape Yahoo! in your own image. Join our Network Research Panel today! http://surveylink.yahoo.com/gmrs/yahoo_panel_invite.asp?a=7