He has created and curated an amazing thing. Kudos to him. Few people have the dedication he shows. I respect his achievements immensely. He's in the top 5 contributors to modern computing in my books.

That doesn't make it respectable to be an asshole though. The world has enough assholes. Being the smartest asshole still makes you an asshole.

Maybe he couldn't have done it any other way. Maybe he could have. Don't think we'll ever know. I still wouldn't hire him, and there's no risk he would have ever been looking to work for me anyways.

I've worked with people like Linus many times; brilliant, know it, and abuse those who suggest ideas that don't align with their world view. They are usually correct when they shoot down an idea. However, I avoid them at all costs, both from a hiring and from peer perspective, because they are a cancer in most organizations, and the long term cost of dealing with their anti-social behaviour greatly outweighs the benefit they bring.

Abuse does not solve problems. Belittling others does not benefit anyone. It's fine to be blunt and explain why the idea proposed will not work, it saves time and energy. It is sign of a deeply flawed personality to attack the person who proposed the idea that won't work.

They accidentally the whole BlackBerry

http://coding.abel.nu/2012/06/programmer-time-translation-table/
4 hours is about the sweet spot.

In seriousness, there are many ways of improving estimates (reviewing past similar projects - you kept metrics right?), appropriate granularity of features and estimates of these features, confidence factors appropriate to the complexity/unknowns of the task (write a CRUD GUI screen? high confidence. Write a new algorithm to combine multiple videos into a 3-d pannable single video? Low confidence), etc. You need to be refining and grooming these estimates weekly as data changes, so at least you fail early.

Mad because a government official cannot wield undue influence over the global, borderless Internet?

Facebook treats all their users like shit, no reason this guy should be special. I agree it sucks for him, but it sucks just as much for the 13 year old whose classmates put up pages mocking him. I don't think you want it to be otherwise. Ideally they would always "do the right thing" but at least they are consistent in ignoring their users and don't play favourites.

How does this translate to real-world user-visible improvements?

Does it scroll more fluidly?
Do apps run noticeably faster? (Were they slow to start with?)
Is there any software out there that takes advantage of it to do more?

Can't see anyone caring about this at all in the mainstream market (you know, the one where people buy 50M+ units).

Also, comparing against iPhone5 is misleading as Android code runs via a JVM and iPhone runs native.

(Not a fan of either, aiming for objectivity)

I didn't bother to read the article obviously, but to compare opening weekend results directly with CAM downloads ignores many aspects. The most obvious to me is the people who did NOT go out to the theatre and who WILL NOT download the CAM, but who WILL wait two months for a high-quality free Blu-Ray rip to appear online. These are potentially lost sales for the theatres.

(Having said that, after going back to a theatre for the first time in a couple years specifically to see Avengers, I still believe the root of their problem does not lie with piracy, it lies with the appalling rudeness found in your average public gathering. For the same price, two months later, my living room is infinitely more comfortable and better equipped to show ME the movie in a manner I will enjoy and not be distracted by phones, screaming children, and poor sound).

Thanks, just ordered a starter kit!

I love the law of unintended consequences.

nLite is your friend. Slipstream drivers, service packs, hotfixes, plus configure/disable many of XP's annoying defaults

http://www.nliteos.com/