W3C home > Mailing lists > Public > public-webrtc-logs@w3.org > October 2019

Re: [webrtc-stats] Exposing RTCIceCandidateStats.networkType might trigger fingerprinting (#374)

From: youennf via GitHub <sysbot+gh@w3.org>
Date: Thu, 31 Oct 2019 21:35:11 +0000
To: public-webrtc-logs@w3.org
Message-ID: <issue_comment.created-548578523-1572557710-sysbot+gh@w3.org>
Below some further questions.

> the exact definition of what should be exposed could potentially be redefined

Would a single bit be sufficient (wireless/not wireless), or something like wired/wifi/wireless? 
We already expose private IP addresses in some cases. Would that information be sufficient for debugging in most cases? all cases? What are the cases where this is not sufficient?
Would some opaque ID that would identify the first hop/subnet be sufficient?

> The networkType stat is leaked to potentially and very indirectly improve the user experience in the future.

I understand that this was useful in the past and may be useful in the future.
So, how many times was it useful in the past? Do we have an idea of how much it is/will be useful? Can we compare the number of users benefiting from it compared to the number of users impacted by it?
If this is mostly useful for enterprise clients, should this stat be restricted to those cases only?

-- 
GitHub Notification of comment by youennf
Please view or discuss this issue at https://github.com/w3c/webrtc-stats/issues/374#issuecomment-548578523 using your GitHub account
Received on Thursday, 31 October 2019 21:35:13 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 19:22:29 UTC