michael_nielsen

(No criticism of Jupyter or @distillpub intended - I love both (& am on distill's steering committee). But both are, intentionally AFAICT, more oriented toward being widely-used platforms than being experimental media forms.)

To put it another way, the world described in your tweet has been here for decades in physics. And the pace of innovation in physics publishing is absolutely glacial. That's because that market is almost entirely broken and non-competitive.

As a comparison: I believe @github has done at least as much to change the publishing of open source software over the last 3 months as the arXiv has done over the last 20 years.

(Apologies. I really am ranting - I'd forgotten just how much the lack of competition and innovation in science publishing ticks me off!)

I see your point, but it is true that arXiv fairly recently started to experiment a little. Still conservative though. @inspirehep has done nice things to integrate data and code as first class citizens into the literature system. And ADS has also been innovating.

It's nice that they do that; the point is a systemic one. The rate of innovation in other parts of the world is 100x as high. And the fact that's not true in science is (really, really) bad for science, and bad for humanity.

Part of the reason for the systemic difference is that the market actually works in other areas. That means that capital is available to support innovation. The market is so horrid in science that arXiv - which generates billions (ultimately trillions, IMO) in consumer surplus -

is barely able to support itself. This is insane (and reminds me of the situation with SSL and the Heartbleed bug).