On Thu, May 5, 2022 at 3:39 PM David Arroyo <dr...@aqwari.net> wrote:
> Go programs run on a single computer. What if a Go runtime was built that > ran Go programs across many computers? I don't think this would pose any problems to this topic specifically. It would pose many problems, but once you have them all solved, this topic wouldn't be any harder than it is today. In particular, in this scenario you'd already have to solve pointer-comparison and dereference and itables and function calls and once all of those are solved, you can just apply those solution to the arguments here. Which is to say, the Go spec already basically assumes that your program runs in a single memory space. > Or, put another way, what if a system architecture emerged where the > instruction access time varied so drastically across CPU cores that it made > sense to duplicate functions across cores' fast memory regions, so that the > receive operation in the above example actually received a duplicate copy > of a function? I will admit that closures with mutable data segments > already complicate such an optimization, but function equality would thwart > such an optimization altogether. > > David > > -- > You received this message because you are subscribed to the Google Groups > "golang-nuts" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to golang-nuts+unsubscr...@googlegroups.com. > To view this discussion on the web visit > https://groups.google.com/d/msgid/golang-nuts/f566007f-aa47-4aaf-bc9e-a0fa9dab1e62%40www.fastmail.com > . > -- You received this message because you are subscribed to the Google Groups "golang-nuts" group. To unsubscribe from this group and stop receiving emails from it, send an email to golang-nuts+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/golang-nuts/CAEkBMfE6iDgOpMuFs-2k4mQA5G0q-zWhSGO9_6eTH_3LMuew_g%40mail.gmail.com.
