Wander Hillen
Wander Hillen
I agree that high performance applications probably *should* use something other than large strict bytestrings and they do already mention that it'll take O(N) time. They don't mention that (for...
I was thinking more >512 MB or something like that :sweat_smile:. The crypto example is really very low since it's so cpu intensive even for fairly small chunks of data....
@dvandersluis I looked at the failure but it does not look like it should be caused by the changes in this MR. I tried reproducing the error on my machine...
I'll have a look at it this weekend.
Thanks for the help and guidance!
I rewrote the `IOBuffer` class in several different ways but did not manage to get it faster than the current implementation. I suspect that the bookkeeping in Ruby land introduces...
@MSP-Greg I haven't used the NIO class in particular but this kind of buffers/stringbuilders were what inspired the issue in the first place. As I mentioned above I could not...
I finally got around to benchmarking a few more variants: - Changing IOBuffer to just an Array of strings and then `iobuf.each {|str| fast_write(io,str)}` had only about 30% the req/s...
And `stream.to_socket`!
I came here to say the same thing, the error handling should be inside the method we provide IMO (in this case `streaming_get`). I do think we should at least...