One Simple Chart: how open source projects interact with users

Core members of an open source projects make an effort to interact with users through a variety of means. They give talks at local events (now moved online), they answer questions online, and they resolve issues identified by users (using tools like GitHub). As far as answering questions from users, over the past couple of years I’ve noticed more open source projects using Slack to interact with users. Previously, project contributors and community members were more likely to share information and answer questions on forums and mailing lists, and track and resolve issues using GitHub and similar tools.

Slack provides (new) users access to core contributors, and depending on the level of engagement, they might even get answers to their specific questions immediately. Interactions on Slack can sometimes have the feel of “real-time debugging sessions”. The downside is search is clunky and a lot of information remains buried within Slack threads. In addition, not all open source projects have the funds to use the premium edition of Slack to save their archives.

Q&A sites (Stack Overflow, Quora, etc.), forums, and mailing lists are indexed by search engines, thus information posted on these channels are discoverable. Users who are fairly new to a project are able to discover useful information through their favorite search engine.

I recently ran a Twitter poll (which drew 676 respondents) to gauge if my hunch was correct. Sure enough, about a quarter of respondents (24%) stated that they used Slack or Gitter to interact with their users. In contrast only one-seventh (14%) stated they relied primarily on forums or mailing lists. However the combination of forums, mailing lists, and Q&A sites – indexable and searchable channels – combined for 26% of all respondents:

There is an opening for tools that combine the best of these mediums: immediate and personalized interactions (Slack) ➕ search/SEO/discoverability (forums; mailing lists; Q&A sites). I hope someone builds such a solution!

Scalable machine learning, scalable Python, for everyone: Join the creators of and core contributors to several leading open source projects including Apache Arrow, scikit-learn, spaCy, Horovod, Hugging Face, Spark NLP, and more → Attend the Ray Summit, a FREE virtual conference which takes place Sep 30th and Oct 1st.