Re: [w3c/editing] Change default paragraph separator from <p> to <div> (#146)

> It also affects what happens when the user presses "Enter" (which is equivalent to calling execCommand("insertparagraph")). Do any editors use this, or do they all override it?

CKEditor overrides it totally. I think that some editors may try to post-fix this (so base on the native "insertparagraph" algorithm, but change the created block afterwards), but those should also be safe.

> Do editors all override hitting Return, Backspace, etc.?

CKEditor 4 overrides entire enter key behaviour and backspace in certain cases (mainly in lists, tables and in Chrome when it comes to [merging blocks](http://dev.ckeditor.com/ticket/9998)). CKEditor 5 overrides everything.

So: 

> Not sure what @Reinmar think, but I would say for those just try to be as similar as Chrome given that that is the currently most popular browser on those

This wouldn't be fully true for us – as I mentioned above, Chrome has certain behaviours which we cannot accept. But it's hard for me to tell which browser's behaviour would suit us the most, cause we override too many so I don't even know how certain browsers behave :).

> The top priority right now would really seem to be to get the beforeinput event implemented in Firefox in the same way that Chrome is currently implementing it. Or what would you say, @Reinmar?

I agree. Beforeinput would be the biggest improvement for us. In CKEditor 5 we intercept all input, because it must be carefully applied in the custom data model (which is based on operational transformation). Right now we intercept the input based on available events (e.g. `paste`), but most of the typing can only be intercepted by mutation observers. This is pretty tricky, though, if you consider composition, autocorrection, spell checkers, etc. and `beforeinput` would a huge relief :).

Actually, once we're done with the current sprint, we hope to create a PoC of https://github.com/ckeditor/ckeditor5-typing working based on `beforeinput` available in Chrome.

-- 
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/w3c/editing/pull/146#issuecomment-244114607