Jump to content

DomC

Member
  • Posts

    646
  • Joined

  • Last visited

File Comments posted by DomC

  1. @lgoodkind
    Can you provide picture and an Idea how the algorithm logic could be. That would be nice. If I adjust the node to grab randomly your diameters. The result is as following:

    image.png.9c49220186c1935d33f4d4aacc7ad99d.png

    still there is white space which could filled up with an enhancement of the algo. But this is quite complicated. Thats why this node is beta version and not finished.
     

  2. On 8/22/2021 at 1:39 PM, Raph said:

    Hi dom

    Thanks for your answers and your interest.

    first of all, did you update your "WorksheetInputExample_v2" file?

     

     

    I was able to reproduce the "freeze" on my computer also with VW 2020 and VW 2021. It seems it freezes by wrapping the Network to a Object Node before the code runs. So the Code is all right, it is the Wrapping-Technique, which produces a problem. I will submit a Bug for that.

    I deleted the empty pass nodes and deleted the name of the Wrapper of the Worksheet Wrappers. Those Wrappers and nodes produces empty outputs in the network, which maybe is results the appHang.

    New Version 1.0.1 of Worksheet Input Example Created for Vectorworks 2020/2021

  3. OK
    Just tested that application.
    1. Import units should be set in the settings to mm
    2. The sheet(board) must be added or imported by dxf

    So far, then it works. It seems to be a try and error algorithm, because it uses quite a while. Would be nice, to have a tool like this directly implemented in Vectorworks (also something, that can be handled by script/marionette)
     

    Bildschirmfoto 2021-08-09 um 18.32.01.png

    Bildschirmfoto 2021-08-09 um 18.44.44.png

    Bildschirmfoto 2021-08-09 um 18.45.27.png

     

     

    Source code nesting algorithm, seems to be java:
    image.png.9dbb209bd8779234ff1472d21ad7ac65.png

    • Like 1
  4. Cool

     

    Spontaneously  I would say, It has to go this way since we have no other Marionette:
    1. Unfold (Extracting Tool) Click on every shape and extract planar shapes
    2. Put them on screen-plane (Infopalette)
    3. Or alternatively there would maybe be a quite simple script to put every shape on a different layer (maybe not, because the pathes may be NURBS and they are not CNC-compatible)

    4. Next exporting DXF and (I don't like to say) import it in an other software to nest those polys. 
    maybe this one, looks free and looks at it runs on macintosh:
    https://deepnest.io/

    just downloaded it looks promising good!
    Edit:
    Seems that every DXF have to be imported one by one manually. Also it needs some settings (units). Did not testes, It did not work after one Minute.

    5. Some CNC-CAM-Software could automatically extract your 3D Model (export Parasolid)

    The really cool thing would to have a poly-nest similar to the rectangle-nest. There is open-source-code for this available so it would not be a project from scratch. Also this would allow us, for automating the complete process. That would be my next project if I would find the time because that is also a kind of project that is interesting for me.

     

  5. Hello
    Aha, you really take the Original File. So the issue could be related to the following:
    1. Platform (Mac Win)
    2. Country-Settings (dezimal seperator is point here)
    3. Vectorworks Version (I have German SP4)
    4. Program Settings (more or less Standard)

    Ah see the issue while I am writing here. On my windows computer it hangs since over a Minute. Seems to be the issue can be confirmed on windows here too. Thanks, I will update for any news.
     

     

  6. On 8/6/2021 at 9:44 AM, Raph said:

    Now I have a problem with the "worksheetInput_Example" file.

    This File "WorksheetInputExample_v2" (without any changes) runs without issues here in 2021. Can you attach your File here?

  7. On 7/22/2021 at 11:00 AM, MJU said:

    I also tried the standard 1.03 for VW2021 and noticed that all the Spaces get the same size, while the attribution works fine

    Ahm ... good found
    I provided a new Version (1.0.4) which handles editing of the space boundary different and this seems to work (hope for many years :-)

    • Like 2
  8. 20 hours ago, MJU said:

    Hi DomC,

    first of all thank you this is great work and so useful. From your point of view would it be possible to modify the script in order to create rectangles with Data Tags?

     

    I'm trying to follow a more 2D oriented hands on workflow because with lots of changes and adjustments spaces get laggy and messy very easily. But I still don't want to miss out on your great script and the ability to analyze the data.

     

    Would be nice to have a qualified opinion on this approach.

    Thank you. I wonder, why the most usefull things always have the lowest rating here 🙂

    Using Rectangles instead of Space PIOs could make definitely sence.
    To have rectangles instead of Spaces you could modify the script as following:

    1. Create a Symbol with a Rectangle inside and link it to your record format. So you save that part of setting record to objects and tagging objects by script.

    You can Tag the Rectangles at the end with the "multi-Tag-Methode"


    2. Modify the brown Wrapper like this. It would result in a much more simple script

    image.thumb.png.2e0044a9ff919da3bfa82b29f120afdf.png 

     

     

    3. Then you have to edit the "Set Record Field" part of the script with your own record name instead of "Space".


    Other Options also possible but I think this is the most easy workflow. 

     

     

     

     

     

     

    • Like 1
  9. On 12/7/2020 at 10:26 AM, t.aichele@freudig.com said:

     

    Hi Dom, thanks for the good puppet! I am currently using this on doors. Can it be that this puppet only allows a shift from left to right and from bottom to top? Furthermore, I cannot calculate correctly with the script, which math functions or symbols can I use for them?

    The actual Version, is able to shift from right to left, top to buttom also. +-*/ and all parameter input names should work.

  10. Hi
    You could directly connect the xlsx reader instead of csv reader, this works perfectly.

    483867189_Bildschirmfoto2021-04-18um22_45_38.thumb.png.f1db872cd754017ca4b6257c39fe91fb.png

    I would first try this one because if you have an excel File this would be the direct way (If it works, because text encoding is a real pain). By the way your worksheet for me looks like an encoding issue too 🙂  Maybe not just islandic?

    Also the First version v2018, uses a Vectorworks Worksheet for input. This still works and you could just connect the worksheet reader instead xlsx or csv reader. 

    735736101_Bildschirmfoto2021-04-18um22_50_37.thumb.png.f21f91d6dd5eb0711b978ff3dcd28d80.png

    • Like 1
  11. 12 hours ago, t.aichele@freudig.com said:

    Can it be that this puppet only allows a shift from left to right and from bottom to top? Furthermore, I cannot calculate correctly with the script, which math functions or symbols can I use for them?

    1. Shift the Geometry by half of the Parameter, so the object would keep his middle
    2. For Calculation Extract the Input Parameter with Get XY. Also to could feed the input Parameter with Point 2D If you don't want to show Parameter Name
    image.thumb.png.1476ff4a56179c6e87628329eacb1490.png
     

  12. On 10/15/2020 at 9:55 AM, elepp said:

    Hi DomC,

     

    great node. I even managed to attach an ifc-record to the object. We encounter though one problem. We tried to use a symbol instead of group as control geometry. It worked for rescaling it but interestingly enough it looses its graphic attributes. In the screenshots you can see the marionette object in white and inside the control geometry you can see the original colors. Any idea why that is or how to solve it?

     

    Thanks in advance!

     

     

     

    Hi @elepp

     

    Hi If you put the complete Symbol into the Node, it will "flatten" the geometry inside of it.
    You have to Make a blue Symbol. Then attach the IFC Date to the PIO container with the Parent PIO Handle. Like this:
    image.thumb.png.a11dc6a7cd9be36283ac8a9c78deba36.png

     

    If you just want to have the Object is an IFC Object without any Detailed Informations, you could map IFC Informations with the Datamanager class Object. 

  13. Hi @Niels Timmer

    The input of the node has to be a group. You don't need to ungroup before.
    image.thumb.png.c65ddfcb6bc0476698bcf5f9a27db7f1.png

     

     

     

    The Issue seems not bei have a solid Operation inside of the Control geometry. Seems the issue is, that the Node does not handle the "One Big Volume" of the vertical Frames by reshaping the z- direction. I think it is a "Bug" in the Node or a "Limitation". I would suggest the "Workaround" to have two seperate vertical Frames (extrudes, Volume Substraction etc. does not matter). I am not sure where the error is. I am sure, that I have examples with Glass inside and also this is "One Big Volume" so it should work but not in this example. Also there seems to be different Versions around of the Node. I will look into the node and maybe provide a new Version maybe in some month, when I have a little more time.

     

    image.thumb.png.84f319f97f7364198d934464dfedcc5a.png

    • Like 1
  14. Hi
    First lets talk about what is an Object List? 

    A List in Python ist like ['Symbol1', 'Symbol2', 'Symbol3'] Used by the any input it will produce one line for every symbol name. This ist the input, the symbol node wants

     

     

    image.thumb.png.aa46ec67d0f4dacc67df2a520d1c5177.png

     

    But yes, we can input a List of objects. Example:

    image.thumb.png.01068f78667b7d92f9b3103140b0dd37.png

     

     

    This node was made to produce over 15'000 furnitures at once from an external list of objects.
    image.thumb.png.c84b9eea3dd7687d2d3490aec861e313.png

     

     

     

  15. @Luisa

    I am back from vacation so sorry for late reply.

    Basically this Example do the following:

     

    1. Collect Objects in the drawing with the criteria node

    2. Read some informations of the objects and put them together with a viewport on different sheet layers.

    3. additionally it creates a title block with some informations about the object

    You can do this for all objects independent if they have a record or not.

     

    It is just an example. Because it is a Marionette feature you could enhance or change, what the script should do (Maybe group all spaces from one layer on the same sheet layer, maybe collect additionally informations from objects which are in the space or on bound of the space, etc. etc.)

     

    To your question:

    All spaces are already linked to a record format (Datenbank).  

    Specific the Example with the spaces, uses the room Number for naming the sheet layers. Also it reads data for write them to the title block. 

    If your spaces have not the informations as the script excepts (No Room Number, No Name etc.) it maybe will not run. So if your spaces and informations are not the same type as in this example the script has to be fit to your space informations. You could share some of your spaces, that would help understanding if an issue appears

     

    • Like 1
  16. Hi Pier

     

    The Node used the outdated vs.GetType() command which do not work in 2019 and also produced wrong handles in 2018. New Versions uploaded. I fixed this long time ago but forgot to update this here in the forum. Beside that, this node is in use in several real live projects (as Example the fassade (Window Elements with additional frame and all ifc informations) below is made with this) 

     

    1049462788_Bildschirmfoto2019-08-24um22_19_05.thumb.png.77aab8715f00f8c43f48fc2f35d2b12b.png

  17. Hi Onyx

     

    It is everything possible, what we can also solve with Vectorworks functions. The Question here is, how to make views from a single furniture and hide all others. The standard workflow I recommend here is, to draw every single position on different layers. So we are able to work on every furniture/position undisturbed from the rest of the drawing. Also we can rotate them in a front view to work on it. Then make Design layer viewports from this furniture and assemble it on a master view. That's the way I would structure it. 

     

    From your structure it maybe could be solved as following:

     

    1. Different classes for every furniture. Maybe automated class-naming with the position descreption or furniture name and then Automated layout with class visibility.

    Or

    2. Sections around the furniture. But I am missing my KnowHow how to make sections with a script and also the issue is, that the sections will cut the furniture which is beside the required furniture.

     

    So #1 or a better Idea could be a base for a Marionette automation.

     

     

     

     

     

  18. @Gescher

    Hallo

    Ich kann mir das gerne anschauen. Es gibt zwei Tabellen-Inputs. Einerseits die Teileliste, andererseits eine Plattenliste um für verschiedene Materialien, verschiedene Plattengrössen zu definieren und um zu sehen, welche Platten bestellt werden müssen. Wobei man auch bei der Lager-Liste Stk 1 verwenden darf, wenn man das nicht braucht.

     

    Also Board-List ist der Platten-Lagerbestand und Stückliste-Küche ist die Stückliste. Es könnte dann Fehler geben, wenn die Spaltenreihenfolge nicht stimmt, die Zahlen mit Komma getrennt sind (7,1 statt 7.0). Poste doch mal Deine Beispieldatei und Beispieltabelle.

×
×
  • Create New...