WebPlat charter: Include retrospectives

[LJWatson]

Feedback from the AC review: It would be good to include retrospectives on previously published specifications, to understand whether they have been useful to the community.

[LJWatson]

Retrospectives are useful, but it isn't clear what the metrics for gauging usefulness to the community might be?

The method for conducting the retrospective would also need to be considered.

[edent]

There are a couple of (admittedly loose) metrics that could be used.

• Browser adoption. For example, srcset isn't supported in several prominent browsers. Is that a problem with the spec - or something else?
• Impact. For example Ink Markup Language was published in 2011. Has it had the impact that the authors wanted? If not, what should have been done differently.
• Security fallout. For example the Vibration API has been well adopted, but has continually been a target for spammers and has some security implications. Given what we know now, what would the authors change?
• Community adoption. It might be interesting to work with GitHub to see if their users have started using specifications. For example, EmotionML was published in 2014 - but today there are only 70 files in the whole of GitHub which use the appraisal-set attribute. Most of those seem to be referencing the specification.