W3C home > Mailing lists > Public > public-webappsec@w3.org > June 2014

Re: Reducing reporting noise

From: Devdatta Akhawe <dev.akhawe@gmail.com>
Date: Fri, 20 Jun 2014 16:16:02 -0500
Message-ID: <CAPfop_0Bt35fFN+hy=snCgftDBib+J4gJnCgvW2+dm=OBqp77g@mail.gmail.com>
To: Daniel Veditz <dveditz@mozilla.com>
Cc: Neil Matatall <neilm@twitter.com>, Glenn Adams <glenn@skynav.com>, "public-webappsec@w3.org" <public-webappsec@w3.org>

I am not sure we need a new "dont-report" or "freq-xxx" directives. I
agree with Neil that it is easy for the server to filter these out. If
this "server-side" filtering is too painful, relying on the JavaScript
interface (say an event for each violation) makes more sense to me. A
declarative mechanism like "dont-report" or "freq-xxx", imho, won't be
flexible enough.

It seems to me that such a JS mechanism can be achieved via a
CSP-report-only header that has no report-uri. The
SecurityPolicyViolationEvent listener can then filter and aggregate as
needed and send the data to the server (modulo being pwned by XSS).

I agree that, per priority of constituencies, capping reports to
conserve bandwidth makes sense. But, I think this should just be left
to the browser with the spec only saying some sort of MAY wording
about capping reports.


On 20 June 2014 13:56, Daniel Veditz <dveditz@mozilla.com> wrote:
> On 6/19/2014 7:00 PM, Neil Matatall wrote:
>> I feel it's the job of the reporting endpoint to make the decision to
>> drop a report on the floor. I realize this is not consistent with the
>> goal of reducing the number of reports sent, but hey.
> It is considerate of the user's bandwidth to avoid sending reports that
> the content author knows are just going to get dropped anyway.
> -Dan Veditz
Received on Friday, 20 June 2014 21:16:49 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 18:54:39 UTC