sbp commented on issue #317:
URL: 
https://github.com/apache/tooling-trusted-releases/issues/317#issuecomment-3538941996

   This is a combination of two behaviours. We set `SameSite=Strict` for the 
ASFQuart session cookie, which means that the browser does not send its 
existing session cookie when you follow a link from e.g. a vote email or from 
the page of this GitHub issue thread. On most routes that would result in an 
immediate prompt to log in, because most of our routes require the user to be a 
committer. That's not true of this vote page, however: it's a public page. This 
combination means that the user is effectively logged out in that whole 
cross-site browsing context, and is not prompted to log in on this individual 
page.
   
   What would be ideal is if we could detect the existence of a session cookie 
even if we could not read it, and then force the authentication. I doubt 
(without checking) that the browser vendors will have allowed this exact 
behaviour, but we could mimic it by setting a cookie called `has_session` and 
setting that to `SameSite=Lax` or even `SameSite=None`. We would then detect 
this and prompt the user to authenticate (or clear that cookie) on public 
routes where an enhanced interface is available for committers.
   
   Alternatively, and much more easily, we could make the page available only 
to committers, but our [voting 
policy](https://apache.org/foundation/voting.html) does not preclude 
non-committers from voting. They would need an ASF UID anyway to record their 
vote on ATR, but they would not need a UID to vote on the mailing list. 
Therefore we will probably have to implement `has_session`, but we will have to 
think carefully about the security implications.


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]


---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to