I use R multiple hours every day, I love it, I love the ecosystem, but I can't help thinking that it's showing its age. It does a great job on static analysis a

I use R multiple hours every day, I love it, I love the ecosystem, but I can't help thinking that it's showing its age. It does a great job on static analysis and visualization of smaller (< 1 gigabyte) data sets, but is seriously challenged by anything significantly larger, and is unfit for purpose if the data is changing rapidly (eg streaming). I unfortunately am slowly coming to the conclusion that Spark and Flink style tools are where data science will be at in a few years time, and while I know you can use R as a layer on top of these, I think other aspects of R also hold you back, paradoxically, things like the excellent base and ggplot graphics, which are rightly lauded as excellent, but are very low-dimensional in a world where tensors increasingly rule. I think R will remain hugely relevant for a long time, and is, what I tell people, like Excel^2, but it's getting to point where the world is moving on and it will struggle if it's not rewritten from the ground up with a much faster, multicore, threaded, distributed implementation.