Marius Millea

I forget if this is Julia specific or Jupyter, but there's a different behavior if you click stop once vs. 5 times in succession within a couple of seconds. Does...

Added a barebones set of JLArray tests.

Ok, found the magic code which gives no allocation or speed regressions vs. current master for Arrays of any dimensions and which works with CuArrays, its this: ```julia map!(view(Base.ReshapedArray(result, (length(result),),...

Don't want to have this lost after the holidays, could this get looked at again?

> adding definitions of copyto! with Tuple to CUDA That seems like a strech to me with CUDA tbh. Although these things aren't specified perfectly, `copyto!` seems aimed to be...

Here's an issue with the `copyto!` option. How do you implement `extract_gradient_chunk!(::Type{T}, result::AbstractArray, y::Real, index, chunksize)` so thats its inferred and allocation free for Arrays? Closest I can come up...

I should also note that one nice thing about the `copyto!` option is that it would allow the specialized `copyto!(::CuArray, ::Tuple)` to pass the tuple data into the kernel by-reference...

Thanks, `fill!` doesn't have a variant with an offset / chunk size though unless I'm mistaken?

> No, not that I know of. But couldn't you fill! a view? No, thats allocating for multidimensional Arrays, which is the entire issue that led to the solution currently...

> I still think that it would be nice if we wouldn't have to use slightly unorthodox implementations, just to be able to add implicit support for another package. Yea,...