wbthomason

vor 4 Jahrendiscuss

I have similar feelings toward ROS, but I don't have experience with industrial alternatives. Would you be willing to elaborate (insofar as you're allowed to) about what made the "step-function" level of difference between ROS and the internal systems you've used? Or on what you would want to see in such a next-generation system?

wbthomason

vor 4 Jahrendiscuss

One small downside of Blaze vs. Eigen for robotics is that Blaze seems to lack the highly convenient geometry types and operations that Eigen ships with. This wouldn't be difficult to build on top of Blaze (and could lead to some interesting performance comparisons), but does present some additional work necessary to use Blaze in many robotics applications.

wbthomason

vor 4 Jahrendiscuss

Thanks, this looks promising indeed! Especially because it similarly supports custom scalar types: https://bitbucket.org/blaze-lib/blaze/wiki/Vector%20and%20Ma...

wbthomason

vor 4 Jahrendiscuss

Huh, interesting - this is news to me, as I use Eigen all the time/see it used all over for robotics. Is there a good replacement for robotics-specific operations/small matrices generally (I see some people mentioning DirectXMath?)? Or is the tradeoff just between spending the time and effort to write SIMD intrinsics yourself vs. lower performance but greater convenience with Eigen?

One advantage of Eigen's approach that I haven't seen mentioned here is that its templated design makes it easy to substitute custom scalar types for operations, which helps enable straightforward automatic differentiation and other such tools (e.g. I'm currently using Eigen to make a tracing JIT for computations like FK, etc. over scenegraphs).

wbthomason

vor 5 Jahrendiscuss

Totally fair, yep. Thanks again!

wbthomason

vor 5 Jahrendiscuss

This is an interesting read, thanks!

I'm a bit surprised by https://viralinstruction.com/posts/badjulia/#julia_cant_easi... - I've been considering embedding Julia (as opposed to LuaJIT) in a project I'm working on that requires fast, mostly numerical computation.

Based on https://docs.julialang.org/en/v1/manual/embedding/, this seemed fairly simple, and as though I could avoid repeatedly paying the compile time latency cost by just calling my embedded functions at least once before the performance-critical loop (to ensure that they're compiled). Is that not the case? There's definitely still the high memory overhead to consider, but (admittedly without having tried it yet), embedding Julia doesn't seem too terrible to do. Worse than e.g. compiling a shared object library or something, but not at all unreasonably hard.

comments