Re: [csswg-drafts] [cssom-view] No event to track window position (#7693)

The CSS Working Group just discussed `[cssom-view] No event to track window position`, and agreed to the following:

* `RESOLVED: PURSUE EVENT TO MAKE REPORTS MORE ERGONOMIC TO USE`

<details><summary>The full IRC log of that discussion</summary>
&lt;frances_> florian: window attribute has objects and event when size changes but not when position changes, proposal to add move event that works similar to resize event<br>
&lt;frances_> florian: nuance while position attribute is on object, browsers are not required to report or have a global coordinate system and hide them, not reliably available<br>
&lt;frances_> florian: if not possible to land html patch, might differ things, could possibly land them in css patch<br>
&lt;khush> q+<br>
&lt;frances_> Alan: Alice's prs are closed why?<br>
&lt;frances_> florian: suspected glitch of rebasing, sat for a while and appears closed<br>
&lt;astearns> ack khush<br>
&lt;frances_> khush: Do all os's have this event? Are we expecting all browsers to behave similarly?<br>
&lt;frances_> Florian: fickle to speak of all os's, not finite, up to implementation detail and html loop in rendering, not need to pull more frequently<br>
&lt;frances_> florian: theres a limit on how frequently<br>
&lt;frances_> khush: if window position is changing, how can we minimize wakeups on the screen and not rely on it in case os is not giving the event<br>
&lt;frances_> Alan: possibly resize event has timing specified<br>
&lt;frances_> florian: specified similarly<br>
&lt;frances_> khush: doubt to have the problem on the os when window is resized<br>
&lt;frances_> florian: as mentioned, linux on Wayland does not have os tell changes on coordinates, windows are composited, no requirement to report numbers<br>
&lt;frances_> florian: up to implementation discretion, numbers are already exposed<br>
&lt;frances_> Alan: Igalia is possibly looking for an okay and no objections to adding the event<br>
&lt;frances_> florian: possible limited appetite to the event, make it convenient to not waste energy in actively pulling<br>
&lt;frances_> Alan: any other comments or thoughts?<br>
&lt;frances_> Daniel: is it set in such a way to detect whether you will ever get a callback or expect it?<br>
&lt;frances_> florian: there is nothing on the event itself, possible window attributes maybe undefined<br>
&lt;frances_> florian: spec should give an answer<br>
&lt;frances_> Daniel: that is my only concerned, sort of imperfect, might not return anything, need to make it ergonomic with the known limitation<br>
&lt;frances_> Alan: put something on the spec such as exposing window position and does it get updated, put event on the window object or it shouldn't be there to add the existence of the event<br>
&lt;frances_> florian: trying to find part of the spec to find where the information is, spec says you don't have to report. Does anyone remember?<br>
&lt;frances_> Alan: these are details that we can work out<br>
&lt;frances_> Daniel: Wayland limitation is possibly substantial, might never work it is an important constraint to be aware of<br>
&lt;frances_> florian: important on window object itself<br>
&lt;frances_> Daniel: we already expose possibly bad information, need to be more ergonomic<br>
&lt;frances_> florian: trying to find the part of spec, possibly open separate issue to track and detect<br>
&lt;frances_> florian: is it relevant to know difference between  0 0 corner? open issue and track<br>
&lt;frances_> Alan: should we land a spec for the event and then ask for a privacy review or vice versa?<br>
&lt;frances_> florian: event doesn't make information worse, need to add way to report information with privacy considerations<br>
&lt;frances_> Alan: information available but hard to use, might make it detrimented<br>
&lt;frances_> florian: might make it inefficient for the programmer<br>
&lt;frances_> Ian: trivial, not difficult to access<br>
&lt;frances_> Daniel: agreed<br>
&lt;vmpstr> this says 0 if not available, not sure if that's the right spot https://drafts.csswg.org/cssom-view/#dom-window-screenleft<br>
&lt;frances_> Alan: let's finish up discussion<br>
&lt;frances_> RESOLVED: PURSUE EVENT TO MAKE REPORTS MORE ERGONOMIC TO USE<br>
&lt;frances_> florian: should pull request wait?<br>
&lt;chrishtr> Agree to land a PR<br>
&lt;frances_> Alan: pull request then work on detials<br>
&lt;frances_> florian: move on css or pursue html request from get go?<br>
&lt;frances_> Alan: do both in both specs directly<br>
&lt;frances_> florian: agreed<br>
&lt;frances_> Alan: no objections<br>
&lt;frances_> Alan: next issue. Zoom issues for Chris.<br>
&lt;frances_> Alan: issue 9644 on iframes<br>
</details>


-- 
GitHub Notification of comment by css-meeting-bot
Please view or discuss this issue at https://github.com/w3c/csswg-drafts/issues/7693#issuecomment-1876144667 using your GitHub account


-- 
Sent via github-notify-ml as configured in https://github.com/w3c/github-notify-ml-config

Received on Thursday, 4 January 2024 00:20:21 UTC