Re:ModelingToolkit.jl, an IR and Compiler for Scientific Models
by ChrisRackauckas on Friday May 17, 2019 @08:25AM
(#19938816)
Yes, you would need a commercial license for suite sparse I think, so that would be a good reason to rewrite its core methods to a new license.
Re:ModelingToolkit.jl, an IR and Compiler for Scientific Models
by new4thaccount on Friday May 17, 2019 @11:43AM
(#19940658)
Thanks for clearing that up. I think someone else on here pointed that out as well as the fact that Matlab and some other tools provide that for you as part of your license/subscription fee. I started digging around and found out that it seems like the suitesparse author is one of the top authorities in this area and therefore, it might not be super realistic to get something as fast or faster on your own. Maybe a big enough open-source project could do it though. Any thoughts on this?
I'm really excited for what y'all have coming in the future and will stay tuned. There is some literature on Newton-Krylov solvers for my domain (power systems) so I'll have to brush up on that math.
Help me understand one thing. I suppose if we could one day compile Julia to a binary and sell software (just curious...I don't sell software today), would you have to secure a license for suitesparse? Is that another benefit from rewriting?