Re: [whatwg] window.opener and security

2008-07-29 Thread Ian Hickson
On Tue, 20 Mar 2007, Hallvord R M Steen wrote: > > when a new window or tab is opened by a page it normally has a > "window.opener" property that points to the window object of the > original tab. Indeed, this is now specced. > If an origin check fails when comparing the locations of the old w

Re: [whatwg] window.opener and security

2007-03-20 Thread liorean
On 20/03/07, Thomas Broyer <[EMAIL PROTECTED]> wrote: 2007/3/20, liorean: > Some thing I would like to add here, is that your "solution" doesn't > do anything to solve the actual l problem case. Even if window.opener > would be read only, that is just a reference to a window object. Even > if tha

Re: [whatwg] window.opener and security

2007-03-20 Thread Gareth Hay
If the primary domain is www.example.com and the other domain is help.example.com the UA clearly should allow them to communicate by request. Believe me, nulling window.opener if origin check fails will break MANY sites. This is not the point I am making, and I feel we are not understanding on

Re: [whatwg] window.opener and security

2007-03-20 Thread Thomas Broyer
2007/3/20, liorean: Some thing I would like to add here, is that your "solution" doesn't do anything to solve the actual l problem case. Even if window.opener would be read only, that is just a reference to a window object. Even if that property would be read only you could still write to the lo

Re: [whatwg] window.opener and security

2007-03-20 Thread Hallvord R M Steen
>> 1) Either it is your responsibility to handle the nulling of the >> property *or* >> 2) It is the UA's. > > The UA can not do this. It would break legacy pages by resetting > window.opener if content comes from a different server. If this is a security point, which I take from the subject "w

Re: [whatwg] window.opener and security

2007-03-20 Thread Gareth Hay
I was clearly mislead by the "window.opener and security" title then On 20 Mar 2007, at 15:51, liorean wrote: On 20/03/07, Gareth Hay <[EMAIL PROTECTED]> wrote: As was clearly stated, I showed a workaround and then suggested it should be up to the UA to handle this situation. It is not helpful

Re: [whatwg] window.opener and security

2007-03-20 Thread Gareth Hay
On 20 Mar 2007, at 15:45, Hallvord R M Steen wrote: 1) Either it is your responsibility to handle the nulling of the property *or* 2) It is the UA's. The UA can not do this. It would break legacy pages by resetting window.opener if content comes from a different server. If this is a security

Re: [whatwg] window.opener and security

2007-03-20 Thread liorean
On 20/03/07, Gareth Hay <[EMAIL PROTECTED]> wrote: As was clearly stated, I showed a workaround and then suggested it should be up to the UA to handle this situation. It is not helpful to deliberately misunderstand points, and quote them out of context. I suggest you re-read my mail. You showed

Re: [whatwg] window.opener and security

2007-03-20 Thread Hallvord R M Steen
1) Either it is your responsibility to handle the nulling of the property *or* 2) It is the UA's. The UA can not do this. It would break legacy pages by resetting window.opener if content comes from a different server. I personally think the UA should handle it (as stated previously) **BUT** i

Re: [whatwg] window.opener and security

2007-03-20 Thread Hallvord R M Steen
> > > http://my.opera.com/hallvors/blog/2007/03/14/window-opener-and-security-an-unfixable-problem > > javascript: void(window.open( 'http://hallvord.com/temp/redir.php')) I don't know what GMail is doing, but I think a window.open('','_self') would destroy the original window.opener. That's a

Re: [whatwg] window.opener and security

2007-03-20 Thread Gareth Hay
It would appear that at least the WebKit team agree about the window.opener being read-only. It has resisted all attempts by me to null it or re-assign it, and as soon as the domains no longer match exceptions are thrown. From a security point of view I think this is sufficient to prevent

Re: [whatwg] window.opener and security

2007-03-20 Thread Gareth Hay
I think you are deliberately missing the point now... On 20 Mar 2007, at 14:50, Hallvord R M Steen wrote: On 20/03/07, Gareth Hay <[EMAIL PROTECTED]> wrote: Anyway, for use case 1 - If you are worried about phishing attacks, you should be using some sort of onunload handler trapping to null wi

Re: [whatwg] window.opener and security

2007-03-20 Thread Martijn
2007/3/20, Hallvord R M Steen <[EMAIL PROTECTED]>: On 20/03/07, timeless <[EMAIL PROTECTED]> wrote: > On 3/20/07, Hallvord R M Steen <[EMAIL PROTECTED]> wrote: > > http://my.opera.com/hallvors/blog/2007/03/14/window-opener-and-security-an-unfixable-problem > I believe you'll find that Gmail doe

Re: [whatwg] window.opener and security

2007-03-20 Thread Hallvord R M Steen
On 20/03/07, Gareth Hay <[EMAIL PROTECTED]> wrote: Anyway, for use case 1 - If you are worried about phishing attacks, you should be using some sort of onunload handler trapping to null window.opener. Yet you are arguing that it should be impossible to set window.opener. If you had your way tha

Re: [whatwg] window.opener and security

2007-03-20 Thread Hallvord R M Steen
On 20/03/07, timeless <[EMAIL PROTECTED]> wrote: On 3/20/07, Hallvord R M Steen <[EMAIL PROTECTED]> wrote: > http://my.opera.com/hallvors/blog/2007/03/14/window-opener-and-security-an-unfixable-problem I believe you'll find that Gmail does not have this problem, because when it uses window.op

Re: [whatwg] window.opener and security

2007-03-20 Thread Gareth Hay
Well, I don't think it is off-topic. You are trying to justify writing to a property I think should be read-only. I am asking you why you think this should be possible. Anyway, for use case 1 - If you are worried about phishing attacks, you should be using some sort of onunload handler trapp

Re: [whatwg] window.opener and security

2007-03-20 Thread timeless
On 3/20/07, Hallvord R M Steen <[EMAIL PROTECTED]> wrote: http://my.opera.com/hallvors/blog/2007/03/14/window-opener-and-security-an-unfixable-problem I believe you'll find that Gmail does not have this problem, because when it uses window.open, it opens a gmail page which then triggers a serve

Re: [whatwg] window.opener and security

2007-03-20 Thread Hallvord R M Steen
On 20/03/07, Gareth Hay <[EMAIL PROTECTED]> wrote: Well, window.opener is conceptually a link from child to parent. Can you give a valid use-case for adoption of the child to another parent? Again: We are off-topic. This isn't what I'm trying to discuss in this thread. However, here are two us

Re: [whatwg] window.opener and security

2007-03-20 Thread Gareth Hay
Well, window.opener is conceptually a link from child to parent. Can you give a valid use-case for adoption of the child to another parent? On 20 Mar 2007, at 13:00, Hallvord R M Steen wrote: On 20/03/07, Gareth Hay <[EMAIL PROTECTED]> wrote: window.opener should be read-only and attempting

Re: [whatwg] window.opener and security

2007-03-20 Thread Hallvord R M Steen
On 20/03/07, Gareth Hay <[EMAIL PROTECTED]> wrote: window.opener should be read-only and attempting to write to it should throw an exception. I don't really see why setting opener would be dangerous, so I disagree that it should throw. Anyway, that is a different issue. What I'm talking about i

Re: [whatwg] window.opener and security

2007-03-20 Thread Rimantas Liubertas
2007/3/20, Gareth Hay <[EMAIL PROTECTED]>: window.opener should be read-only and attempting to write to it should throw an exception. It was possible to set window.opener in IE, alas, I do not remember which version :( But it has been fixed, AFAIK. Regards, Rimantas -- http://rimantas.com/

Re: [whatwg] window.opener and security

2007-03-20 Thread Gareth Hay
window.opener should be read-only and attempting to write to it should throw an exception. This is a similar issue to window.history, in certain browsers you can write to this with js. It has no effect, but does persist across domains. The webkit team decided to just throw an exception if a

[whatwg] window.opener and security

2007-03-20 Thread Hallvord R M Steen
Hi, when a new window or tab is opened by a page it normally has a "window.opener" property that points to the window object of the original tab. This happens whether the new window is opened by a JavaScript calling window.open or by a link or form with target attribute set. If an origin check f