RE: Defining mutation events

> We might learn something important there, I wouldn't be surprised if there are things missing from the web platform.

I suspect many of the usages could be solved by adopting a custom element and switching the mutation events to the various custom element callbacks.

As an aside, at a previous web components face-to-face, I was pretty convinced that generalizing the custom-element callbacks to any element was a bad idea because it essentially propagates mutation event-like patterns, and we weren’t convinced those were needed elsewhere (though arguably the timing for custom element callbacks is “better/safer” than mutation events). Is this still the current thinking today?

From: Philip Jägenstedt [mailto:foolip@chromium.org]
Sent: Tuesday, August 23, 2016 3:18 AM
To: Olli Pettay <olli@pettay.fi>; Anne van Kesteren <annevk@annevk.nl>; Elliott Sprehn <esprehn@google.com>; Dominic Cooney <dominicc@google.com>; Domenic Denicola <d@domenic.me>; Ryosuke Niwa <rniwa@apple.com>; Travis Leithead <travis.leithead@microsoft.com>; Rick Byers <rbyers@google.com>
Cc: www-dom@w3.org
Subject: Re: Defining mutation events

I think we need someone to own this issue for Blink, and the DOM team led by Dominic is the obvious candidate. Dominic, do you have someone who could take a look at the situation and come up with a plan forward together with engineers from the other engines?

FWIW, I'm skeptical about deprecation messages being able to seriously move usage, or at least I don't think it can be the only tool if we're serious about attempting a simplification or removal of mutation events. At the top of my list of things to do would be to find big sites or frameworks that use mutation events and to understand what it would take for them to migrate. We might learn something important there, I wouldn't be surprised if there are things missing from the web platform.

On Tue, Aug 23, 2016 at 12:10 PM Olli Pettay <olli@pettay.fi<mailto:olli@pettay.fi>> wrote:
The use of mutation events has gone down significantly and if we could get browsers other than
just Firefox to warn about use of mutation events, the usage might go down even more.
All these deprecations can't work if UAs don't warn about use of such features.


-Olli



On 08/23/2016 11:23 AM, Anne van Kesteren wrote:
> Elliott, you mentioned a few times that you want to slightly change
> the timing of mutation events. Is that still in the cards?
>
> I posted a summary of the current state of things around mutation events here:
>
>    https://github.com/whatwg/dom/issues/305

>
> It seems like browsers implement different subsets, so there might be
> some room for simplification. But it doesn't seem that the use
> counters are low enough for removal for a couple of them so we'll
> either have to experiment with timing or just suck it up and define
> them as-is.
>
> Thoughts?
>
>