- From: Ryosuke Niwa <notifications@github.com>
- Date: Wed, 12 Aug 2015 17:55:31 -0700
- To: w3c/editing <editing@noreply.github.com>
- Message-ID: <w3c/editing/issues/66/130493164@github.com>
> We need to distinguish between those editing hosts which can serve as editing hosts for execCommand and those that cannot serve as editing hosts for execCommand. The one in cE=true can serve as such an editing host, whereas those in cE=events cannot serve as such an editing host. At least we cannot guarantee that they do. Right. > One solution A) would be to not use the term "editing host" for cE=events containers. Another solution B) would be to call them two different types of editing hosts. I am open to either one of those solutions or for an entirely different proposal I'd prefer B to avoid introducing two made-up terms that refer to very similar things. > True. Given that the ultimate goal is to eliminate cE=true, I think the best is to keep the definition as simple as possible. Something like I don't think we necessary want to eliminate `contenteditable=true`. We're adding `contenteditable=events` now because spec'ing the former is really hard, and we can't easily reach an interoperability but we should aim to eventually reach interoperability between browsers. In a way, you could think of `contenteditable=events` as "desugaring" `contenteditable=true`. In that case, spec'ing `contenteditable=events` and adding more primitives should help browser engines to slowly converge to an interoperable implementing of `contenteditable=true` over time (like in 10-20 years). --- Reply to this email directly or view it on GitHub: https://github.com/w3c/editing/issues/66#issuecomment-130493164
Received on Thursday, 13 August 2015 00:56:11 UTC