@unknownuser said:

like your plugin,i will study it,

Thanks, but current release is only linehatches orginating from TIG's mentoring.
It probably won't help you.

I havent released tiling for Components onto surface yet since I thought it was to slow when scaling non-uniformally.

Aligning a Component to the center of a Square face and scaling it to fit is relatively easy. But not when face is tapered. Mostly common in spheric objects and bulges.

As you most certainly have discovered there is no API-method for tapering.
So that means when a face is not paralleogram(?) one has to go in and alter each tiled Component individually == make Component uniq.

Maybe you have found some nice solution for this, I don't know.

In my case I came up with a linesweep algoritm that scanned the face vertices and scaled the Component to be tapered. As mentioned going in doing this sweep and translating Components vertices X n.times is gonna be slow nomatterwhat.

If this could be taken externally in C perhaps would help, but still won't be faster than API methods. I reckon.. Kind of a Catch 22.

If looking on the positive side of things, since I'm starting to sound a bit negative.
It's better to sit and wait a couple of minutes than putting the Components there manually As long as Sketchup doesent freezes

Anyway, will be interesting to see what you come up with.

Interesting, regarding your trial downloads. That pretty much proves donation buttons won't get pressed often..

who owns what -> Thanks for the input. It made sence.

Still waiting for other peoples input on this. Might be somewhat touchy subject...

@unknownuser said:

download the utility and run it

Later, I'm not at that stage in working on that problem just yet.

@unknownuser said:

Of course running 2 versions will complicate maintaining the code.

True. Especially if you already have a finished product.
Guess it depends on the plugin code structure. Main focus will have to be on the
PRO version.

Anyway, I don't know anything about the subject, just speculating as usual

Plugin store model is a great initiative. Keeping commersial plugins here is a good idea.
It's not just headed for donation-based, but also fixed priced plugins ?
Havent read the fineprints..

Interesting topic. I'm interested too in this.

I have a hunch, relying on donations is waste of time. It's against human nature paying for something when you have the alternative not having to.
Yet we see topics people say they sent donations but I guess they are few.

Maybe Tig and Thomthom's input on this will help, since they made many plugins and have
donations all setup.

An alternative is perhaps to have a free-version and a PRO-version.

That way no time trial, and you will make more people happy giving back to the Community
whilst getting YOUR efforts payed for if user wants to upgrade to the PRO-version.
Now there has to be an delibarate balance of features so the user desire to upgrade VS they find your free-version interesting. That goes without saying...

That's probably what I would do should I ever get in that situation...

Regarding RBS protected code, how does that work actually? Havent even looked at that.
Don't one have to license the code? Just because one cannot see the code doesent mean it's protected more than it cannot be copied and pasted?

And then comes the question, since probably all in here at some point has borrowed code, or gotten help in this forum, can we use part of code provided in topics safely without the risk of getting sued when including it in our library? I'm not talking about stealing long methods, but solutions to problems probably only solvable in 1 way..

I'm guessing you are a more experienced programmer than me(cant see your code cause it's scrambled), but I believe that.

Since we have been working on similar things. I mean, if you want any snippets of code, ideas, thoughts or discoveries found out during tests during development of Hatchfaces. Let me know. PM me.

If you wanna do your own thing, that's fine too. I understand.

Better interface. Hm, If it can be avoided, it's easier.
But have a look at Thomthom's new widgets.
They might make webdialog business much easier.

Panelization Plugin..
You mean like this :

http://sketchucation.com/forums/viewtopic.php?f=323&t=38637&start=300

(scroll bottom)

BTW, the time it took to render was due to the fact that Components adapted scale to nonuniformal(Square) faces.

Transform! executes/calls the method wheras transformation = is assigning it.

tested your code and
transformation = , did give me a same result as transform!

The principle still is call method with object.method.

Oh, and to answer the question. In this case were dealing with Geom::Transformation object. So were not calling a method directly(I think..).So its more like

Group.Geom::Transformation = myspecial.Geom::Transformation

I hope I explained it correctly..

If you test t0.class you will see it's a Geom::Transformation.object

Floating Point comparison confuses the hell out of me.
Think you had a similar disucssion in an earlier post somewhere some time ago.

This is reassuring, thanks.

@unknownuser said:

SketchUp's internal tolerance is 1/1000th of an inch. Comparing two Geom::Point3d objects using == will use this tolerance.

Getting position values from a mathematical computation one should convert results to Point3d objects before comparison, right?

Right, I see... There is some quite complex code going on..

Actually I was probably too fast earlier. the API actually says:

@unknownuser said:

you can use the Array class in place of a Point3d or Vector3d as a way to pass coordinate values.

If you got duplicates Array.uniq! should do the trick though

can't you just do an Array.uniq! on the Arrays, or will that mess up the face construction?

I think there can be a difference in using arrays of Points coming from a math computation VS using real Point3D.class objects regarding the use of API methods and Unit tolerance, No?

Hi.

As Dave pointed out you will have to unzip the files to Plugin directory.
Let me know how it goes.

Hey wait! Did you put this together yourself ?

Did not wanna sound disrespectful earlier.

Anyway. GOOD JOB!

Working nicely now with Ruby, JS and CSS. Very calm look and nice to the eyes.

Anything dealing with custom made 3d pattern will have have to to some kind of boolean or intersect.

Sketchups API methods for doing thoose kind of things on a large entities Collection is just to slow and unpredictable on detailed meshes..
In fact it doesent really have to be detailed. A simple quadface intersecting near the edge-border of a face could cause trouble.

Been dabbeling quite a bit with that for Hatchfaces.

The best solution would probably be to build premade patterns library(like it looks like Stepháne is doing) and draw them within face boundries.
Easier said the done..

JSONPadding Will try that.
Thanks a lot!

I'm setting this aside for the moment. Getting errors in crome debugger.

"Origin null is not allowed by Access-Control-Allow-Origin."

Crome has reputation being very hard on applying the "Same Origin Policy" to local files.

But this still indicates that it could cause trouble for me later on..

Yeah..

Well I guess I'll stick to the regular defaults then. To much hazzle..

Can't navigate anyway with all these new colors

No. I did not understand it like that, and I still don't not know what you mean by "loading script element dynamically" and scr="JSON/myjson.js".

That would cache the whole file loading it in the first time anyway, no?

I'm sure your thinking of something more cleverly then I, so exuze me if I sound off.

I don't consider this being a problem, more of an interesting solution if I can have it working properly..

edit: you mean with "document.write". That's seams to be sketchy.?

Hi Aerilius!

@unknownuser said:

during a session
it is not likely all nodes from that library would get used. There could be hundreds+!
So it seams like waste of energy for the browser caching all that..

Like I said, there is no problem having it in a separate JS or even in the main js file.

I just thought I could get better performance that way, by not loading in the static configuration data unless requested for.
If keeping it in the js there would be no need to use JSON though..

There will be many Icons and SVG curve rendering so it all will add up. At the moment the GUI is not swamped with stuff so I can't really see the performance hit as it will get. Maybe I'm just over complicating the matter..
As usual the big bottleneck will probably be in SU somewhere..

@unknownuser said:

For local files it behaves sometimes different. (I also ran into similar issues, and got it working but now it isn't working anymore. I'll look it up and post back.)

Ahh. OK.. It's working for me though right now. Are you also writing?
I'm just interrested in reading files.

If there are any issues I'll just dump the idea, no biggies..

Thanks for the help, appreciate it!

edit: Btw might mention that I'm creating this JSON file myself, so there isen't a dependancy for a framwork or anything like that..

Looking forward to your ideas on the matter, AE

Is there any way to only target Ruby for style? It messes the highlighting up for the others..

Otherwise, quite cool.
Thanks!