1. Just finished Laputa, Castle in the Sky again. It may not have the strongest plot out of all the Myazaki films, but it does have, in approximate order of appearance:

    • Crazy awesome airships
    • Mysterious crystals
    • Flying pirates
    • Bad guys who wear dark glasses at night, inside and underground
    • Pseudo-Welsh miners
    • A hermit who talks to rocks and has a beard
    • Robots. Lots and lots of robots. One of them looks after birds nests and picks flowers. Others destroy things with their crazy laser vision.
    • Stylised lightning which looks like dragons
    • And at the end: a flying island which looks like a giant jellyfish, floating gracefully into orbit

    The robots look like this:

    My point being that anyone who demands any more than this from a film probably doesn’t deserve it.

  2. Ben Werdmuller: @barnabywalters Btw, I don't actually see datetime format guidelines in the mf2 spec. Hoping moving to the time attribute helps. 8m

    @benwerd explained here and here — essentially <time class="dt-*" datetime="yyyy-mm-ddThh:mm:ss±hhmm">human representation</time> is the preferred format.

    Sorry about the messy state of the mf2 docs at the moment, we’re working on it!

  3. @benwerd loving your work on idno! Just had a look at the source, great that you’re using 2, I have some suggestions/corrections:

    • .h-entry is better off where you’ve got .idno-entry so then the author .h-card can be scoped into the entry
    • add .p-author to the .h-card for each .h-entry to explicitly declare authorship
    • put .h-as-* on the same element as .h-entry .idno-entry
    • put .u-url where you currently have .dt-published, move .dt-published to the time element

    Thanks to Aaron Parecki you can see how a page is parsed here, or use my php-mf2 demo sandbox for experimentation by hand.

  4. Aitor García Rey: In fact, I’ve been playing for a very long time with the idea of creating a service for lowering the entry barriers to use personal domains.

    @_aitor what are you up to on the 22nd/23rd? We're remote participating indiewebcamp.com, it'd be great to talk about this stuff with you, drop into the office if you're free?

  5. @scottjenson RE google maps, I hear you. This particular problem could be solved by an app which remembers your speed, then displays the concentric rings. It assumes internet access/cached maps, GPS data and a device capable of displaying it — what if the device transmitting the information was a pedometer/similar which knows my speed but not location, has no internet access or way of displaying maps?

    I’m a fan of more ambient approaches like this because they enhance my own senses (in this case my poor sense of timing) without trying to run my life, as apps seem to want to do. I see it as a fundamentally different approach; apps make me perform a task and give me output. Ambient information enhances my senses and gives me more context within which to make decisions.