Re: [dom] DocumentFragment mutations

On Thu, Mar 22, 2012 at 5:48 AM, Anne van Kesteren <annevk@opera.com> wrote:

> Hey Adam,
>
> So I looked again at these algorithms after our IRC discussion.
>
> 1. I managed to avoid some of the DocumentFragment-related duplication as
> discussed: http://dvcs.w3.org/hg/domcore/**rev/ac51d41fb636<http://dvcs.w3.org/hg/domcore/rev/ac51d41fb636>


Looks good, and thanks for the Note regarding ignoring the suppression flag.


>
> 2. I did not find transient observers actually requiring additional text
> because concept-node-insert references concept-node-remove for the nodes
> removed from DocumentFragment nodes and concept-node-remove then takes care
> of adding the transient observers.
>

Yeah, this is a nice advantage of avoiding duplication.


> Now having fully defined all this and given point 2 above it seems to me
> we can inline the "add a transient observer" algorithm. Do you agree? It
> also seems to me we can either remove the "transient observer" concept in
> favor of just having the "transient" variable or the other way around.
>

Not sure that I see much difference between these two, but it definitely
would read better to me if the thing that stayed was "transient observer".
But don't we then have to keep it in sync with
http://dvcs.w3.org/hg/domcore/raw-file/tip/Overview.html#registered-observer?
That's why I went with the two-stage definition originally.

- Adam