On Sat, Sep 6, 2014 at 6:33 PM, Erik Moeller erik@wikimedia.org wrote:
< Flow is a long term bet that an architecture of structured comments
will ultimately have fewer hard and fast limitations on how collaboration in wikis can work, and will accrue usability benefits very quickly (as it already has done, like faster posting and replies)
< due to its architecture ... some rebalancing of effort towards the short
term may be valuable, and may lead to interim milestones that impact users today rather than years from now.
Testing this by irreversibly replacing existing unstructured talk pages seems likely to be a hard transition and hard to evaluate, since it breaks workflows as it enables new ones.
Two less dramatic ways to test such a bet:
1. Experiment with annotation and inline comments
This is one of the oldest forms of curation; a rapidly growing area of knowledge production online; e.g.: one of the early and beloved features of G!Docs, and the central feature of Genius which has its own communities curating interleaved and overlaid knowledge.
MW currently has little capacity for annotation, beyond inline footnotes. An improvement in that area would be welcome. The annotation use case is also a bit better-defined – more universa!ly a thread of individual thoughts – than the broad range of uses for talk pages.
Over time I could see many current uses of talk pages, including the simplest and most common ones, shifting to inline comments.
2. Experiment with UI overlays on top of current talk pages where possible.
For instance: the way talk pages and their tables of contents and section headings are displayed, where links to "reply" are displayed, how signatures are generated, the way a textarea is presented for adding a new comment.
Similarly, font / whitespace / layout changes are general UI shifts, and could be tested now without changing the function and data models of talk pages.