If you're someone whose username I'd recognize, you can send me a resume (my e-mail is in my HN profile) and I can refer you. That sends you to the top of the pile that HR looks at (technically, I think it's two separate queues, and they empty the referral queue before they look at the slushpile), though you still have to get through the interview process. I got in through a referral - I'm guessing it helps quite a bit. And there's a nifty bonus for me if you're accepted. :-)
Does Google even hire people who care any more? Does it only hire people who interview really well? From what I've gathered, most people who get through the HR disaster there these days (old times are exempt; decentralizing hiring is the first sign of impending disaster) really aren't that smart; maybe they've merely perfected a sort of slimy smooth-talking (literally speaking), typical of most salesmen.
As much as I want to still love Google, it's getting harder every day. Even my original suggestion regarding pixel space amounts on the search results page template seems to have been implemented recently:
Yeah, I'm a wee bit irked; I think I deserved a bit more than the lie-wielding form rejection email from my application somewhere back in early 2005 or so. And don't even get me started on the fact that Google went and doubled the limit of receiving pay on my AdSense account sometime last year; so I've still not even seen anything from that.
So as much as I still want to love Google, it is getting harder every day. Your comment regarding "Nooglers" is actually pretty funny considering how many of them, like me, have probably been really diligent in their quest to get hired just about anywhere.
I know Nostrademons in person, who was recently hired by Google, and I can vouch that he does care, is extremely smart, and isn't a slimy salesman. Though it's getting larger, and hence more bureaucratic (from what I've heard on the outside), Google still has the best data in the world and hence attracts world-class engineers.
If you believe my recruiter, it was previous startup experience, strong CS fundamentals, and good JavaScript knowledge. In other words, I had a skill set that they needed and seemed like a person who could get things done. Pretty standard stuff.
The interview process was less difficult than I thought it would be. 1 phone screen, 4 (supposed to be 5, but one never showed up) in-person interviews over a day. I'd thought I'd flubbed it when I got out, but apparently the interview feedback was quite positive.
If I had to guess what makes most candidates bomb, it's that they don't have sufficient depth in the positions they're interviewing for. For example, at the end of one of my interview problems, the interviewer asked me "And how would you attach that event handler?" And I replied "Well, if it was IE, I'd use element.addEventListener('mouseover', fn, true). If it was a standards-compliant browser, I'd use element.attachEvent('mouseover', fn)." Note that that's wrong - it's backwards. But the interviewer didn't care. He just wanted to make sure that I was aware of those functions, because so many of the people he interviews would just say element.onmouseover = fn.
Similarly, most of the disastrous product-manager interviews I've heard of have been because the candidate would give handwavey answers and couldn't clarify what he meant or figure out how to support his opinions with data, even when prompted by the interviewer. Google culture really prizes precision, perfectionism, and strong opinions backed up with data - if you're used to finding the first Google hit on the web, implementing something by copying and modifying that, and then saying "Okay, it's done," you really won't fit in here.
I'm kinda curious why you would ever use onmouseover when attachEvent/addEventListener are available? I just took out the remaining calls to .onEvent handlers in websearch's JS library - the only browsers that don't support the W3C/IE event models are so ancient that they don't get JS anyway, and there's not much point using a method that's unnecessary on modern browsers and does the wrong thing anyway.
So yeah, my interviewer wasn't implying that onmouseover is used only by bad programmers, but I am. ;-)
Maybe I only need one event handler, or maybe I've baked a function that calls another function?
Besides, I'd rather use one standard method that works across browsers than choose one based on the user's browser. Would you call that bad programming? :p
Nowadays I generally use jQuery (.click(fn)) though.
Ah, okay. Most of the stuff I've worked on adds events through a library (eg. jQuery.click(fn)), and there's no reason for that library to use onmouseover etc.
However, won't HN people applying through that method just be sent in to a black hole?