W3C home > Mailing lists > Public > public-media-capture-logs@w3.org > September 2016

Re: [mediacapture-main] Reinstate strong language on permission ending when tracks stop, lost by editorial mistake

From: jan-ivar via GitHub <sysbot+gh@w3.org>
Date: Tue, 06 Sep 2016 17:13:25 +0000
To: public-media-capture-logs@w3.org
Message-ID: <issue_comment.created-245021280-1473182003-sysbot+gh@w3.org>
> As to the original intent: I think we changed the language to say 
that when you check whether you can open a device, you first check 
whether it is already open, and if it is, you skip the check against 
permission. This preserves the previous behavior,

@alvestrand It does not. Checking whether a device is already open 
affects *overlapping* gUM requests only (i.e. the app requests gUM 
again *before* calling stream.stop(), a case we agree on, but a 
separate issue). It does not preserve the February behavior that 
requesting gUM *after* stream.stop() MUST cause a prompt "unless there
 is a stored permission" (stored by the user).

If this edit was a mistake, it should be undone. If it wasn't, then I 
can't find the necessary group consensus that we meant to change this 
required behavior.

@jyasskin Thanks for the link, but I disagree `http` vulnerability is 
the spec's only concern here (the other being privacy). As I mention 
 the spec establishes privacy as a concern when it suggests users 
should see a "device accessible" indicator (separate from "live 
recording"), and introduces concepts to imbue such an indicator with 
the necessary power, chiefly the "stored permission".

This is clearly echoed under [Best Practices: Stored 
 where it says "the choice [to store permission] needs to be apparent 
to the user",


"When permission is not stored, permission will last only until such 
time as all MediaStreamTracks sourced from that device have been 

Combined, this guarantees users that access cannot resume once the 
"device accessible" indicator goes away, unless they've made an 
"apparent" "choice" to allow such a thing.

This is consistent with the February language, but the edit is not.

GitHub Notification of comment by jan-ivar
Please view or discuss this issue at 
 using your GitHub account
Received on Tuesday, 6 September 2016 17:13:44 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 16:27:30 UTC