Hey All, I've had some discussions with people over the years and have met people with all kinds of varying opinions about Heron and it's usage. A few opinions that stuck out to me:
- People typically were not as excited about Heron as it had the same type of API for creating jobs as Storm. To them, they might as well use Storm. - The underlying architecture and enhancements that Heron brought along with weren't that important to most developers/development teams, etc. Don't shoot the messenger, I'm only relaying things I've heard from others. With the above being in context, does anyone think it is worthwhile resuming effort on the streamlet API? Maybe add value to Heron elsewhere as Nick mentioned concerning metrics? Something else? We have a select few convenience binaries ready for public consumption once the vote passes here and in general@. Those convenience binaries are - install scripts - maven artifacts for creating jobs - an official Apache Docker image Our first release of convenience binaries in the incubator is a huge step for all of us and will help with lowering the barrier of using Heron. What else do you all think we could add to Heron that would increase it's adoption? What would make Heron stand out from the rest? - Josh
