- From: Jan-Ivar Bruaroey via GitHub <sysbot+gh@w3.org>
- Date: Mon, 12 Apr 2021 23:03:13 +0000
- To: public-webrtc-logs@w3.org
Sorry for the late reply. TL;DR: This seems like an imperfect solution to three problems, two of which I'd rather not fix. > I was not there when label was initially designed but, for capture tracks, the localized device name is used. This makes it potentially easy for a web page to display the source of the track. I think it was the initial intent. Yes, the sole intent behind label was for local display, to help end-users remember what tracks they've chosen.ยน `label` is neither localizeable nor unique across localizations, in the sense that it only matters that the end-user recognizes it as representing the source they originally chose (whether in an in-content or in-browser picker). Having applications override them runs counter to that intent. > Analytics ... Capturing applications can gather statistics over what applications its users tend to capture. Applications overriding labels cannot solve this, except for a subset of sources (doesn't solve native sources). Labels were not designed for analytics, but to the extent analytics work for camera and microphone today (by accounting for their potential variance across locales), I'd suggest https://github.com/w3c/mediacapture-screen-share/issues/128 or opening a separate issue if analytics is a pressing issue. > Cross-Tab Communication With several preexisting mechanisms through which tabs can communicate, I'm skeptical of new ones. The only thing unique about this one, seems to be its ability to help force the outcome of *getDisplayMedia* (as in: "please pick again"), which I think is a poor solution to https://github.com/w3c/mediacapture-screen-share/issues/143#issuecomment-818115676, and better solved by *getViewportMedia*. > Conditional Tab-Focus Change I consider the whole area of interaction between tabs around *getDisplayMedia* an anti-pattern, due to the unique [security concerns](https://w3c.github.io/mediacapture-screen-share/#capabilities-depending-on-elevated-permissions) around capture of web surfaces. User agents are [supposed to](https://w3c.github.io/mediacapture-screen-share/#dom-mediadevices-getdisplaymedia) _"discourage their selection on the basis that these represent a significantly higher risk when shared"_. We should be scaling back (not ramping up) such usage of *getDisplayMedia*, and work to standardize *getViewportMedia* for these apps. --- <sub>1. To illustrate the narrowness of scope: we even considered inventing special opaque-to-JS text-elements that could be displayed locally only, but not inspected by JS *(dropped for font issues, other complexities, and being quite overboard of an idea)*</sub> -- GitHub Notification of comment by jan-ivar Please view or discuss this issue at https://github.com/w3c/mediacapture-screen-share/issues/159#issuecomment-818298411 using your GitHub account -- Sent via github-notify-ml as configured in https://github.com/w3c/github-notify-ml-config
Received on Monday, 12 April 2021 23:03:16 UTC