Please see comments.

Added comments.

This patch is managed in devs/hermet/lottie

This patch is managed in devs/hermet/lottie branch.

remove unused parameter warning.

In D7494#131094, @cedric wrote:

I think this is generally a good idea. Reducing event being triggered in the hot path is a good thing. There is a few things that I don't get in this patch, so please have a look at that.

Also what happen if you render an object where no commit call has been made ? I would expect that until you call commit, you render the previously committed path.

An alternative solution would be to have a PATH_CHANGED triggered only once, when rendering is requested and that the path did change. Guaranteeing at max one update event per frame. Not sure really which one is best at the end. But I definitively agree with the goal of this patch and will to any of the above solution being implemented.

update code according to review comments.

In D6919#122570, @JackDanielZ wrote:

Hi Hermet,

How can you abandon it? It is already in master. Did you revert it?

Daniel

In D7494#131078, @segfaultxavi wrote:

Since this is an optimization, I would expect a test case or example which can be used to show how this patch reduces the CPU usage. Some figures obtained with that test case could also be added to the patch.

eh? Could you please explain how to run enlightenment with xdebug.sh more detail?

added api doc.

updated some code.

updated code.

In D7456#130257, @cedric wrote:

Wouldn't it make sense to go with an Eina_Inarray structure then ? Would avoid the addition of all the expansion logic.

revert unnecessary changes.

1. Can't figure out issue in your sample.
2. If you say, map coordinates weren't updated properly before clip_recalc_do(), we can just update it's coordinates (without checking inactive statement) before that. But it must be updated when we call evas_object_map_set() or evas_object_map_eanble_set()

updated commit message.

updated commit message.

comment+

list view sounds nicer than view list.

Curious, "Is on Plane" would be a generic method or just only for image type....

Strange... this native surface has been used long time in Tizen without any leak issues. It would be better if introduce use-case to assume this issue scenario...