Pepijn de Vos
Yes, npm gave several warnings for packages that want 0.10, you command "fixed" it.
Maybe the watch function should pass the wrapper reference?
The verilog backend in that commit (which I do not claim to fully understand) does seem to "shim" over the lack of flooring division in Verilog, so I think it...
Probably not...
Difference 2 is handled in Clojure by having [partition-all](https://clojuredocs.org/clojure.core/partition-all) and [partition](https://clojuredocs.org/clojure.core/partition), where base `partition` maps to Clojure's `partition-all` behavior. Difference 1 & 2 are grouped together in Transducers under [Consecutive](https://juliafolds.github.io/Transducers.jl/dev/reference/manual/#Transducers.Consecutive)...
The transducers equivalent of this seems to use `Val` in a wrapper constructor that supposedly gets inlined. https://github.com/JuliaFolds/Transducers.jl/blob/2a51f8dbd6b4408063e52108c3e36006f63b6b0c/src/consecutive.jl#L41 I currently have a custom PairIterator that I would gladly replace with...
It seems you could just wrap `SlidingWindow` in `takenth` to cover the full `partition` power at several magnitudes better performance.
I have written something similar in the past and would really like something like this to get merged. It seems to me the fa & fb are a bit extra,...
Double32 is actually really great for GPU where doubles are 64x slower than floats on consumer hardware. But that's a bit besides the point of this bug which happens regardless...
As far as I remember, yes. But the point of this bug is that length of a range is a BigInt, on both Double32 and Double64