Yesterday, the New York Times R&D Lab announced streamtools, a web-based graphical tool for working with streams of data. It lets users create dataflow systems to re-mix and process the streams they believe most data will consist of in the near future.

Streamtools looks nice, but I do have some questions:

  1. What does it offer that Yahoo! Pipes didn't seven years ago?
  2. (How) have they addressed Pipes' shortcomings? (See for example the Stolee and Elbaum paper from 2011, or my post from 2011 about If This Then That for science.)
  3. Why would I use any of these systems rather than something like the Kinetic Rules Language?
  4. Why will people who don't use Pipes, IFTTT, or KRL (which is almost all of us) use streamtools?

This post originally appeared in the Software Carpentry blog.