Twan Koolen
Twan Koolen
Yeah, changing things so that it actually does become `+(1,2,3,4,5,6,7,8)` doesn't look like a good idea. In any case, seems to me like this should just be a call to...
> Generally we've let Base and LinearAlgebra be precise and StaticArrays be fast Yep, sounds good. > if people felt strongly about it I definitely don't.
To what extent are these speedups due to using `muladd`? StaticArrays currently doesn't use `muladd` at all, and using it constitutes a tradeoff.
From `?muladd`: > The result can be different on different machines and can also be different on the same machine due to constant propagation or other optimizations.
I don't know if it was intended. I tend to agree.
Same with `update_settings!`. It would actually be quite nice to be able to set settings before calling `setup!`. In fact, if OSQP itself supported it, I would personally opt to...
`perf_read!` does a `seekstart` at the beginning, but `perf_read!` is not used for the `readstring` benchmark, only the `read` benchmark. So if the `read` benchmark is done before the `readstring`...
Howdy Dehann. > We PR `lcm-gen` to add Julia support to LCM itself Are you aware of https://github.com/ZeroCM/zcm/pull/163? > `convert(::Type{Vector{UInt8}}, ::MyType)` and `convert(::Type{MyType}, ::Vector{UInt8})` I feel the problem with that...
> I could maybe help out with the autogeneration of the LCMType code from an .lcm definition Sounds good. @rdeits at some point suggested starting off by making a string...
Any idea what's up with the OSX failures?