- From: Andriy Tyurnikov via GitHub <sysbot+gh@w3.org>
- Date: Mon, 03 Feb 2020 01:38:37 +0000
- To: public-houdini-archive@w3.org
@nhiroki , well, sounds like you are aware of a lot of nuances. As a software developer quite often I find myself to busy dealing with consequences of trade-offs that I've made. And in such circumstances it is not unusual to see (obviously flawed) choice as acceptable one for wrong reasons. As current spec is a working draft, it seems like it is a proper time for being vocal about issues with current version of the spec. I thought the very idea behind "Working Draft" is to avoid obviously suboptimal choices. While I am obviously unaware of rationale behind current solution, I have to highlight that API and implemented lifecycle mechanics seems overcomplicated when compared to plain function call. And perhaps it is worth to discuss rationale behind current mechanics of worklet initialisation/lifecycle. And as there is no shortage of suboptimal design choices in history of web standards, I see need of being vocal. While I don't see easiness of implementation and spec change as a right goal here as present mechanics doesn't looks simple (how all this worklet registration dance could be simpler then function call?) and if we would be silent, web developers of the future will pay the price. If there is a need to do something to avoid problems of present draft solution - your input/ideas are very welcome, please know that you are very welcome to share your thoughts via public or private channels. P.S. I honestly don't understand how current mechanics of `registerPaint` and `CSS.paintWorklet.addModule()` may be perceived as 'simple'. To me it looks like people got lost in trade-offs, and we should help them by being vocal about obviously trivial options. -- GitHub Notification of comment by andriytyurnikov Please view or discuss this issue at https://github.com/w3c/css-houdini-drafts/issues/979#issuecomment-581204305 using your GitHub account
Received on Monday, 3 February 2020 01:38:38 UTC