Sensitivity to community perceptions regarding use of "PyData" trademark
wesm opened this issue · 8 comments
While this initiative is important, keep in mind that given that "PyData" is a significant community brand (with conferences, etc.) that some people might see "PyData API Consortium" and ask "who put you in charge". By utilizing the mark there is a pseudoauthority that some may not be happy about. To be honest it makes me a bit uncomfortable personally.
In any case, it might not be the worst idea to consider giving a more neutral name to this effort.
@wesm very good point. We definitely need to get official permission from NumFOCUS, which is on my todo list.
I'd also be happy with a better name, it's hard to name things well. Thoughts that went into choosing this name:
- We'd like a name that captures what this consortium does
- This effort is focused on the Python numerical computing / data science / ML ecosystem. The umbrella name for that was "the SciPy ecosystem" earlier on, but with data science and ML becoming as important as (or even more important than) scientific computing, there was a widespread feeling that SciPy was too narrow of a name, and PyData was the best alternative.
- A while ago there were some discussions about writing documentation about the whole stack, as an entry point for newcomers. After trying some new domain names, the idea there was also to move it to the main page of pydata.org, and move the conference info there to pydata.org/conferences. This was positively received by NumFOCUS board members.
- We'd like this consortium to grow over time, become more organized and open. And ideally leave space for other similar initiatives (say scikit-learn et al would want to do something around ML APIs).
"who put you in charge"
Like with anything, it's more a "who wants to work on this" I think. As long as NumFOCUS (who owns the PyData trademark) is happy, then I don't see an issue here. Especially given that many leading people in the community are involved.
To be honest it makes me a bit uncomfortable personally.
Would that still be the case if NumFOCUS formally endorses the name, and we address the motivations, ways to join etc. in a blog post?
Would that still be the case if NumFOCUS formally endorses the name, and we address the motivations, ways to join etc. in a blog post?
Still might be challenging, given the "pay to play" nature of the genesis of the consortium (important: this isn't a bad thing, because developer time needs to be compensated to make this successful). However, using the name to create pseudoauthority seems fishy to me viewed from an outside perspective. Perhaps something like "Industry Consortium for PyData API Standards" would do better to express that the desire is to propose standards for the community's consideration, but that the group is not acting officially on behalf of the PyData brand. Or just "Industry Consortium for Python Analytics/Computing/Data API Standards" if you want to nuke it from orbit and not use "PyData" at all
Yeah if we're going to change it, perhaps better not to use PyData at all.
I'm not sure the word "Industry" is that suitable. That almost invites a "big business is coming in to take over" type of criticism. And really that's not what's happening here - this effort is long overdue and many of us are long-time community members.
How about "Consortium for Python Analytics API Standards"? Still long, but captures the essence probably.
That works for me too - slightly shorter, which is good.
I agree on a short and "generic" term, which we can reuse in the future if we tackle for API standardization efforts.
Strong +1 on avoiding the "PyData" trademark in favor of something more generic like "Python Data"
We again discussed this in the meeting on 28 May, and everyone was happy with Consortium for Python Data API Standards. So that's the final decision - closing. Thanks for the input everyone.