[Bigbang-dev] activity in mailing lists and correlation to meetings, draft publications

[Nick Doty]

Hi all,

> On May 22, 2020, at 1:42 PM, Sebastian Benthall <sbenthall at gmail.com> wrote:
> 
> Thanks for the productive call, everyone.
> 
> Here are the notes, copied from the Etherpad.
> I look forward to being in touch with you as things progres..
> 
> Best regards,
> Seb
> 
> NEXT STEPS: RESEARCH
> =====================
> 
> For the next phase of Article 19 supported work, new and earlier BigBang functionality will be used in new Jupyter notebooks answering research questions of community interest.
> 
> What are the most pressing research questions?
> 
> Solicited from the meeting attendees:
>     - From J.S:
> 	• - Correlation between activity in mailing lists and drafts produced by working.
> 	• - 

Regarding this style of research question raised by Joey, I had similar questions about whether mailing list activity would track closely with publication of drafts of W3C specs or with W3C WG face-to-face meetings. There isn’t an API/machine readable data source in the same way that IETF has its Datatracker, but I was able to import dates for f2f meetings and draft publications for the Tracking Protection Working Group (which I worked on and study) and create these timelines:

https://github.com/npdoty/bigbang/blob/tpwg-timeline/tpwg/public-tracking-activity-timeline.png
https://github.com/npdoty/bigbang/blob/tpwg-timeline/tpwg/public-tracking-activity-yearly.png

Just looking at those timelines qualitatively, there are some connections to meetings, where there’s more mailing list traffic leading up to a meeting or following a meeting. 

The publications dates seem to have less connection. In some ways that makes sense to me: Working Draft publications at W3C are supposed to be heartbeat-style (it might be that I-Ds are supposed to follow that style as well) and published somewhat regularly as snapshots. “Last Call” or document transitions (to Candidate Recommendation, for example) are more notable steps in the process, but those also aren’t especially dramatic in terms of discussion activity: for this group (and this should be common at W3C) the issues are being hashed out by the group before those milestones, and while there may be some public comments and responses to public comments afterwards, that’s generally less active because the WG has already discussed the controversial issues by then.

I’d be interested to compare with anyone else doing this kind of analysis, especially if the IETF Datatracker can provide it for many groups at once.

Thanks,
Nick