[webkit-reviews] review denied: [Bug 78445] Web Inspector: include size, failed, loading and requestId in FrameResourceTree : [Attachment 126698] Proposed Change (with build fix)

[bugzilla-daemon at webkit.org]

Pavel Feldman <pfeldman at chromium.org> has denied Timothy Hatcher
<timothy at apple.com>'s request for review:
Bug 78445: Web Inspector: include size, failed, loading and requestId in
FrameResourceTree
https://bugs.webkit.org/show_bug.cgi?id=78445

Attachment 126698: Proposed Change (with build fix)
https://bugs.webkit.org/attachment.cgi?id=126698&action=review

------- Additional Comments from Pavel Feldman <pfeldman at chromium.org>
View in context: https://bugs.webkit.org/attachment.cgi?id=126698&action=review


At this moment, we only add methods to the protocol that are used by the
front-end. This way we can guarantee it is useful and we can test it. Are you
planning on using this data in the front-end or is there alternative front-end
it is for? I am fine with changing the rules to the "we can expose
functionality not used by the inspector front-end", but we need to make sure
these methods are tested.

> Source/WebCore/inspector/Inspector.json:141
> +				   { "name": "decodedSize", "type": "integer",
"optional": true, "description": "Decoded resource size." },

These are going to be wrong for the ports that have network stack responsible
for encoding.

> Source/WebCore/inspector/Inspector.json:143
> +				   { "name": "loading", "type": "boolean",
"optional": true, "description": "True if the resource is still loading." },

I am not sure how relevant this is given the asynchronous nature of the
front-end - backend interaction. There is no guarantee that it is still "true"
at the moment message is received.