On Mon, Jan 3, 2011 at 11:36 PM, Jari Bakken <[email protected]> wrote:
> Den 3. jan. 2011 kl. 23:55 skrev Simon Stewart <[email protected]>:
>
>>
>> The constraint is laid down because we sometimes need to implement the
>> cookie code in pure JS.
>> At that point, the only cookies available are
>> those that are visible in the current domain. The WebDriver API
>> mirrors this constraint. It does mean that in those browsers where we
>> have better integration with the browser the API is more constrained
>> than it needs to be, but it also means that we have a consistent
>> contract between browsers.
>
> For future reference, where is that?

Strictly speaking: that we only return cookies that would be visible
using Javascript from the current frame that is selected. The tests
are located at:

http://code.google.com/p/selenium/source/browse/trunk/common/test/java/org/openqa/selenium/CookieImplementationTest.java

>> Also, no-one's really got a good way of acknowledging HTTP only
>> cookies.
>
> I'm not sure what that means. What are HTTP only cookies?

Terrifying. They're cookies that are _not_ viewable from client-side
scripts. I've not seen them much in the wild yet (no pun intended)

http://tools.ietf.org/html/draft-ietf-httpstate-cookie-09
http://www.owasp.org/index.php/HttpOnly

Simon
_______________________________________________
Wtr-development mailing list
[email protected]
http://rubyforge.org/mailman/listinfo/wtr-development

Reply via email to