Sebastian Markbåge

I find myself nodding along *a lot*.

If you're new to our industry, I highly recommend following @sarahmei. She writes in a very insightful way about counter-intuitive topics...

I have never seen a software patent that didn't have significant, but unmentioned, prior art by other people.https://twitter.com/eff/status/833353201369833472 …

PSA: when developers gripe about vendors "breaking the web", this is the web we're trying to keep: http://www.warnerbros.com/archive/spacejam/movie/jam.htm …https://twitter.com/nolanlawson/status/832779018822787072 …

My biggest successful trait as an engineer is the ability to change my mind quickly. It's also the most frustrating.

Leading to worse upgrade paths than the hard work was meant to prevent.

TBH the Node/ES module unification is running out of time. If it doesn't ratify soon, the community will solve it for us.

Still interesting...https://twitter.com/sebmarkbage/status/530618224946061312 …

I believe @stefanpenner found a similar but slightly different take on this problem too.

It shows up in virtual DOM libraries and many other cases. Creating a solid pattern matching/brand checking mechanism for JS is important.

I'd encourage you to do the same. Thank a browser engineer for the thing you didn't have to work around or could remove the work around for.

Debates about how things should be done sometimes cover up the hard work of little details that unlock large potential.

Thinking about writing a browser appreciation post of all the little improvements I've noticed as a lib author but don't get much attention.

Multicore OCaml has an interesting strategy to parallelism. Could work for JS too. Similar constraints. https://ocaml.io/w/Multicore 

I've made two small PRs to react since the addition of fiber. I'm pleasantly surprised at how approachable the codebase was.

And also vsync aligned rAF.

To Chrome's credit, it has the best low level support for 60fps in active listening mode. Thanks to rIC and off main thread rasterization.