Random Patchwork Sphere

horizontal objects are distributed radially, but, as far as I can tell, at random angles and length,

The latitude/longitude distribution is random. The scaling is not; each instance of whatever you use as the original object is scaled, rotated, and tranlated so as to be effectively "pierced" by its thrust line (the invisible "radius" of the "sphere" at the instance's center).

Thus, it only looks like instances of an original object with a horizontal aspect ratio are distributed radially at "random" length. To see this, first make a Symbol out of your horizontally-shaped object. Then apply the script, then use Replace Symbol to replace the instances with instances of some other Symbol that has a square bounding box (ex: a circle). You'll then see that the instances are not scaled randomly at all. The scaling is applied as percentages of original width, according to the object's longitude. Because Symbol Instances retain their scale percentages as attributes, the replacing Symbol is scaled to the same percentages.

and it seems that the script can be run on any object, not just a circular symbol.

It can. However, that is inefficient, and prevents your exploring alot of other possibilities inherent in following up with the standard Replace Symbol command, or my RandomReplaceSymbol script, and/or the Symbolism tools.

I like this script very much. Thank you.

You're welcome. Let's see some of those fascinating results you mentioned.

JET

Peter,

Good examples. And being differnt from mine, they help show the variety of effects that can be achieved by playing around with simple geometric manipulations like this script. (Imagine the time it would take to generate those same effects "by hand" in the standard feature set.)

Replace the results of running the script on a non-symbol line with a symbol (clumsy sentence, sorry, in a hurry) then the distribution and scaling work, but the rotation and translation disappear, reducing the results to two dimensions rather than three.

That's why I recommend making a Symbol out of your initial object before running the script. As explained in my last post, the script scales and rotates (and translates) copies of the original object you have selected. Illustrator does not store scaling and rotation values for ordinary objects. It does for Symbol Instances.

Each Symbol Instance on the page carries with it its current scale and rotation (and other attributes, such as colorization applied with the Symbol Stainer) as long as it remains a Symbol Instance. Therefore, when you use the Replace Symbol command to "swap out" the base Symbol of an Instance, the new Symbol "picks up" the same transformations that were applied to the Instance.

JET

Would it be possible to post an AI file of your example or a PDF that could be opened in AI? That would be very helpful so that I can spot where I'm going wrong.

Jay,

My posting an .ai file would in no way reveal anything extra about the script. The result of the script is just a set of copies of whatever object you had selected. There is no after-the-fact "live connection" between the script and the objects for you to dissect.

my original circle symbol ends up with very jaggy edges. JETs circles maintain their circular look.

Well, so do mine. Look at the ball of pink triangles. Or the ball of randomly-replaced Symbols (most from the default AI Symbol library) at the very bottom of my first screenshot.

As already explained, the script effectively merely does this:

1. Uses some basic trig to randomly "plot" locations on the surface of the "sphere".

2. For each location, scales, rotates, and moves a copy of the selected object, so as to be properly shaped as if the "radius" of the "sphere" were the copy's "axle". The script does not do any kind of warping to add any spherical surface"bulge" to the copies.

So with a circle as the selected object, don't think of the individual resulting ellipses as renderings of bulging "contact lenses" as they near the horizon of the sphere. They are just ellipses. Think of them as mere circles drawn on the surface of the transparent sphere. Any "bulge" you want to add to those which occur on the horizon of the sphere would simply be a matter of drawing a circle behind the whole thing, slightly larger than the radius value you entered in the script's dialog.

Think of it this way:

1. Draw an ordinary small circle.

2. Now imagine that this circle is drawn on the equator of a globe. From your line of sight, the equator is a straight horizontal line. The only reason the circle's shape appears perfectly circular to you is, you are viewing it "dead on." In other words, that circle, laying on the surface of a globe will only look like a circle to your eye when your line of sight happens to be the one radius of the sphere which pierces the center of the circle--the radius of the sphere that is pointing directly toward your eye.

3. Now imagine the globe spins a bit counter-clockwise (as viewed from its north pole). As the globe spins about its pole, the circle enscribed on the equator becomes more elliptical from your viewpoint, because it is "rotating away from" your line of sight.

4. So what would be the apparent shape of the circle when the globe has rotated a full 90 degrees? From your viewpoint, the circle now lies on the horizon of the globe. You are viewing the circle perfectly "edge on". The circle now looks like a straight line. That's what you are calling "jaggedness." But it's not jagged at all. A straight line is exactly (in orthographic projection) what a circle would look like when rotated 90 degrees from your line of sight.

It's simply that you just drew a circle. The script does not in the least pretend to draw a sperical surface within your circle for you.

For all practical purposes intended by the script, the profile of any such desired "bulging surface" to fill in the ellipse would only be evident at the edges of the sphere. And the radius of those surfaces would be the radius of the sphere. So simply draw a filled circle behind the set of ellipses with a radius equal to the distance from the sphere's center to the corner of one of what you're calling the "jaggies."

Now, of course, for shapes that include straight lines, such as my pink triangle, an increasing slight curvature would have to be imposed on the edges of the triangles if they were meant to accurately depict triangles drawn on the surface of the sphere. That would be an extraordinarily ambitious Javascript. So with straight-sided polygons, you have to think of the script results as a bunch of flat-plate polygons "pasted" or "pinned" onto the surface of the sphere (like the mirrors on the Tellstar satellite), not as polygons actually drawn on the spherical surface.

So using a circle as your initial object, the script results in a quite reasonable reprensentation of circles either actually drawn on the sphere's surface or flat circular "plates" attached to or orbiting the sphere. (The difference is simply a matter of what size circle you decide to draw as the sphere itself.)

Using a polygon, the script results in a reasonable representation of polyonal flat plates attached to or orbiting the sphere, but not of polygons actually drawn upon the sphere's surface.

Some of JETs symbols have a transparent look to them. How was that implemented?

C'mon, Jay, you can figure that out. Simply apply transparency to the initial object.

For the appearance of a translucent sphere:

1. Run the script on your selected small circle. Group the results.

2. Run the script another copy of your small circle. Group the results.

OR

2. Simply make a copy of the first group and rotate it a bit.

3. Draw a circle with a translucent fill for the sphere.

4. Stack and align with the large circle sandwiched between the two Groups.

JET

Try copying the original result, then scaling it and applying Replace Symbol:

Group the first set of results; then apply Transform Effect:

Rotation setting added to the Transform Effect:

(Am I the only one who finds it silly the way this forum assumes every post is a question, and requests the silly awarding of meaningless "points" and insists that even a non-question should be marked "answered" or "not answered"?)

There is a toggle for that under the topic box when you start a new post.

Oh. Thanks. I'll try to remember to watch for it next time I start a discussion. (I still think the whole mess is a bunch of ridiculous just-because-we-can fluff.)

I would mark yours the "Correct Answer," so you can take all those "points", add $5.49 to them and go to Starbucks for a Mocha Latte, but I'll wait 'till I see the off-switch for myself. ;-)

Is there a way to completely opt-out of the "personal messages" functionality? I have looked for that, but didn't find it.

JET

An unforseen, but "happy accident" caveat:

JET

Latitude 59

Well, if that's the actual name of the establishment, I guess that would be appropriate to this thread. ;-)

JET

That's way up there.

Would that happen to be somewhere 'round 151 longitude?

;-)

JET

Since the scripts are no longer publically available, I believe no one should [distribute] them but James...

Thank you, Jacob.

JET