slewis

So are you actually switching off Google?

"improve the platforms all these programmers are going to work on"

I'm intrigued. Do you have specific ideas in mind? Do you mean like self-driving cars? Or in home helper robots? All of the above?

Would be awesome but it seems hard to implement. Do you randomly choose half and tell them "sorry, no school, you must start a business"? If you let them choose which group they're in instead is the experiment biased? Ideas anyone?

I think there are people who have interesting ideas but who can't express them as well as others. The ideas may still be worth discussion. And of course anyone who doesn't want to comment doesn't have to.

Its a bit dissonant when a post is upvoted enough to be on the front page but the highest voted comment is about the post's style rather than content. The post upvoters were probably voting on content, not style.

Its a good point but less relevant to the discussion than some of the other comments here, and currently has 3x the points of all other posts in this thread.

Don't just upvote just because you agree, upvote because a post contributes to the discussion!

A fallacy? Sharing code between client and server is easier when the two are in the same language than when the two are in a different language. How is that flawed reasoning?

One example: you can run the same input validation in the client as you do in the server if you can easily share code between the two.

But there is way more. Here's an example of an application that shares all of its data models between client and server: http://tilemill.com/docs/models.html

Why desktop rather than web?

I'd argue that being in javascript is an advantage rather than a disadvantage. Sharing code between server and client (browser) is much easier, and there is no alternative to javascript in the browser.

I'm not saying javascript is the king of languages, just that we're limited to it because that's what's in the browser, and writing server code in the same language simplifies things.

Apple id seems to implement the proposed solution. They send a verification email but you don't actually have to click the link, you can just ignore it and your account works.

This can turn out bad though. I thought I had an apple-id when buying something on the apple site recently. But my standard passwords didn't work so I reset the password (via an email sent to me personal email address from the password reset sequence). When I logged in I found that my email address was actually registered to someone else, and I had their name, full address, phone number and credit card number but with the first 12 digits X'd out.

The person has a similar name to mine, and my email address is my initials and last name, so I believe they just made a typo in the email address when they signed up. But it seems pretty bad that you can do that without verification when doing so can give someone your personal information.

A motivated scammer could register a bunch of typoed email addresses and try resetting apple-id passwords. Then you have a 1 in 333 chance of buying stuff with their credit card because you have to guess the security code (I'm guessing you get 3 chances but you might get more).

Yeah we're a little concerned about the fax receiving service: interfax if anyone's interested. We've seen fragments coming in a few times and heard one report of a fax that didn't go through.

Ours, while somewhat jokey, is fully functional and we think could potentially have some interesting content: http://www.faxins.com/

Is that so wrong?

Thanks! Neither of us have slept. Wooooooooooooooo

I think that having "The Google-way" is certainly valuable to Google (and engineers). Because there are only a few ways to do things you can build something and then hand it off to another time (SRE) for day-to-day operations. It also makes security and machine management much easier to deal.

I'm not saying you're wrong, just pointing out the flip-side.

Maybe Google should have a separate network with separate virtualized machines and no "Google stuff" for new medium-sized projects. Test whether they get traction there and then port them into "the Google-way" later.

Or maybe "the Google-way" just needs to be fixed up to make it easier. For example you could use AppEngine.

I wanted to point out that donning Color "just some photo-sharing app" is probably not accurate. The investors saw something else. What did they see?

Cool, I just wrote a blog post trying to give Credulity too, but from a different angle. I think the app could be something huge, especially with the people and amazing tech they have, as you say.

http://news.ycombinator.com/item?id=2366137

I think people are missing one of the key features. This is described in the video here http://vimeo.com/21413899, but not discussed in any article I've seen.

When you turn on Color it starts recording all of the pictures that others in your local area are taking. All of those photos _are stored in your "visual diary"_ forever!

This is a "life recorder" app. Remember this: http://www.wired.com/culture/culturereviews/magazine/17-09/p... ?

In the future you'll be making a documentary of your life just by moving through space, and I think that's a beautiful concept. Your grandkids will be able explore how you lived your day to day life as witnessed by those around you.

Try it.

Right, if you are to learn to use a tool in novel ways you need to understand its behavior at a basic level: it is better to experiment with it from the outside until you understand its "Tao".

On the other hand if you need to achieve a task whose outcome you won't rely on to reach other ends, like say buying airline tickets (ha), its probably better if you can offload the thinking to some other entity.

When I was at Google I worked with very large structured protocol buffers in Python at one point. A single piece of data could be hundreds of MB in total, consisting of millions of smaller protocol buffers. I was doing a pass over the whole structure so needed to access each smaller PB from Python.

One day I decided my program was too slow so I profiled it and saw that the hot spots were in the Python protocol buffer implementation. "Easy", I thought, "I'll used SWIGed c++ PBs instead." Made some changes and ran the program again. Almost the exact same run time as before! I profiled again and found that this time the hot spots were in the SWIG layer. I was making so many calls through SWIG to c++ (because I was walking millions of objects), that using SWIGed PBs v. native Python PBs made no difference to my run time. Maybe I could have done some more custom SWIG work to lower the call overhead, but I remember being convinced at the time that SWIG wasn't going to do the trick.

So I ended up writing a 30 line Python extension that processed the protocol buffers in c++ and put the data into Python data structures. Run time was reduced by a factor of 10, hooray!

Sorry, didn't see this earlier. In addition to project awards (which are significant as you say) there are individual bonuses and stock refreshers for top contributors that can also be 'significant'. Sorry for the vague terms. Further, getting promoted to the upper levels can be rewarding as well.

But you are competing against 'top-notch' (to quote op of this thread) engineers for those types of awards. You'll also sometimes hear complaints about politics in the promotion process. It seemed mostly fair to me based on the promotions of engineers whose work I was familiar with.

I used to work there, started post-IPO, left in December 2010. IMO the answer is absolutely. But that's all I feel I should say.

Please put an edit at the top of the article in addition to the one at the bottom. This article has the potential to waste the time of those who don't read all the way through.

Con Kolivas. He's also mentioned in the article.

http://en.wikipedia.org/wiki/Con_Kolivas

Nice work. But be careful about relying entirely on something like this. There are more than just seek time differences between hard disks and SSDs: they have different internal cache behavior, different latency variances, queuing differences (depending on how the devices are configured and how you're using them) etc.

This seems like it'll get you 80% of what you want. But it'd be useful to have an actual disk to test on as well.