In my experience, having released software is first and foremost necessary for establishing a community around code, followed by evangelism (for example: engaging blog posts, distribution of links to the blog material through social media channels, submission of links to aggregators (subreddits, hacker news)). In both cases a steady cadence would serve the project well.
Good documentation is important for sure.
Newcomers will be curious what problems NiFi is solving today where/if deployed, or is aiming to solve soon. As a prospective user I'd really like to understand use cases (do my problems look like that?) and as much detail as possible on how (can this description serve as a solution template for me?). What is the project roadmap? What is the vision for a year or two from now? (Where might we go together if our organization signs on as an adopter, and potentially contributor?) This is in my opinion good grist for further mailing list discussion, and later, material for above mentioned engaging blog posts.
As you have been doing, make all decisions and development transparently and in full view of the project mailing lists and issue tracker. Collegial consensus-driven decision making is attractive. The mailing lists should be a friendly place.
Consider finding public venues for spreading word about NiFi. Once there is a release consider hosting informal area developer meetups to talk about it. And/or reach out to related Apache projects and ask for floor time in their upcoming meetups. In terms of professional venue, the Strata family of conferences comes to mind as a perfect place to talk about scale data processing technology.
A user base has its own gravity. Early effort in guiding prospective users to success will have long term dividends. Committers on the project should team up / trade off to promptly answer inquiries and trouble reports that may come in on the mailing lists once people begin to try out project releases. Consider investing time to help prospective adopters get off the ground however your personal and professional bandwidth will allow. In the earlier days of HBase, my experience, several committers for a time were essentially running a no-cost consultancy to Bay Area tech companies, including on site work. Now years later I still see the resulting "X uses HBase" word of mouth attracting new users. This is an extreme of course, but no doubt part of our success story as a healthy and active Apache community.
> On Dec 31, 2014, at 1:26 PM, Joe Witt <[hidden email]> wrote:
> NiFi Team:
> Priority #1: Community building. Path to get there:
> - documentation on use, examples
> - documentation on how to build processors, contribute, etc..
> - getting a release out and establishing a rhythm
> If any of the apache experienced among you are interested in sharing
> opinions on what the key community building things to do are or have a good
> reference to point at it would be appreciated.
> Looking forward to a great 2015. Have a happy new year!