search

w3c / server-timing

Server Timing
http://w3c.github.io/server-timing/
Other
75 stars 20 forks source link

Editorial cleanup #16

Closed igrigorik closed 7 years ago

igrigorik commented 7 years ago

Outstanding TODO's:

Preview | Diff

igrigorik commented 7 years ago

@marcoscaceres can we bug you for a sanity check + some help on respec IDL front here? :)

marcoscaceres commented 7 years ago

Happy to help.

cvazac commented 7 years ago

Regarding hooks from resource and navigation timing, best I can think of is a note after the "Create" step of the "Processing Model" that says something like: "After the object has been created, and before it has been queued/added, allow other specs to add non-overlapping attributes." I couldn't think of a way (excepting pub/sub) to set up something that other specs could "call in to".

igrigorik commented 7 years ago

@marcoscaceres thanks! In particular, any guidance on https://github.com/w3c/server-timing/pull/16/files#diff-eacf331f0ffc35d4b482f1d15a887d3bR134 would be appreciated. If you have time to do an overall check, that would be awesome too. :-)

@cvazac thanks, updated the ID and removed the URL. Re, hook: we can do a similar approach as we have in HTML + Page Visibility specs:

(HTML, step 6): Run any unloading document visibility change steps for Document that are defined by other applicable specifications.

(PV) If the user agent is to unload the Document, run the now hidden algorithm during the unloading document visibility change steps.

It may also make sense to define two hooks: one when entry is created, and one right before entry is queued, to allow for upstream specs to easily hook in and extend at both points.

siusin commented 7 years ago

@igrigorik What is L30 format: "markdown" for here? Can it be removed? - it will create chaos for the HTML markup.

Let's publish a new WD for the WebPerf WG once this PR gets merged.

igrigorik commented 7 years ago

@siusin same as https://github.com/w3c/preload/pull/104 -- this hasn't been a problem before?

marcoscaceres commented 7 years ago

Sorry for the delay... starting review now.

igrigorik commented 7 years ago

👍 thanks Marcos!

The last remaining bit here is to define a hook in RT and then update ST to use it. We should probably break that into a separate PR however.

@cvazac if the PR looks good, feel free to merge. I'll open a separate issue to track RT hook: https://github.com/w3c/server-timing/issues/19

marcoscaceres commented 7 years ago

I have a PR ready to go in my own fork to apply on top of this one to fix a few more xref nits.

cvazac commented 7 years ago

Should/can we fix the travis failure before merging @igrigorik?

igrigorik commented 7 years ago

@cvazac hm, not sure why that's not happy.. /cc @siusin. FWIW, I don't think we need to block on that. Merging, and @marcoscaceres can they apply additional fixes.