I think the idea is a good one in general, though it would be good to get data from Google on how it's been used in practice.
It's been quite a while since I reviewed the spec, so my thoughts below might be outdated (or I might be misremembering). There were two things that I found unfortunate about the spec. One was that it was a bit undefined exactly which headers to include when. This should not be left up to UA policy, but should be clearly defined in spec. The spec seemed too focused on including CH headers when loading subresources of a page. Too few headers were included when loading the "document" (in practice, when loading the HTML). My understanding is that UA-sniffing is quite often used to determine what HTML to serve to a given client, which I was hoping that CH could help reduce. I know there's some concern that always sending CH with all requests would generate too much header traffic which would be wasteful. However I think something like the following could work: * Include most headers during a navigation when loading a "document". (In gecko terms, when the nsIContentPolicy-type is TYPE_DOCUMENT or TYPE_SUBDOCUMENT). * Default to no CH headers being sent for other requests. * Make it possible for the document response to opt in to which headers should be sent for resource loads that happen inside that document. (In gecko terms, for requests which have a loadingNode that belongs to the given document) / Jonas On Tue, Mar 8, 2016 at 5:05 AM, Xidorn Quan <quanxunz...@gmail.com> wrote: > On Tue, Mar 8, 2016 at 7:40 PM, Gervase Markham <g...@mozilla.org> wrote: > >> On 08/03/16 06:22, Andrew Overholt wrote: >> > Implement Client-Hints HTTP header >> > https://bugzilla.mozilla.org/show_bug.cgi?id=935216 >> >> Well, we are in favour of adaptive content, progressive enhancement, >> responsive images in HTML, and feature detection. The question is >> whether we think that these things cover all the use cases or not. >> >> Do we know whether anyone's using this in Chrome? >> > > Chrome Platform Status has items for this feature: > ClientHintsContentDPR: > https://www.chromestatus.com/metrics/feature/timeline/popularity/906 > ClientHintsDPR: > https://www.chromestatus.com/metrics/feature/timeline/popularity/835 > ClientHintsMetaAcceptCH: > https://www.chromestatus.com/metrics/feature/timeline/popularity/904 > ClientHintsResourceWidth: > https://www.chromestatus.com/metrics/feature/timeline/popularity/836 > ClientHintsViewportWidth: > https://www.chromestatus.com/metrics/feature/timeline/popularity/837 > > So it seems there were some websites tried to use this, which made its > usage up to 0.06%, but then they abandoned. All of the features above is > used only in ~0.0002% of pages surveyed now, which may indicate its > unpopularity. Not sure why it was ever used in 0.06% of pages, and why > those page abandoned it. Probably Chrome team knows. > > - Xidorn > _______________________________________________ > dev-platform mailing list > dev-platform@lists.mozilla.org > https://lists.mozilla.org/listinfo/dev-platform _______________________________________________ dev-platform mailing list dev-platform@lists.mozilla.org https://lists.mozilla.org/listinfo/dev-platform
