Getting Started / Reality Check

Hi, I’m currently in my second year of 3D Animation, and beginning to research what to do for my final project next year. I’m really interested in the power of gpu accelerated programs, and I’m currently looking at how I could use opencl to make some kind of plugin for Maya. I’m interested in fluid simulations so maybe that would be something I would look at, or accelerating the viewports/improving the realtime shading of the viewports.

I was wondering if this would be a feasible project given that I currently have no knowledge of C / opencl (I’ve only coded in Lua and some MEL), and have a year from now to complete the project.

If so, where would be the best place to start?

TL:DR can I learn opencl and create a maya plugin in less than a year with little prior knowledge?

Thanks for any advice/help in advance

Hi kraftman,

I myself got interested into Maya plug-in developments at the end of my 3rd year of 3D animation class. At that time I just learned mel in a couple of months and feeled pretty comfortable with it. Prior to learning mel, I used to code a lot of mirc script (very basic scripting langage), very tiny visual basic and some tool book stuff. All of this are very very basic. Then, c++ arrived. To me, it felt like a colossus punch in the face over and over again. I got to say that I absolutely didn’t know at all where to start and I had nobody to help me. Each time I tried doing anything with it I always felt completely desperate and frustrated. I think that swimming in a pile of garbage would have been something more fun than continuing to try to develop my own plug-in in Maya back then. I don’t know if I can count the number of time I got so frustrated that I just smashed my bed trying to break it. I was very motivated though, so I kept on trying when I had the time (never full-time). Two years later, I think I could say that I’m starting to feel that I know where to start and where to look at when I need to develop a custom c++ plugin. I finaly ended up by writting the first version of mine, which was completely re-written maybe 5 times.

These days, I have the privilege of working on my plug-in full-time and I am starting to think that I might, in a couple of weeks/months try to write it in openCL.

If you compare mel to c++, mel is like a small baby and c++ is a giant brute ready to kill you and all that you love. Let’s imagine openCL now.

I can be totally wrong, but in my opinion. If things go very well and you spend one complete year on this on your own without anyone to coach you (that’s what happened to me), I guess that you’ll end up being able to do some basic maya plugins in c++ and maybe a hello world command in openCL? But that is not very happealing in a demo reel when you want to find a job afterward… especially if you don’t have a finished product that can you demonstrate. Anyway, that’s my guess. It’s up to you to make your own opinion now.