- From: Neil Matatall <neilm@twitter.com>
- Date: Tue, 29 Jan 2013 14:55:01 -0800
- To: Daniel Veditz <dveditz@mozilla.com>
- Cc: Neil Matatall <neil@matatall.com>, Ian Melven <imelven@mozilla.com>, public-webappsec <public-webappsec@w3.org>, Eric Rescorla <ekr@rtfm.com>
Hello Dan, Thanks for the background, I'll try to take off my hat :) Just some questions: > If an attacker can inject a policy what data can be sent where? Is this a threat that we should keep in mind? If you can inject a policy, I would think you likely have bigger issues. And if you're over http://, you have no guarantees whatsoever. On Tue, Jan 29, 2013 at 2:28 PM, Daniel Veditz <dveditz@mozilla.com> wrote: > On 1/29/2013 9:42 AM, Neil Matatall wrote: >> >> As an implementor, I would prefer no restrictions. I feel it is my >> choice to do what I please with the data. In this case, we aggregate >> logs at a central host. In other cases, I shipped data to "cloud" >> services like Splunk/loggly. I imagine this use case will be more common >> than directly logging back to the host application. > > > As browser implementors we don't only worry about people using the feature > for its intended purpose, we have to worry about intentional mis-use of a > feature. If an attacker can inject a policy what data can be sent where? If > the reporting URL is restricted to the same domain that set the policy and > provided the content then I'm comfortable that the data is really being sent > to you, the site implementor; in that case yes, please, do what you want > with the data. If the data is being sent to some other site then I don't > know it's you (might be, might not be) and then my instinct is to restrict > the amount of data available so as not to further additional attacks. The > downside of that is that you might not get enough data to do anything > useful. > > That's the trade-off. Are you getting enough data from the current error > reports? If not can we safely send the additional data just anywhere > potentially to anyone or will we have to restrict the destinations. Is it > worth the spec complications of saying reports may differ depending on the > destination? > > -Dan Veditz
Received on Tuesday, 29 January 2013 22:55:29 UTC