[Plugin] Shape Bender Beta
-
This is what I get and than it start repeating itself
-
Thats perfect, thanks! I'll see if I can find the problem there.
Chris
-
Hmm, that is a progressbar error. What OS are you running?
-
XP
-
I have reinstalled progressbar.rb but still same error.
-
So Chris. . .I got the new version. Works pretty well. Sorry you were working this weekend on it. I hope you still had time to dye a few Easter Eggs.
Anyway, couple of q's. 1) are there supposed to be more than 1 icon on the toolbar. Seems I downloaded several png's. Edit. . .Oh wait! The other pngs are for the cursor. . .D'uh
2) the original Group doesn't erase. Is that a change with the latest iteration or what? -
Hi David, no worries, I had plenty of Easter egg time. It was a good weekend.
Yes, the original component/group is suppose to stay in place now. That was a recent change.
As for the errors you're getting Sepo, I should be more clear, its not a problem with progressbar. Progressbar is working fine. Its something to do with what I am feeding to progressbar. That has always been an issue with the component spray. Could you try running make faces with the ruby console open and see what happens? If it is also broken, then that will help determine that there is still something somewhere messing with progressbar. Thanks!
Chris
And don't hesitate to tell me you're tired of troubleshooting for me. I gldaly would do it myself if I could reproduce the error, so a great big thanks for all your time Sepo!
-
thank
-
i'm wondering how difficult it would be to make a ruby that preps an object to be bent. (it could be a part of ShapeBender or separate as it would benefit fredo's bend tool as well)..
it would basically replace the need for doing some repetitive steps in order to set up certain objects to be bent cleanly.. i can achieve the proper shape via math or the more simpler route of SU inferencing but the ruby could work either way (or can it?)
this ruby has already made it a lot faster for me to draw curved shapes because i can do all the prep work on xyz axis and then bend it later.. something with one compound curve used to take me 20-30 minutes to draw and now i can do it in 5-10... introducing yet another curve (2nd attachment collides 3 different radii) used to take me over an hour to draw and i can do that in ~20 minutes now.
my question is whether or not this is out of the scope of a ruby script. ??
so if the first example is possible, how about adding yet another curve into the mix?
curve faced wall with curved topoh, am i the only one noticing a huge slow down with the introduction of the progress bar? certain actions are at least 10,000x slower now.
[edit] also, those examples show the problem with the softened edge when trying to do precision fitting.. i could extend the straight line to get hard ends but then nothing goes in it's right place.
-
Chris I am not at my desktop at moment, so I can't really test it. I will test it tomorrow on my desktop. And it is not a problem at all. I am very greatful to you for your wonderful free scripts and if I can help with anything please do not hesitate to ask.
-
Sorry Jeff, I just don't see how its possible at all. At least, not so that it would work with any regularity on more than a small handful of objects.
But I do have a question, you say (in another post, and I'm paraphrasing) that the geometry is useless to you the way it currently is working. What are you planning on doing with the geometry after you bend it that makes it so useless?
Chris
-
i have to build it.
pretty much everything i build is curved so in that regard, SU is a weird program to work with but the segmented curves aren't a big deal as long as key intersections/joints are at vertices... unless i go insanely high poly with the drawings, i can be out 3/4" or more which just isn't an option..
also, if i'm colliding multiple curves then the resulting geometry isn't a radius or an ellipse and i have to plot points for construction... if those points aren't in the right spot, my measurements are off again..realistically, i have ways to build this stuff without drawing it first but by working through it on a computer first, i can do it more easily on site and/or have a way to show someone on my crew to make the cuts.. it's still faster for me to spend the hour or two on the computer getting my points in the right spot than it is to build without the drawings but if i can streamline the process even more then i'm not complaining .
[add] - as far as the actual drawing goes, there are problems as well.. these shapes are just shells and i often have to continue on to figure out the framing in tricky areas.. if the planes/lines aren't where they should be, it just doesn't work out too well.
-
Oh yes, now I remember the kind of stuff you use SU for. I think you need an entire plugin of your own for that skateboard park type work. That is definitely well beyond the scope of what I was thinking of with shape bender, as it would require a different type of bending calculation I think. Definitely better suited for a solid modeler or something. Sorry I can't be of more help at this point Jeff,
Chris
-
@chris fullmer said:
I think you need an entire plugin of your own for that skateboard park type work.
oh man, that would be nice.. if times were better, i'd be into paying for 2-3 custom scripts.. maybe in another year or two. :hope:
@unknownuser said:
Sorry I can't be of more help at this point Jeff,
ha, your plugins have already helped immensely.. i used to have to draw stuff like this line-by-line and being able to bend things now cuts more than 1/2 the time it used to take.. don't sweat it..
you gotta realize that i know i'm asking for some weird requests that only I or maybe a handful of others would actually use ..just in case anyone else has the need to get the lines to break like this, i put a little step-by-step together...
custom_arc_tut.skpi actually do it a little bit different than this (esp. steps 4-6) which is a little easier/less confusing but it was harder/more confusing for me to show how i do it.. this should get the idea across though.
-
Chris, you are a machine!! (the good kind, of course)
-
Jeff, I want to emphasize a bug in the code since it might matter to you, since you are making shapes that get built. But the first and last angles might not be calculated the way you would expect - I suppose that depends on what you expect. The image shows the problem. Every sliced part of the model lines up correctly to the curve expect the two ends. That is because I am not actually finding that outer angle at this point. It is a planned improvement
So I thought you should be aware of that so that it doesn't cause you undue pain when you can't figure out why stuff is not lining up right,
Chris
-
yeah, i noticed that.. i didn't realize it was a bug though and i more/less attributed it to being like using the follow-me tool on an arc.
here's a top view showing that problem:
the workaround (and the same workaround goes for shapebender) is to make sure there is some flat before going into the bend.
i guess if you manage to figure out how to fix your bug, pass that info along to google so they can fix follow-me
[fwiw - if you look at the skp titled bend2.skp a few posts ago, you'll see that i added the flat ends.. the flat won't affect the setup i do to get all the intersections to match up because there are no vertices on it.. if i add 4' of straight wall prior to bending, i'll also add 4' of straight line before the curve]
-
Chris
It seems make faces makes error but makes faces nevertheless. Here is an extract from the console.
Error: #<ArgumentError: comparison of Float with nil failed>
(eval):44:in>' (eval):44:in
update'
C:/Program Files/Google/Google SketchUp 7/Plugins/makefaces.rb:146:inmakeFaces14' C:/Program Files/Google/Google SketchUp 7/Plugins/makefaces.rb:140:in
each'
C:/Program Files/Google/Google SketchUp 7/Plugins/makefaces.rb:140:inmakeFaces14' C:/Program Files/Google/Google SketchUp 7/Plugins/makefaces.rb:163 C:/Program Files/Google/Google SketchUp 7/Plugins/makefaces.rb:163:in
call' -
Chris
I'm not sure if this is the core issue, but rotating the arc to mid segment works for this simple try.
dtr
-
dtr...
the problem with doing that is it throws off all the measurements.. the SU arc is only correct at the vertices -- not the midpoints of the segments.
Advertisement