Account Options

  1. Sign in
The old Google Groups will be going away soon.
Switch to the new Google Groups.
Google Groups Home
« Groups Home
mozilla.dev.security
+ new post
About this group
Subscribe to this group
This is a Usenet group - learn more
Message from discussion Content Security Policy Spec questions and feedback
The group you are posting to is a Usenet group. Messages posted to this group will make your email address visible to anyone on the Internet.
Your reply message has not been sent.
Your post was successful
 
From:
To:
Cc:
Followup To:
Add Cc | Add Followup-to | Edit Subject
Subject:
Validation:
For verification purposes please type the characters you see in the picture below or the numbers you hear by clicking the accessibility icon. Listen and type the numbers you hear
 
Lucas Adamski  
View profile  
 More options Jul 10 2009, 3:18 pm
Newsgroups: mozilla.dev.security
From: Lucas Adamski <lu...@mozilla.com>
Date: Fri, 10 Jul 2009 12:18:12 -0700
Local: Fri, Jul 10 2009 3:18 pm
Subject: Re: Content Security Policy Spec questions and feedback
Print | View thread | Show original | Report this message | Find messages by this author
With security, its safer (and more accurate) to assume compatibility  
breakage than not.  Its not just syntax that can change but the rules  
themselves.  For example if we identify new vectors for code  
injection, we might have to block additional APIs thus breaking sites  
that would otherwise effectively support CSP without any change in  
syntax.

Even something as relatively simple to reason about like HTTP itself  
is versioned, that is why each transaction starts with the HTTP  
version for the request.  CSP is a complex security model; I would say  
backwards breakage in the future is inevitable regardless of how much  
we churn on it now, given it will have to evolve hand in hand with our  
understanding of the threat models.  It cannot anticipate attack  
vectors we don't yet know about.
   Lucas.

On Jul 8, 2009, at 9:21 AM, Gervase Markham wrote:

> So the versioning in the UA is to guard against a policy syntax  
> change. But the syntax is so simple a list of (key/value pairs) that  
> it's very, very hard to imagine a requirement which would mean we  
> *had* to break the syntax. And yet, every request the browser ever  
> sends acquires another six bytes, until the end of time. (This is  
> not a UA token which changes over time as browsers change, like OS,  
> it's one which has to be present for ever.)

> I don't think the risk of needing a breaking syntax change is worth  
> it. In that very unlikely event, we should instead plan to deploy a  
> new header, as you say below. There's more downstream bandwidth than  
> upstream, and there's more every year.

>> The other approach is to version the response, a few extra bytes only
>> when a server supports CSP. Yay, bandwidth win! But then what do we  
>> do?
>> How does the server know which version to send? Should it send every
>> version it knows about, and the client process the highest version it
>> knows how to process? That means if we ever have a CSP-2 either  
>> clients
>> are sending two complete headers (or three, or more) or they're  
>> sending
>> their preferred version and users of clients which only support CSP-1
>> get zero protection rather than the 99% they actually support.

> But this scary scenario fails to take into account the frankly tiny  
> chance that we'll need to make one breaking syntax change, let alone  
> two. Even with the spec as it is. Careful design can reduce the  
> chances even further.


 
You must Sign in before you can post messages.
To post a message you must first join this group.
Please update your nickname on the subscription settings page before posting.
You do not have the permission required to post.
Create a group - Google Groups - Google Home - Terms of Service - Privacy Policy
©2012 Google