EricBurnett

Indeed. And even at Google scale where 'good enough' is an art form, ECC memory is deemed to be worth the money. The paper you cited does its study on Google hardware, and I can confirm that it's still used today.

It worked for me earlier in the day when I tried it at work, but it doesn't appear to work any longer.

As a Googler I found this pretty funny myself. Most people won't even understand it without explanation, and those that do are likely the right sort to also find the decade old event it's a play off. I doubt too many people are going to take the wrong message away from this.

It's nice to see people are still allowed to have fun with their products.

I see about 2.8 million results from that query, but I'm not surprised the number changes a lot. The result counts are a very rough approximation calculated on the fly, and shouldn't be relied on for more than order-of-magnitude estimates.

The last name method is precise, but of questionable accuracy; the result count is probably more accurate, but very imprecise. Take your pick :).

Can I get a translation please? I read the article, but all I really got out of it was that there is a different way to configure traffic shaping now. No real clue how it's different, better, or why I should be interested.

Am I correct in saying that this is a minor feature change, only of interest to people using traffic shaping in BSD already?

as in 10^6 possibilities, ignoring backgrounds? For the curious, that means collisions are expected >50% of the time with ~1200 samples, and someone could easily generate collisions on purpose with little effort.

So it works well for a simple signal (which is the goal), but shouldn't be relied on if people might try to abuse it.

I'm trying to figure out exactly what this is, to be honest. To me it looks like their version of Amazon Web Services, for sharing compute resources exclusively within their community. But I can't tell whether it's just providing secure hardware to throw VMs at or if they've built the support infrastructure like a fault tolerant file system as well.

Edit: From http://nebula.nasa.gov/services/ it appears they have a distributed file system, but otherwise it's simply a set of networked machines.

I can't speak to what chrome does, but usually there would be rollup patches to take advantage of, cutting down the number of patches to apply dramatically. E.g. To get from 23 to 76 you might only need to apply 23->25->75->76. This would be slightly less efficient than going directly both in terms of time patching and patch size, but not terribly so (one can assume most patches are decently disjoint). Most importantly, this keeps the total number of supported patches manageable.

Check the block list on http://www.google.com/reviews/t . If it lists the pages you blocked through the extension, you'll be good to go. Otherwise you can always add them manually.

An NDA would not and could not cover abuse, and I highly doubt it covers your impressions either (e.g. "it was terrible and the interviewers were all rude!"), so I don't really agree with your first point. For your second I can see where you're coming from, although at least in this case I don't see droves of people interviewing and then saying the process was terrible. But there should be (and perhaps is?) an easy way to complain about bad interview practices companies have without running afoul of their NDAs.

For your third point, I don't understand what you're getting at. What logic are you referring to that is broken?

Sorry, how is that evil? If questions are known they can't be used anymore, and often confidential information can come up in interviews, so it seems only sensible. I don't remember hearing of it ever being used against someone, so it's really just a precaution against someone finding out and then disclosing something really important. I suppose I could have said that not discussing people's questions is only polite, for all the weight it carries.

*EDIT: added last sentence.

FYI, I believe interview questions are covered under the NDA, so if you did interview (and presumably had to sign an NDA at the time) you shouldn't actually be talking about the questions you were asked.

>and given global knowledge of where people are and where they want to go, it could plan the absolutely most efficient route.

Random digression, but I think you will find this is an NP-hard problem - I'm quite sure you can encode the Travelling salesman problem[1] within it. So the absolutely most efficient route is probably not feasible, although the system could certainly do a good job optimizing over what we have now.

[1] http://en.wikipedia.org/wiki/Travelling_salesman_problem

Are you objecting on the grounds that you think you know how they should interview better than they do, or that you think they actually are hiring the wrong people? I've always heard that Google is chalk full of amazing programmers doing really good work, so I'd say evidence is on the side that whatever they're doing is working.

FYI, search history is one of the parts of the account you have to put your password in again to view. So it should be safe from prying eyes.

Not really. Personalization, location, etc. affect a lot. I can't find the page I was reading, but it appears that in just running through all the first letters and seeing what is suggested (a is for...) some users are seeing personalized results. And if personalizations are showing up right from character 1, caching results can't really play a big role. Not at the user level anyways; results can certainly be cached further back, but multiple computers will be used just to decide if you deserve personalizations.

Even if that were true, then the value of clicks would decrease, prices would drop, and a new balance would be achieved.

Contrast this with http://freestylemind.com/how-much-do-you-value-your-time , where Oscar contends that the internet is filled with mediocre content already, so traffic-wise you are better off to write less often but spend more time on each post than attempt to be prolific.

They are certainly trying to make it as secure as possible. It is open source, they've held at least one contest to try to break the security of it (http://code.google.com/contests/nativeclient-security/faq.ht...), it is running inside a sandbox with limited permissions, etc.

While it is always risky to introduce anything new, that is a price we pay for progress. And that is true for anything - at this point, I don't think using it would be too much more risky than using a new javascript engine, for example. Aside from never trying anything new, what else can we hope for?

Also, I will point out that even in the dev channel it is still disabled by default. So there is some time yet for the security to be evaluated, to see if there are any endemic problems.

Just curious, what terrible things has DoubleClick done since being bought by Google? If anything, I'd say Google 'tainted' them and not the other way around.

Same question I've been asking all along. Why did Mozilla choose to spotlight these specific features? I assume there is a strategic answer, but I don't know what it is.

I'm sorry you think I was trolling, as I really was asking an honest question, and I appreciate your responses.

  > ... and learn about what you're talking about before you speak.

I have an interest in browsers and browser technology, but I don't develop any of them, no. That doesn't mean I am not allowed to discuss them, however.

  > Ok, if you want to know all the features ...

Please don't mistake me here. I am not at all trying to slight or belittle Firefox 4. As I dig into it, I think it has a great feature set, and I'm excited to see it come out. My question was specifically about the features chosen for the spotlight page.

Now, back to the discussion.

WebM: You are correct, Chrome 5 does not support WebM, although it is in the dev version of Chrome 6. I do think, however, that the perception of the feature is it was Google's first, considering its their codec.

WebSockets: Please don't try to pretend Chrome is ignoring the standards changes. The new version is in Chrome 6 for you to try, if you're interested. My point is that this is a feature they drove to market, and got everyone else to support, hence my perception of it.

IndexedDB: Again, the specific technology is beside the point. Google has been pushing for offline storage since Gears, and now everyone else is implementing it (albeit a slightly different backend). You see where I'm going with this.

Crash Protection: I stand corrected. I always think of Chrome being the browser pushing the segregated processes, better plugin separation, etc, but it appears IE8 was first. Still, Firefox is definitely not leading the market on this one.

JetPack: as per aboodman, I will call this a wash.

WebConsole: It is a bit specious to give Firefox credit for a 3rd party extension (FireBug), but regardless, this feature doesn't support my point, and I never pretended it did.

HTML5/CSS3: Every browser coming out has been touting these two technologies for ages, most of the time without a clear consensus on what they even entailed. At this point you couldn't not put them on a feature list. So I don't particularly consider it valuable or surprising to see them.

From my view, the highest highlighted feature was the tabs, followed by the 3 you mention. I counted WebM as Chrome, given that it is Google's codec, but I'm not sure its in Chrome 5 either :(.

Looking further down, I was thinking that WebSockets, Indexed DB, Crash Protection, and the key feature of JetPack (Add-Ons not requiring restart) were all pioneered by Chrome. That leaves The Web Console, which I have not yet tried so can't really comment on either way, and the two givens (HTML5 and CSS3). To my eyes this seemed odd. But I take it you disagree?

When I look through the new features spotlighted (http://www.mozilla.com/en-US/firefox/beta/features/), it seems to me that the vast majority are already in Chrome, some copied directly from it (tabs on top).

I'm genuinely curious why that is. Options I can think of are a) they are spotlighting features that take away a perceived edge that Chrome has, b) they really are playing feature-catch-up and letting other browsers determine the roadmap, or c) some features (the web console?) are far more ambitious than I've given them credit for.

Can anyone shed some light on what is really going on here?

I would also recommend The Metamorphosis of Prime Intellect:

http://www.kuro5hin.org/prime-intellect/mopiall.html

It is a (quite graphic) novella about the human race after the technological singularity, and I think one of the best pieces of writing on the subject that I have ever read.

Regarding the title of the article, the attackers stole the source to "Gaia", the single-sign-on system used by Google, and could potentially find previously-unknown security holes in it.

I think you may have misunderstood how the algorithm works. N is simply the number of items remaining in the working set, so over the course of the algorithm N goes from n to 0.

You might be thinking of when √n - 1 = 16, (i.e. the maximum size of the initial length the pivot is picked from is 16). In that case, n = 17*17, or 289.

Does that help?

Yes and no...it does use a pivot, true, but is a bit more distinct than just that. For one thing, it is iterative where quicksort is recursive, so it doesn't need any external storage for a stack.

Wow, I never would have guessed Flash was once called "FutureSplash". What a terrible name.