Hello James, Hello Al,

@James: The problem is caused by the onAdded function in the Concat2 node. Reloading a nodefile with a Concat2 node, causes a setup of the output to the default type numarray. It seems that onAdded is responsible for creating new nodes and also for recreation from disk. If a wire exist in the nodefile, the nodeeditor seems not to check it again. But If you delete the wire a reconnection isn't possible for pointarray and Objectlists. I hope the workaround in the node will do the job.

@Al: Could you please post the problematic files with the Fillet node - Fillet is very experimental and I have to investigate where the problem is. I will have a look at it in the next week.

A nice weekend to all
Karsten

Hi Karsten
here is the test file of fillet
I hope you will serve to understand the problem
Have a nice weekend
al

Hello Al,

the problems seems to come up with the v.0.85. The edge names will be stored in the node file, but only the first element/edge name will be read back. I found a workaround by placing a leading number, so the node editor don't suppress the reading of the complete array of edge names. Please test the version in the attachements.

Have a nice weekend
Karsten

Hi Al et el,

I have been playing with patternSelect to generate radial forms.
See attached. It is best to delink extrude when when adjusting input dials
and then reconnecting. Otherwise it will hang for some time.



Hi, Marco

patternSelect is a brilliant node. It would be useful to an open/closed pulldown
so that when "closed" the indices prior to start were added to end of array.
The work around is not difficult, as in my .nod file; but the inclusion would be
productivity enhancement. Also, as I have suggested before, it would be useful
to have a num array input for pattern (it accepts 0s & 1s)

Keep on noding
-James

Hi James,
> "...It would be useful to an open/closed pulldown...it would be useful to have a num array input for pattern..."

I hope to find soon some free time to put my fingers on PatternSelArry in order to work with num array.

See you,

Hi Karsten
I replaced costruction 2 with your last
it seems to run better .....
File opening the output of the second Fillet
it does not work... ( picture-1)
I redid the connections, and it works , but creates
the surface Brep source ( why ?)( picture-2)
Saved, and reopened it's ok....... (picture-3)
I think which you , maybe I should spend a little more time
for the exact functioning of Node Fillet,
which it is really very important-
I saw the new Offset node
I hope functions ,
then we resend-
Good work..Great Karsten
best
al

Hello Karsten
with input curve works well ,
with line does not work ......
to be rewritten ????
Or there is something in the connections that do not understand
have a nice day
al

Hello Al,

thanks for the details. I had to change some internal things of the node to get it to run under that version. Sometimes I have to fight with some sideeffects and it is difficult to test all configurations, in which a node has to work. Some of the bugs took hours to understand them. Some nodes make problems again and again. Anyway ...
I redid the connections, and it works , but creates
the surface Brep source ( why ?)( picture-2) >>> Did you cloned the BrepNameSubObjects? If you do that, both will have the same basename for naming the sub elements. That can cause the problem. You can change it manually or create a new one for that. Moi's API isn't made for that stuff, so it is difficult to create that functionallity. Otherwise I saw also some unselectable ghost-objects while using the node editor - not selectable - like display artefacts - with other nodes. Is this actual problem repeatable?

I saw the new Offset node ....
I forgot to delete the offset node - it isn't finished - it works only in a first step. The flip for the side isn't stable. For test purposes you can show -with the frame output - the decission of the offsetfacory where the offset is placed. Ok - for a flip I have to calculate a frame on the other side of the input. But for a line the point is lying on the curve - typically you decide it with a pointer, but in the node editor isn't any interaction;-) These are the problems ...

Have a nice day
Karsten

Hello Friends
starts again week, continue the tests ....
a little exercise with input curves in Moi
and a little test SubD
best
al

Hello Friends
another test : Flow with Surface deformation Target
best
al

Hello Al,

(In Italian)
Stai diventando un drago con il NodeEditor!! :)

(In English)
You're becoming a master with NodeEditor!!

Ciao!

(Italian)
Marco amico mio , grazie
ma sono ancora piccole cose, rispetto a quello che si potrebbe fare..
Grazie sopratutto a Karsten, James, te , ed il grande Max per lo sviluppo
del sistema..
speriamo che tutto proceda per i verso giusto...
saluti
al
Ps- ho visto nel Forum , che sei interessato anche da Thinksolid...
Bel Modellatore sto iniziando ad usarlo anch'io
(English)
Mark my friend, thank you
but they are still small things compared to what you could do ..
Thanks especially to Karsten, James, you, and the great Max for development
of the system..
we hope that everything goes the right way ...
best
al
Ps- I saw in the Forum, you are also interested by Thinksolid ...
nice modeler I'm starting to use it too

Hi Al,

Thanks a lot for your kind reply!

> "Ps- ho visto nel Forum , che sei interessato anche da Thinksolid...
Ps- ho visto nel Forum , che sei interessato anche da Thinksolid..."

Hemmm.....how to say it....Sincerely I don't know ThinkSolid :)
Maybe you are referring to a different product or you're getting me mixed up with a different user,
or maybe I'm doing confusion... :)

Anyway...if you are referring to SolidThinking Evolve (SOLIDTHINKING) I can say you a couple of things.

1) I tried the product for about 1 month and I got rid of it without doubt.
Ok, very very good integrated renderer and interesting PolyNURBS tools (similar to T-Spline).
But, TERRIBLE, terrible UI and workflow!!
I can't imagine nothing worst.

2) Very bad forum and very bad support.

See you,

Hi Marco
Yes Mark , I'm sorry for the inaccuracy
I meant precisely SolidThinKingEvolve , and Inspire ,
I agree with you with regard to the interface
somewhat antiquated the WorKflow and also to
the Forum ....
But I find robust enough the Engine "Siemens" unless
mistake a derivative of NX or Parasolid
that responds very well, is also an interesting
variety of proposals surfaces, which when combined with PolyNurbs
in my opinion it makes the product interesting enough-

best , and good Work
al

Hello,

caused by the thread Placing spheres on knots/vertices and the very interesting skript fileToShape from libelle (many thanks for that!!!) I have made a draft for some nodes to read coordinate files back. (import/CSV-file - ArraysEXT/idxReplace)

Have a nice day
Karsten

Hi everyone.

I repeat here the same message that I wrote in the topic "Placing spheres on knots/vertices",
because this is the right place.

Here's the version v.0.85mod5 of "Project Elephant" : http://take.ms/Pxp6A

It's almost identical to version v.0.85mod4 posted by Karsten, with the addition of the "objectsExt.js" file which includes
the "SelectedObjExt" node that I wrote some times ago.

This node is an updated version of the standard "Selected" node, which allows to specify the order of selected objects : "By Selection" or "By Creation".

It's very useful in order to obtain different elaborations of the same data.

Best,

Hi everyone and good evening.

I'm planning to write a new utility node for the Project Elephant's "pointArray" structure,
but I don't know if already exists something similar within the latest official Max's v.085 version
or within the latest community version v.0.85mod5 (downloadable here http://take.ms/Pxp6A)

So, I tell you what I want to do.
First of all I say that this new node should be based on the great Max's "MathPts" node.
That is, in order to write this new node I will use most of the "MathPts" code, plus some new ad-hoc javascript code.

So, this shoud be the new node : "MathTrasformation" and here's in brief what it should do :

- Input :
     - PointArray
     - Math formula to apply to "X" values of each array element
     - Math formula to apply to "Y" values of each array element
     - Math formula to apply to "Z" values of each array element
     - Math formula to apply to "AngleX" values of each array element
     - Math formula to apply to "AngleY" values of each array element
     - Math formula to apply to "AngleZ" values of each array element
     - Math formula to apply to "pSize" values of each array element

- Output :
    - The resulting PontArray

I repeat, I don't know if already exists something similar, but I think that this "MathTrasformation" node would be a very very powerful tool
in order to do almost infinite kind of PointArray eleborations.

As I said many times in my previous posts, unfortunately my math skills are very very basic,
so in order to write this new node I will leverage only on my "creativity and intuition" :)

If any of you want to write this same node I will be very happy to share the results.

Have a nice day.

Max's Array3D script?

Hi bemfarmer and thanks a lot for the suggestion!!

Yes!!!
I mean definitely something very close to a special version of Array3D script translated into a node,
with the opportunity to include into every single math formula, a reference to the fields of the PointArray structure.

For example, this could be an hypothetical math formula applied to "X" values of each array element : "(sin(ptY) * AngleX)"

Where "ptY" and "AngleX" are special names that refer to "ptx" and "AngleX" field of the PointArray structure.

The node I'd like to try to write should include also the opportunity to use additional input fields like the "Basic->Math" node.

Thanks and have a nice day.

P.S. A little correction : Unlike the "Array3D" script, this new node, probably named "ArrayMathTrasformations",
will not generate an array of pointArray, but simply, given 1 input pointArray it will returns 1 transformed pointArray as output.

- Marco (mkdm)