@wuzelwazel So sorry for not replying back to you. I just had a look for it now (February last year), wow almost a year has gone by. Glad to see your still working on it. I will see if I can find some time soon to have another look as well. As far as the UDIM work goes I did something similar many years back. I extracted every texture for every triangle and wrote them to a new texture map. And as each triangle texture is extracted to the texture map I use its location for the newly created UVs. I used this for an OpenGL preview as well as rendering. Note that this is not baking at all. Since I am using actual ptex pixel image size directly for each triangle extracted to the texture. So the UV coordinates are based on the actual pixel size in the ptex images. As soon as one texture was full I created another with a new selection set for the triangles and its own material. I had planned on doing this again and converting the OBJ files to C4D files in the process, to make them smaller and faster to load. Which is what you are doing with Redshift proxies I believe? I will reply to your email as well so you have my contact details. Cheers, Kent

@PluginStudent Thanks for the suggestions. I will definitely use your super obvious (in hindsight) recommendation to check the emptiness of the original String! Could you encapsulate briefly why I might or might not want to take ownership of the CString array? I have a gut feeling that using the GetCString() method you suggested is safer but I have no knowledge to back that up Error handling is definitely something I need to better familiarize myself with. Lastly, when would I need to use the maxon namespace? I sort of assumed that these two would be equivalent: const String fileString = fn.GetString(); const maxon::String fileString = fn.GetString(); // significance of namespace here? Is it a situation where if String also exists in another namespace then not providing the explicit namespace would result in ambiguity for the compiler?

Just FYI: you find that information regarding ./res/libs/win64 in the documentation: Development for Microsoft Window

Super, thank you!!

@kbar said in Generator crashes and guidance: @zipit said in Generator crashes and guidance: OBJECT_DONTFREECACHE Yes that is the flag. If you enable that flag then you own the objects. And you pass the pointer as the result of the GetVirtualObjects call. But you MUST ensure correct deletion of the objects when the generator is deleted. And you must handle all read/write/copyto calls correctly for your generator. Thanks Kent for providing additional notes, but I think I need to clarify a bit about its real scope. The OBJECT_DONTFREECACHE allows generators to maintain their caches on their own which means that Cinema will not handle the cache deletion upon its generator object has been deleted. This flag is highly dangerous and should be used only in combination with a correct memory management design that secures against memory leaks. That said, this has no relation on the option to safely refer a BaseObject in the scene by directly storing or accessing its pointer. Again, safely accessing a reference to a BaseObject must be done via BaseLink. Cheers, R

@Cairyn Hi, Thank you. Solution: Float GeIncircle(Int32 n, Float r) { return r * cos(PI / n); } Float max_rrad = GeIncircle(n, r);

@r_gigante Thanks for this info. In two of my plugins I am already handling the TagData::Message to check for point and polygon changes (and other things as well). To my understanding, all this would be performed on the object remaining after the merge. What happens to the objects being merged "into" the remaining object, and especially the tags assigned to them. I am still wondering how to actually be able to merge the data from the other tags. Say you have 3 objects: cube. sphere, cylinder. Each has a custom tag assigned. To keep it simple let's say that the tag stores a user specified value. User has entered value 10 for the cube, 15 for the sphere, 12 for the cylinder. Now, when the 3 objects get merged (via connect or connect + delete or SendModelingCommand(MCOMMAND_JOIN ..., ...) the end result is the cube object receiving all the polygons and points, etc ... from all 3 objects. Internally the UVW tag is merged, etc ... The custom tag from the original cube remains as is, with value 10, and could react to i.e. MSG_POLYGONS_CHANGED in the TagData::Message. But ... how and where would I be able to handle the merging of values 15 and 12 into the custom tag ? Only when perform a SendModelingCommand(MCOMMAND_JOIN ... am I able to actively merge the values, as I know which objects will be merged, and thus can obtain the values from their tags before the joining is performed. With a connect or connect + delete, how to know about the other objects, and especially their custom tags? I don't expect this to be available inside a TagData::Message

Hello, as Riccardo has pointed out, AdditionalIncludeDirectories has nothing to do with DLLs. It is used to define search paths for header files. See Additional include directories. It was never valid. Also, looking at the error, the plugin was installed into the C:/Program Files folder. In R21, there is no reason to install plugins there; one can install plugins at any location and just tell Cinema 4D the plugin path. best wishes, Sebastian

@r_gigante Hi Riccardo, thank you for your detailed response. That allowed me to find the exact formula. l / (2 * tan((180 / n) * 3.14159 / 180)) Regards, Mustapha

Hello, Q2 : filterwhirl.cpp is a internal file you don't have access to it. And we don't have tutorials. (this comment will be removed to avoid confusion) Q3 : CallCommand should be use on the main thread, that's maybe why it's not working, you can use ExecuteOnMainThread in your job. Cheers, Manuel

@r_gigante Maybe a little late, but thanks for the info anyway. Always useful, and much appreciated! I went with a custom implementation using a BaseArray to store the information, and from there used the specific index in the array as the value in an HashMap Instead of the wanted: maxon::HashMap<Int32, DataStruct> mymap; I went with: class MyHashMap { ... maxon::BaseArray<DataStruct> mArray; maxon::HashMap<Int32, Int32> mArrayIndexMap; }; This is more or less what you have described. The only drawback here is that I still haven't found out how to implement range-based loop support ... but I went with using the old-fashioned indexed loop usage. Good enough for its purpose.

@r_gigante ok thanks! I managed a workaround by filling my dummy spline with two zero length segments at the claculated bounding box limits.

@m_magalhaes Hello, Thank you so much.

Yes because I've done it :)

Thank you @m_magalhaes! I needed to open the plugins.sln solution instead of the project file. I feel like I've learned something, but I'm not sure what

@m_magalhaes Thanks for bringing this up, as it seems I had overlooked it on multiple occasions. Which also points me to the fact that the R19 SDK did have a String::operator == () Also overlooked that all those years. I need better glasses.

@s_bach said in Communication between DescriptionTool and GeUserArea: The typical modern approaches are patterns like Model-View-Controller or Model-View-Presenter. In such models, you don't "send data around". Thanks for mentioning this. As stated in my opening post, this is what I have been using, where the "model" is a singleton class, available as a global variable to all plugins to synchronize and share data. Maybe not quite the exact wording. I was just wondering if there is another means of providing this "model" instead of using an instanced (singleton) class via a global variable? So in your example, the interaction with the DescriptionToolData would write data into the model. And the GeUserArea would read that data from the model. The example I provided isn't making use of a "model" as it already represented too much code just to point out the bug that slipped into R21. There I indeed used a very simplified solution by "sending data around". But an overall "thank you" for bringing up the model-view discussion, as this made me realize I could use an observer pattern in another project of mine which would fix its current poorly chosen design. Daniel

@m_adam said in GeUserArea lag in R21 versus R20: Thanks a lot for reporting this kind of issues, it's really relevant for us :) My wife is "jokingly" telling me to send in an invoice for all the effort I am spending in "beta-testing" this official public release. Not so sure if she really means it as a joke, or just being sarcastic ...

Hi, @r_gigante Thank you for your reply. With your help, I finally traced the source of the problem. As you said, the reason for the error(Critical Stop) is that there is no Container at all. Because of a very hidden error in somewhere, I wrote the wrong data type to BaseContainer.

@m_magalhaes I'll write it in your way Thank you very much. Cheers!