IntelLabs
Packing FLRC so that it can always be revisited in future benchmarks
Too few people, IMHO, work to make FP fast, so we should preserve all attempts for future comparisons! For example, we've been trying to collect Docker configurations that build such compilers reproducibly for microbenchmarking purposes.
The existing .travis.yml file seems like it could be trivially converted to a Dockerfile and then built and made available on Dockerhub. I can send a PR, but it will take some clicks on the part of a committer here to add it to dockerhub for automatic builds. (@RyanGlScott do you have commit access to this repo, btw, following your internship?)
This. Exactly this is why I worked with the original authors to gather all pieces that made up Reduceron and got it running (and have maintaining it since). The original sources would be a lot of work to get working today. (Sorry, I just had to chip in. FLRC looks interesting).
@DrTodd13 Giving @RyanGlScott commit access sounds like a good idea to me, since he had worked on this and can help maintain it. What do you think?
