And another related question -

which version of Poser was the first to allow Mac syntax runtime-relative pathnames in the runPythonScript statement in a Poser file ?

If I use Windows style path syntax (I'm on Windows 10) it still doesn't work

But with the full absolute path starting "C:\Users\..." it works.

Of course that's no use in a freebie as I have no idea where users will put the file.

So the new question now becomes "can Poser 6 handle runtime relative paths from a runPythonScript statement in a Poser file ?"

Y-Phil posted at 3:44 PM Fri, 30 August 2024 - #4488923

I have always found Poser rather slow at object deletion... My impression is that it kind of reload a certain number of things...
There's a free tool you could try: Snarlygribbly's Scene Fixer. You will find it in the Poser 11 section here. It contains many tools, among them the ability to delete characters, lights, props, etc...

So when Poser deletes attached figures after clicking OK below it's clearly doing a lot more than just that.

I think I'm happy with including a warning and a delete script with my freebie.

However, there's clearly an underlying problem in Poser.

I've never really been bothered about the time things take to delete until now, and it's only the apparent doubling in time per additional nested figure that piqued my curiosity... that and the possibly related Poser crashing. So I'm mainly wondering why the deletion time appears to double for each additional nested figure, and it definitely appears to double - I did another quick test while I was working on progressively larger stacks, estimated the deletion time with the old-school counting "one thousand two thousand..." approach:

10 nested figures (40 books) - about 10s
11 nested figures (44 books) - about 22s
12 nested figures (48 books) - about 38s
13 nested figures (52 books) - about 84s

For 14 nested figures I deleted the top half first (a few seconds) and then the rest (another few seconds).

Thanks for the suggestion too. I've used Scene Fixer quite a lot in the past and found it, like all Snarlygribbly's stuff, very useful. 

I'd originally thought that the Poser crash was related to running my scripts, but in the test I did in the post above I didn't run any scripts. And I've just been playing around loading all the CR2s and running my scripts, but NOT deleting anything, and there's no sign (yet) of any problem.

It's not conclusive, but I have a gut feeling that the Poser crash is related to the deletion problem.

(I'm using Poser 11)

Right, that didn't really help.

But editing the CR2 reminded me that each bone has a joint (and joint parameters) with its parent, and another joint with each child. So it would seem that jointx, jointz, and twisty are for the the joint with the parent bone (just one parent), while bone_2_twisty, bone_2_jointz, and bone_2_jointx are the joint with the child bone, the bone id being required because there can be multiple children (and the names without bone ids are already used for the joint with the parent anyway).

I'd guess that when you look at the joint parameters for a bone in the Setup room you're seeing the joint parameters for the joint between that bone and its parent, which appear in the CR2 as the  jointx, jointz, and twisty within that bone(bodypart/actor) but also as the bone_N_twisty, bone_N_jointz, and bone_N_jointx in the parent bone(bodypart/actor).

So while the x/y/zRotate of the second, third, etc bone in a figure bends the joint between it and its parent, the x/y/zRotate of the first bone in a Poser figure doesn't bend any joint, but simply rotates the bodypart. That's perfectly logical.

I think I need to insert a dummy bone. :)

Except that with the Sixth Sense you knew you had to re-watch the whole movie, whereas with the short story you didn't realize that chapter 2 was the same as chapter 1 until you were some way into it and suddenly thought, "hang on a mo', I've read this before..."

I've been racking my brains to try and think what the story was about, but no luck. The nearest thing that comes to mind is the movie The Sixth Sense.

The only differences I can see in the CR2 between bone_1 and bone_2  that seem to relate to this are highlighted in yellow on the comparison screenshots below - bone_1 on the left, bone_2 on the right.

Firstly, the twistY, jointX and jointZ are in a different position and order in the channels section and are named in a different manner

Secondly, the twistY, jointX and jointZ of bone_1 all have two extra parameters - flipped and primaryParm.

I've checked Kuroyume's unofficial CR2 spec but 2004 was a long time ago and it doesn't really give me any clues, except that the order of jointX and jointZ is important

While I wait for somebody who knows what they're talking about I may just try editing a copy of the CR2 to make the bone_1 joint parameters similar to the bone_2 ones... what's the worst that can happen ? ;o)

The reason I'm asking is that the problem came up in post #67 on my topic over at Hivewire My rather infrequent, and possibly apocryphal, 'books project'. | Page 4 | HiveWire 3D Community. I know that I can sidestep the problem by inserting a dummy bone between 'Body' and 'bone_1', but I'd prefer not to do that if it's possible.

This is a screenshot of the "3 Pool edge random brown" shader. Change the four colours in the ColorRamp and the Mortar_Color and Mortar_Thickness in the Tile node as you wish. 

And here's a screenshot of the "4 Pool tiles random blue" shader. It's exactly the same, but with different colours in the ColorRamp and a different tile size - the 40 which the big red arrow is pointing at. Change the four colours in the ColorRamp and the Mortar_Color and Mortar_Thickness in the Tile node as before, but you can also change the tile size, although make sure it's a multiple of 4 otherwise the tiles on the sides of the pool will be misaligned. 

Pool mesh - diving board, ladder, and water plane are all part of the same object (1,358 vertices, 1227 faces)

Closer view of diving board and ladder

Nine material zones (ignoring 'Preview'). All except the ladder and three diving board ones are UV mapped (by which I mean they have UV mappings that I deliberately created. The "non UV mapped" ladder, diving board, and previously mentioned boing text are actually UV mapped, it's just that I didn't do it - it's just whatever Blender did when I created them, which is probably each quad mapped to the full UV square)

The ground (material zone name mis-spelt), water, and pool floor are UV mapped from top view

The pool sides UV mappings were lovingly hand-crafted so that the Poser procedural Tile node (or indeed my also lovingly hand-crafted random colour tiles stuff) would work nicely on it. It's basically a cylindrical mapping from a point inside the pool constriction, manually adjusted so the tiling remains consistent.

And finally the pool edge, which is the 64 quads around the pool rim plus the bevels, is mapped to an 8x8 grid ! I did it this way so that I could use the Poser procedural Tile node (or indeed my also lovingly hand-crafted random colour tiles stuff) on this too.

Lightweight "Boing!" mesh (732 vertices, 724 faces)
Not UV mapped.

Includes two props

and several materials for the different parts of the pool.

The freebie's been approved, so here's the link to it in the Renderosity Free Stuff: Simple toon pool (for Poser*)

Please feel free to post here any comments (good or bad - constructive criticism preferred), suggestions for improvement, and/or questions about this freebie.