Re^3: [OT] Testing race conditions in web pages

[Jenda]

Even if the CSS is indeed dynamically generated and user dependent, the race condition still makes no sense. Unless of course he's crazy enough to keep the URL of the CSS static and generate CSS for the last person that requested the page or something similarly insane.

[Corion]

Well, not to claim that this is sane, but you could serve different CSS to different clients based on the cookie they request the URL with...

[markong]

Given the poor information the OP has shared, I think that the only way to interpret this "race condition" is that the friend is serving resource's URL(s) which maintain the whole request state consistent if they are hit in some defined consecutive and predictable order, which given the distributed nature of the HTTP protocol he should not assume.