I think it checks the referral page you came from when clicking it and shows the {"is_fb_employee":"maybe soon? <a href="https://www.facebook.com/careers/department.php?dept=engineering" rel="nofollow">https://www.facebook.com/careers/department.php?dept=enginee...</a>} when coming from HN.<p>If I open a new window and copy/pasta the URL, it'll say false :)
Now, if you were to browse Facebook through a proxy that always tampered with this result in transit to say you were an employee... might some stray client-side code do anything interesting because it trusted that response?
That's cute but it's not accurate. I interviewed there <i>over two years ago</i> and I appear to be blacklisted, since whenever I look at any of their job descriptions I get this:<p>"Hey, we have reviewed your application and unfortunately don't have an opening for you."<p>I can't really square that with "Maybe soon?"
I think this is_fb_employee variable is to check if they should be running the internal testing version of Facebook. Facebook has a subdomain which, in their offices everyone is redirected to. It's something like 'preview.facebook.com'. It houses the latest testing build of Facebook. This way all the employees are testing Facebook just by being on it, and they have other people testing their new builds for short periods of time (~2 weeks). Chances are, the server checks if the user is at an IP of a Facebook office, and that's the only condition where this is true. This would make sense because if they just redirect users to Facebook.com in their offices to preview.facebook.com, then nearly anyone could do it. This would also help prevent leaking of new features as well, because employees wouldn't be able to access them outside of Facebook.<p>They mention this in the Facebook Effect (<a href="http://www.amazon.com/Facebook-Effect-Inside-Company-Connecting/dp/1439102112" rel="nofollow">http://www.amazon.com/Facebook-Effect-Inside-Company-Connect...</a>). Or, at least the part about an subdomain for testing their website in-house. Everything else was me analyzing that.
According to Facebook, I might be soon an employee.<p><pre><code> {"is_fb_employee":"maybe soon? https://www.facebook.com/careers/department.php?dept=engineering"}</code></pre>
It appears that this web service has a rather obvious defect, the Content-Type is set to "text/html; charset=utf-8" yet, the response body seems to be JSON rather than HTML. The proper Content-Type should be "application/json" with Content-Disposition to "inline". Perhaps they didn't do this since some browsers ignore the Content-Disposition with this Content-Type, and prompt to download the content regardless.<p>Even so, "text/html" is still wrong. Since the content actually isn't intended for a JSON parser, but, a human, "text/plain" would be the most conservative (and not wrong).
Ok looks like if the url is referred from Hacker News, then they are displaying, {"is_fb_employee":"maybe soon? <a href="https://www.facebook.com/careers/department.php?dept=enginee..." rel="nofollow">https://www.facebook.com/careers/department.php?dept=enginee...</a>}<p>Even if you are logged out.<p>I am not an facebook employee so if i just paste the url in address bar and enter it, it shows: is_fb_employee: false with my UID.<p>If i logged out, it shows UID: 0 with is_fb_employee: false.<p>It's a nice idea to attract ppl from different site and based on the site domain, give them relevant career page url. (May be an intern project)<p>So for HN, it's engineering career page, if the referral site is relevant to some other domain (i.e. sales/marketing) then they will give <a href="http://www.facebook.com/careers/department.php?dept=sales" rel="nofollow">http://www.facebook.com/careers/department.php?dept=sales</a>.<p>Not sure how much data facebook have of other websites for categorizing majority of the websites in to different domain, but i feel Google can do much better with the same concept.
I hope there isn't a flash app on Facebook that is using just that to decide whether to show an employee/admin interface.. Will be quite easy to spoof the result of that page if it is client-side.
Who cares? Getting excited over some random graph value used internally... okay?<p>I mean, I already assumed FB had a staging server. Why does this interest people so much?