Must say, I find their measure a bit odd. I wonder about those cool new ordinal analyses we can do eg journals.sagepub.com/doi/pdf/10.1... (although I've yet to try bayesian...). Maybe a Friday night explore with obligatory Aldi prosecco.

Ouch, not sure their measure lends well to errors bars...

Pass over the r file? Ill have a look tomorrow if I figure out if this contains what we need d3nkl3psvxxpe9.cloudfront.net/documents/To...

Don't suppose you'd be tempted to add error bars? :) sorry about being pedantic :s

Nice one Tom

I made it so that I can access my 3d printer (wifi) and my regular printer (lan) but it's proven useful in other situations. e.g. when developing service workers.

Gosh, utterly relatable too

I sneaky checked out your talk slides -- speakerdeck.com/wsvincent/dj... . I've been pondering on this sort of thing, eg time to load ML model and share it across django instances, trying to use a GPU (gave up!!), trying to use multiple ML models (ouch). Tricky tricky tricky! :)

Hey Matteo. I'm intrigued by stationary attractors, from a perception perspective. Does this explain (pareidolia) why faces or figures are perceived in random or ambiguous stimuli? Hope cool to ask. We've just a paper out where we mention this.

Cheers. There's a curious OOP thing here I feel, where one can more easily swap out children/ancestors. Will have a play!

The similarly to html scares me a bit. I ponder if it's harder to 'spot' the Django

I don't quite understand the appeal. Is it the power of being able to pass complex content inside a node <node>complex stuff</node>

I feel that might allow curious composition. I must think.

I thought to give it a go. We submitted a paper with Multisensory Research, Bril. I suspect Bril owns Kodos, and fancies scraping just a little bit more money from academics. Beh. Deleting my account.