The Implausible Impossible, the Plausible Impossible, the Plausible Possible and the Implausible Possible.

Like conscious competence/known unknowns — there’s a pattern emerging here.

I’m noticing a #taproot pattern emerge whilst writing the simplified auth code: multiple event listeners which don’t know about each other working on the same object, augmenting and changing it.

E.G. RememberMeListener looks for an encrypted cookie with a URL (my user ID of choice) in — if it finds one it makes an ActivityStream person object and puts it in request.attributes.user.

Then, in the same event chain but at a lower priority level, the Contacts module looks in request.attributes.user for a URL. It looks up the URL in my people DB and, if there is anyone, augments request.attributes.user with all the extra info (full name, roles, photo URL, rel value, etc.)

Then, another listener could run, looking for request.attributes.user with only URL — and look the URL up on identengine.com, caching the response.

Other example is @-name autolinking, working on a similar basis of: basic transformation (raw data => common data format), then progressive augmentation adding URLs, names and rel values.

I think this a very powerful and flexible pattern and something I will make a founding principle of Taproot.

I’m specifically looking for examples which efficiently present post metadata. Feel free to mention any which come to mind!

People: I’m building a library of blog header patterns. Tweet me your blog URLs and RT!