Without changing the concept or algorithm much, I'd like to see a three column version, with the left-most column being for all edits -- with speed smoothed out over time (time delay 30 seconds, average it out); the middle one being edits changing over 100 chars that aren't immediately reverted (time-delayed 1 min?), and the left column being edits changing over 1,000 chars that aren't quickly reverted (time-delayed 2 minutes?), and aren't by bots or huggle.
SJ
On Thu, Jun 16, 2011 at 12:40 AM, Ed Summers ehs@pobox.com wrote:
I've been looking to experiment with node.js lately and created a little toy webapp that displays updates from the major language wikipedias in real time:
http://wikistream.inkdroid.org
Perhaps like you, I've often tried to convey to folks in the GLAM sector (Galleries, Libraries, Archives and Museums) just how much Wikipedia is actively edited. GLAM institutions are increasingly interested in "digital curation" and I've sometimes displayed the IRC activity at workshops to demonstrate the sheer number of people (and bots) that are actively engaged in improving the content there...with the hopes of making the Wikipedia platform part of their curation strategy.
Anyhow, I'd be interested in any feedback you might have about wikistream.
//Ed
Wiki-research-l mailing list Wiki-research-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wiki-research-l