Retrospective - (the good, the not so good, and the puzzling)

Question

Our community is growing! We started our public beta in April and we now have ~2000 questions. As the Substrate/Polkadot ecosystem grows, so will this site. It is also worth noting that we currently have 250+ stale, unanswered questions.

This is a good time to do a litmus test on how we are doing and what we could be doing better.

(And remember, when I say "we", that means you the community as well - not a sub-set of members)

I invite you all to comb through our questions and ask yourselves what could be done better? For example, should we write additional help documentation on how to ask a certain type of Substrate/Polkadot question? Can you identify categories of questions (e.g. "IDK Rust" questions, node troubleshooting questions) in which we can retrospectively analyze and come up with solutions such as a template on how to ask a "node not working" question - e.g. the author must include their OS, Polkadot version, rust version, etc.

Please share your retrospective here along with ideas on how to improve.

Answer 1

My retrospective:

Overall I am happy with our progress. I see value for the community in the questions/answers we already have.

However, I do think that we have reached a point where we have enough data that we can start seeing some common patterns or themes developing. And I propose we start identifying these patterns.

For example, here is one I have identified:

• What to do with non-reproducible unanswered questions?

Would love for others to identify other patterns in which we can take some action on.

Answer 2

Hey all I have been working on an initiative to extract more data/trends/patterns from this site and I found an exciting way to mine data from Substrate Stackexchange (Stackoverflow): Look for our subdomain site (Substrate & Polkadot) here at: https://data.stackexchange.com/

You can make SQL-type queries over the questions and answers: https://data.stackexchange.com/substrate/queries

This will be helpful in technical issue discovery and addressing issue commonality. It would be great to go treasure hunting for good trends and ideas.

For example, see this one I generated to find the "Most Controversial posts" on Stackexchange (Sum of up and down votes respectively):

Answer 3

I think the question base is starting to mature. We probably need to cull tags with only one question at this stage. I see a steady stream of new users asking first time questions which is encouraging. Personally I'd be ok with allowing subjective questions that don't have just one answer. As long as the question is respectful I think it should be entertained but I realise I seem to be in the minority here. Maybe these types of question should get redirected to the polkadot forum as we close them?