Update pota, implements throwing moveBefore

[titoBouzout]

This change implements the new "node state preserving api" moveBefore. see: whatwg/dom#1255

This is the set of changes I had to do potahtml/pota@51da4b8

Unfortunately, they are making it throw., you check a ridiculous amount of conditions or just wrap it in a try and catch, I suspect the performance of the swap rows will degrade on frameworks implementing this. The benchmark unfortunately only tests 2 moves[when the framework does the optimal thing]. So I'm not sure if it would be representative of the degradation. I am still curious what the results would be.

Thanks!

[WebReflection]

imho, we should not implement stuff that has not officially landed yet or we'll make "their game" in doing so, when the "it's already shipped, too late to change it" card will be thrown next round of complains.

[titoBouzout]

The DOM is a dictatorship, this api, custom elements, you name it. They ask for feedback but only listen to each other. The more broken the thing becomes the better, because the sooner will get replaced for something better by some next generation of developers.

I was not made to deal with the politics of this bs. I put users first, they don't, I will use whatever they give me, accept trade-offs, workaround what I can, because I'm powerless, and I value more the users that want to keep things focused when stuff updates in the screen than the purity of their bs and smell of their 💩 apis.

They have strong opinions on problems that they admit they don't understand. whatwg/dom#1255 (comment)

Maybe that's why you don't see too many framework/library developers giving feedback in such threads.

[WebReflection]

I want to believe until it’s out it’s not as bad as it currently is, seeing lit, or others, already dogfooding that broken method name hurts back, doesn’t feel or sound as a win for anyone involved

[titoBouzout]

You are right, just reverted, thanks for your comments