Brandon Simmons

Hi Luke, that sounds like a cool project; definitely please report back if you can think of anything you think should be added to the benchmark suite. Also, this is...

A good place to start might be to try to copy some of the more interesting benchmarks used in **The LMAX (don't-call-it-a-queue) Disruptor™** [here](https://github.com/LMAX-Exchange/disruptor/tree/master/src/perftest/java/com/lmax/disruptor), as it sounds like they've put...

Hi qrilka, thanks for reporting. Yes, I think your understanding is correct; a successful `read` operation does not write to the array segment to remove its element, so while the...

> One thing we can certainly do is document this better and add another `read` function that clears out its cell. Oh actually this makes `dupChan` impossible, so we'd need...

I might not understand what you're asking, but yes that comment is correct.

Oh okay, so all of the variants so far use a linked list of array segments. These are freed by the garbage collector. An OutChan is a counter + a...

I've started on a new variant in which reads don't block (related to #3). This will also support a much faster path for single consumers (SC), as well as pre-interleaved...

Per [this post](http://www.haskellforall.com/2014/11/how-to-build-library-agnostic-streaming.html) Gabriel pointed out that my stream type is essentially `ListT IO a` from `pipes`, so it should be possible to make the `streamChan` function agnostic to the...

Actually I won't be including streaming library -agnostic streaming in the next release; waiting on Gabriel439/Haskell-Pipes-Library#131

> (is there a class for types convertable to ByteString?) So what this is _almost_ just `Builder`, where we could use a [BufferWriter](http://hackage.haskell.org/package/bytestring-0.10.6.0/docs/Data-ByteString-Builder-Extra.html#t:BufferWriter) to write the buffer to our array,...