I'll second the Shapeways option,
interesting business model, but tricky to navigate site, they have an API available that needs a guru like yourself to package for us SU minions.
I've been looking for a 'water-tightness' ruby that passes there upload criteria, it seems a bit hit and miss at the moment (not just with SU models)
this hard to find page is worth a look
http://www.shapeways.com/tutorials/design_rules_for_3d_printing
john
@d12dozr said:
Thank you for the detailed explanation.
no problem, don't know if it helps Al (I know next to nothing about UV mapping, but I just saw this sitting on the desktop and may as well add it
@d12dozr said:
And just HOW do you do that?
I believe that is what he means...
probably the hard way,
I strip out the annular rings from one sphere, make a group, rotate copies until I like the pattern, explode the lot, slice off the top (3 or 4 rings), drape with sandbox, slice off the top of a plain old 'follow me' sphere (just 1 or 2 rings), offer up the sandbox hat (Gai was right, just a hat), decide how far to use the pattern, trim, explode, add projected material off a number of different angled surfaces I paint ( and make unique before projecting), decide what I like, purge all the test materials, and that's it...
I make these cages for pipe along path objects, and just thought I try draping and rendering one night,
foolishly I always bin the makings and have to start from scratch each time, but I'll keep this drawing and use it as a component, nearly everything is grouped so have a play.
You only need to change the top 1 or 2 rings to completely change how the texture gets projected, but I have no idea why.
On this one I also copy pasted the cylinder off the standard one because the texture worked better on that one, again no idea why.
john
2004 by Gavin Kistner- Geodesic_SketchUp.rb
don't remember where or when I got this, have you tried it?
john
[quote="Gaieus"]Hi John, If you mean this image - I guess, yes./quote]
Cheers Gai,
I've been using the feature only since since reading this post, find it quite handy and wondered if a Cardinal Points Highlighter ruby exists, or was easy to implement.
I might even try myself if it 's not available.
john
Hi Tig,
Just to confirm it all seems to work on the mac, tool bar stays between sessions, no new clashes, needed a few more 'first rail reversals' then previous, but I just randomly chucked down and rotated 2 lines to test on, so not surprised, just gets better... cheers.
one thought, is it possible to allow navigating to see if you need to reverse a rail, from some views the new mesh can hide an issue?
john
@bob james said:
@driven said:
I'm not sure I understand your suggestion, but thank you for it.
Bob
Hi Bob,
I see TIG has covered what I meant by pop-ups, but in case it's a more overall misunderstanding.
A few more bits...
Any ruby can only do what YOU tell it to do, based on your understanding of what the script author has told it to do, based on his understanding of SU team has told SU to do, and on and on....
So as non-authors the only bit we can do is either randomly stumble about using a script and hope it works (which is quite fun) or we can methodically evaluate OUR input and keep track of what happens when we do certain things. Then if we come across something 'odd' we can then ask someone else to try exactly what we've done to see if it might be a 'bug' somewhere up stream or simply a local issue.
Because Mac's can have there own set of problems with SU I have been using the above approach on many rubies to make sure problems aren't just me, or my mac's, and it's a great way of getting your head around what can and can't be done with a new ruby.
I also check results against other methods and sometimes (but not with the ground breaking one like this) find that the vanilla SU tools or a discarded 'oldskool' ruby do what I've been trying to achieve for weeks much faster.
cheers for a another great script TIG...
john
@bob james said:
This is just the latest in many tries.
Help :?
Hi Bob, some more things to try...
There are basically 8 ways ( that could give your expected outcome) for any 4 loose profiles (16 if you add surfaces to each, 32 if you don't mind Scaling the length or width back after processing )- Most give either normals or reversed identical copies, some differ and that might be the one you want.
What I've done with stubborn ones is
save them into a new drawing, scale x 10, explode curves, re-Weld, moveCopy of each line out-board by a fraction
you now have 8 line segments, 4 attached + 4 unattached, name or number them, (e.g. L1,L2; T1,T2; R1,R2; B1,B2),
if you really want to be methodical, make 8, 16 or 32 sets of this, side by side, cut and paste in place (just to be sure there clean)
go through all combinations i.e. L1-T1-L2-B2 and name the new group after each combination is used, if you don't want to scale back to size use attached lines as you rails for the first 8 sets
If none work at all you probably have a ruby class, i get some issues if I've used certain tools to make the curves, but closing and reopening the drawing clears those and I also had a problem selectonopen. rb so I binned it and use the space bar.
this plugin is worth getting to know, it's very,very useful ...
john
PS. to speed things up click no all pop-ups, and then you have a proper record of what does what.
@unknownuser said:
Seems there is a little bug!
might be related to this
in all the test I've done on ramps, curtains, TIG's rails and more, I seem to only have a problem with one direction of selection, (other then complete cock-ups)
I don't know if it's a clockwise V counterclockwise issue or what, but one way round seems to generate hidden grouped and ungrouped geometry DURING the operation that seem to account for the result, the only rail never effected in the above test is still in red,
when the hidden geometry was behind any of the other three i had one of the odd results, and it wasn't there at the start of each run. often there were 3 or 4 sets of hidden bits. A ran this particular drawing in six different drawings 20 or 30 time, the problems were always on the same selection.
Only got the points cloud the once?
TIG, one thing I would appreciate is an escape key, if you accidently choose 2nd rail twice (instead of the profile) it can generate vast numbers of edges
john
@august said:
Hi John,I did some digging to see if I could find CadKey or PRT-import for Windows.
August
Hi August and anyone else watching this topic,
happy new year,
I was just having another look at the files and notice there's a windows CADKEY to DXF convertor enclosed, called ALCADMY.EXC, I don't know if you can run that, I can't on the mac and have no PC access at all, 4 Mac family I'm afraid.
john
Hi again,
I certainly don't NEED it enough to warrant any effort, importing via MeshLab as either odj, dae or dxf is very painless on a mac, for the odd occasion that I'd want to.
the cow, which I copy pasted into a text file, re-named cow.ply is the only .ply I've had any success importing to SU and it won't open in MeshLad (I get a missing header message)
I'm more in need of 'skink-wrapping' of complex engineering geometry which I can't seem to find for Mac SU usage.
john
how do I revert to standard import to see if it bring faces?
john
Hi,
this could be very handy, but the skinning is not working on mac? is it possible to tweak the code?
john
hi Kirill,
1.2, when did you post this? I can't keep up... looks even better, but I quickly downloaded and tried it on the curtain skp and it's not recognising the group that I was using with 1.1 so I explode, re-grouped added a second group, and still get the message to choose groups? I had this with 1.1 but can't remember the work around...
I'll have a play
john
sorted, DrivingDimensions was interfering so I disabled it and 1.2 appliqué is 'part of the fabric', literally. very, very nice...
after having a play your skp I noticed I already had a slight rotation on the group from 1.1 test and I must of done it to get it to run...
it seem to be what was causing the lifting at on end, so I re-aligned it and added a single square at right-angle to the text face (out of the way) and it processed fine.
Tips for anyone giving '2D' text a try-
leave the 'square' till last to avoid needing to use outliner to select the text group
apply material to recipient mesh first, so you can you 'select face material' and ' erase co-planer', to aid clean-up of mesh
I found, using cut, paste in place, explode into the mesh group gave a very clean join.
leave the back of the mesh default to aid changing the face colour texture, (or hide a flat colour swatch in the group)
Tips for playing around with my skp-
each side is now 3 groups on separate layers, so you can toggle front, back, or thickness independently to have a look.
these 6 group are inside 2 groups, so you can use 'scale' tool to open and close the curtains.
if I was to use it in a model I would delete the front and thickness groups and apply 'fabric' to both remaining faces(less polly's)
Now I'm off to 3D Warehouse to find a car to stick 'DECALS' all over...
cheers again Kirell
[attachment=1]
Hi Kirill,
1.1 is even better, but I guess that's the point... well done.
I think anyone on the mailing list may have not know about the update yet because your edit of the top frame doesn't appear to send a 'new message' notification and I got a you tube 'new movie' email and came looking for it.
been playing with using it to apply decals, etc to objects. You can sort of do it with photo textures, but grouped or imbedded geometry is way nicer, and easier to change materials, so I'm very happy.
The text is single surface, no thickness
On my first test, which I binned (because it had a typo) I used- Yes check for non terrain, Front and Super, when it was cooked (much faster now) I exploded both groups, smoothed, softened and textured. I then did a volume check and there were no holes at all, perfect first time, except for the typo.
I binned it, changed the font and size, same settings. When cooked it sat off the surface (.25), with one end further away then the other? so I went though everything I could think of until I eventually got one that was almost right, added .002 thickness and merged that.
Nowhere near as nice as the first so I tried scaling down the text used the original settings and ended up with a perfect profile sitting exactly in the other surface, but I couldn't the two to intersect, so this one is actually .00025 in front off the drape ( so that you can see ti from all angles) and it's still grouped separately unlike the first.
if you want to have a go at merging these two here's the file
@red_shift said:
Any idea why on a MAC (os 10.6.2) and SU 7.1.58 the EEbyR plugin runs successfully then causes SU to go into a stall and then crash?
no, but I could test your 4 profiles on 10.5.8 and see if it's a 10.6 issue, if you want.
at first I had problems which I now think were caused by using one of Kirrel's or one of Chris's tool earlier in the session, since then I've been restarting once I've done the curve construction and haven't had any problems.
john
@tig said:
driven
Thanks for all of the permutation testing...
My pleasure, I know there's a few people needing very accurate meshes for different things and quantifying what affects the outcome adds POL to your excellent script. {Principle of Least Surprise~ to quote Kirril}.
My contribution is minor compared to the effort you put in, and I don't write scripts (yet) so the least I can do is pass on some learning (and point out my errors).
@tig said:
I'm unclear about this one missing small triangle - did EEbyR miss it out or did you ?
I deleted the the inner edge of the wayward point (instead of fixing it) to keep tabs on when it did and didn't appear during my testing, and then decided it was tidier to leave it exposed then to fix and label it, I don't need to use the drawing.
why it's happened? I initially used 'paste in place' to add the top line sans surface for the top rail, later on I used 'paste in place' for the whole profile, to use it as the top rail. I was using group layers to keep everything separate, but when I did my final cleanup I had two identical ungrouped but separate polylines occupying the same space on my drawing layer, so maybe it was there all along and SU got confused by that, I certainly am...
I take it the file link works?
john
