ComponentDefinition.insertion_point returns the 'origin' axes of the Component Definition as it was originally made or they where later relocated - which affects the subsequent insertion point of instances.

ComponentDefinition.insertion_point=[12,34,56] <say> will appear to reset this BUT when you repeat the above although it returns this new value it doesn't move the axes or where subsequent instances are placed...

Weird...

I know what a Hash is in Ruby, but are you guys talking about the same "hash"? It's not clear how I could use a Ruby hash without running into the same problem; needing to search Entities for the matching ID.

I wanted one of my plugins to handle selections in undo and redo operations, so I wrote a Selection Manager to handle it. It appears to work just fine using the entityID of each selected entity stored in an array of arrays; and stepping back and forth thru the stack.

Second issue, there appears to be a que-ing, or scheduler in Sketchup? If I call Sketchup.undo followed immediately by my undo, the entities have not been "undone" in time for my undo to find them. I hacked it up using...

Sketchup.undo UI.start_timer(0.1) { my_undo}

Any alternative?

Ah, I used Array.instance_methods but I see that it requires an agument of "true" to get the inherited and mix-in methods also. ( 98 methods )

(Array.instance_methods(true) - Object.methods).sort ["&", "*", "+", "-", "<<", "[]", "[]=", "all?", "any?", "assoc", "at", "clear", "collect", "collect!", "compact", "compact!", "concat", "cross", "delete", "delete_at", "delete_if", "detect", "distance", "distance_to_line", "distance_to_plane", "dot", "each", "each_index", "each_with_index", "empty?", "entries", "fetch", "fill", "find", "find_all", "first", "flatten", "flatten!", "grep", "index", "indexes", "indices", "inject", "insert", "join", "last", "length", "map", "map!", "max", "member?", "min", "nitems", "normalize", "normalize!", "offset", "offset!", "on_line?", "on_plane?", "pack", "partition", "pop", "project_to_line", "project_to_plane", "push", "rassoc", "reject", "reject!", "replace", "reverse", "reverse!", "reverse_each", "rindex", "select", "shift", "size", "slice", "slice!", "sort", "sort!", "sort_by", "to_ary", "transform", "transform!", "transpose", "uniq", "uniq!", "unshift", "values_at", "vector_to", "x", "x=", "y", "y=", "z", "z=", "zip", "|"]

I use SnagIt from TechSmith - it has an option to scroll windows to capture them. Very useful for long lists like that.

Bob

interesting... although it seems very cheesy...

the modeling isn't very impressive.

Hi Jim

Very cool, I'm planning to model a Fender stratocaster.

@+

I took a look in the model.options list, and nothing there looks promising (I didn't expect it to, since what you want aren't really model options. If I have a chance, I'll do some more checking tomorrow, but I don't expect positive results...

It has "issues" with certain graphic-heavy PDF files, and printing limitations. No frills is what you get...Foxit is bigger but handles more PDF types.

It is not a sort of perversion to use Excel as a 3D program? 😆
Seems a very strong plugin! 😮 😲 ☀ 😎

@kwistenbiebel said:

@fbartels said:

They just need to get that roof curving in multiple directions ...
Fred

Building cost X 2. (+ a damn good structural engineer) 😉

Kwist,

Probably but perhaps not, although I have a bunch of work ahead of me to prove that.

I've been plugging away on experiments with PVC pipe structures and although the progress has been slow I've been reasonably pleased with the results. I'd make faster progress with a bigger research budget but that might jeopardize my marriage. 😞

Fred

Just for reference, the following works for scaling along the Group's axis:

m = Sketchup.active_model e = m.active_entities s = m.selection tr = s[0].transformation trs = Geom;;Transformation.scaling( 2, 1, 1 ) s[0].transformation = (tr * trs)

but this scales along the global axis:

s[0].transformation = (trs * tr)

as does this:

s[0].transform! trs

Let me clear up a remark I made at the top of this thread, and even my followup posts.

When I said "that is not the topic", I thought I was posting to Lewis's thread about a Ruby vulnerability that applies to people who use Ruby on a server for websites.

However, this thread, your topic, certainly applies, and yes, you can do just about anything you want with Ruby.

Ok, good. I wasn't sure if asin was simply overlooked, or if you needed something different.

Documentation error. Page.layers should logically return an array of layer objects rather than the layers parent object, and it does. Just one of those cut-paste-forgot-to-edit errors, most likely.

You are welcome. I thought I had already done it. Thanks for the tickler.

Todd

if you select open and un-check "always ..." still has the problem ? usually this appears when MIME setup is wrong

Yes, I noticed LightUp was already in there on my system. Probably not a wise move.
It harks back to the problems people had when copying their entire Plugins folder from one release to another...with the Extensions like Sandbox not showing up etc.
The present separation allows Google to safely upgrade some of their own rubies and dlls...tucked away in the Tools folder...without the danger of people then reinstating older versions back when they copy over the Plugins folder.

The messageboxname argument works with MB_MULTILINE messagebox type, but not with the presumably more frequent MB_OK type, which is the default.

Bug. Annoying.