jamie

Well, they have to catch up to Facebook's overdoneness.

http://www.contemporist.com/2009/10/28/facebook-headquarters...

Is there any notable innovation here over tumblr?

This is tremendous advice:

"Jeff makes an insightful point that, when designing for scale, you should design for expected load, ensure it still works at x10, but don't worry about scaling to x100. The problem here is that x100 scale usually calls for a different and usually more complicated solution than what you would implement for x1; a x100 solution can be unnecessary, wasteful, slower to implement, and have worse performance at a x1 load. I would add that you learn a lot about where the bottlenecks will be at x100 scale when you are running at x10 scale, so it often is better to start simpler, learn, then redesign rather than jumping into a more complicated solution that might be a poor match for the actual load patterns."

Probably zynga.

This is at least interesting; articles from the last 90 days are for-pay, but once that window closes, it will be available for free.

How long until there's a reverse auction or minimum clearing level for articles? It seems the future of publishing news should be: pay for immediate access, as soon as the value (utility) drops to zero, make it free. "Free after 90 days or the first 10,000 subscribers see it"

Yes, but you can commute in it year-round. And with a fixie or single speed, the hills add at least a modicum of challenge.

There's no reason to say that SF proper is too small to take advantage of; I think the appeal is that here, you have year-round, every day exercise. That little bit adds up, just like every day of brutal cold in Chicago or NYC changes you.

That effects people's psyche, and I think that's a major reason why the attitude out here is different.

I'm amazed how caught up in the SF tech-scene people can get. Sometimes, it feels like anyone who ever used vi or emacs moved to the Mission or SoMa.

I know those are trendy places for engineers to live, but this city is really a bunch of small villages. Pac Heights, Nob Hill, NB, Russian Hill, Cole Valley, Bernal, Noe, and Hayes Valley are all wonderful, clean, "diverse" neighborhoods (well, diverse in the sense that your neighbors aren't engineers). I think a lot of people get myopic when they either live in the same neighborhood they work in or cluster too closely with people from their industry.

While Lovefest was indeed this weekend, so was Oktoberfest, the Fillmore Food and Wine festival, and the Hardly Strictly Bluegrass festival. The Bluegrass festival was free, outside, and featured some amazing music. http://www.hardlystrictlybluegrass.com/

There's a lot to appreciate here in the city that isn't tech related!

Don't forget that the weather here means you can comfortably go running year round. In fact, the san francisco half is in January. That's a huge advantage that SF-proper has. Soccer, running, and cycling are all activities you can do every single day in SF-proper.

For those that don't remember, the early days of TechCrunch were all about "who is getting bought today!" A big part of the "sellout to google" dream started from TechCrunch making it seem like Google would buy any and every startup.

Here's what TechCrunch wrote in 2005:

And as an interesting side discussion, Peter Rip, a Riya investor (and a friend of mine) takes us all down a notch (TechCrunch is specifically mentioned) saying “Every report was factually incorrect at the time it was printed. I did not see one accurate characterization of any discussions Riya.com may or may not have had.”

http://www.techcrunch.com/2005/12/18/google-passes-on-riya/

Yup, and while you can fault them for investing in phone operating systems and two aborted attempts at a phone os, arguably both of these represent the real threat to the MS cash cow.

It's execution (and fear of killing the cash cow), not market selection that's the problem.

What's striking to me is that the CEO letter mentions Moritz directly. I'm sure this isn't the last we'll hear of this story.

  > We asked our board members what they thought of the opportunity. Michael
  > Moritz, who represents Sequoia Capital (one of our investors and board
  > members), wrote the following: "You now have the opportunity to accelerate
  > Zappos' progress and to make the name and the brand and everything associated
  > with it an enduring, permanent part of peoples' lives... You are now free to
  > let your imagination roam - and to contemplate initiatives and undertakings
  > that today, in our more constrained setting, we could not take on."

http://blogs.zappos.com/ceoletter

  > At least two sources who do not hold board seats, but are directly involved
  > with Zappos, indicated that Moritz and Zappos CEO Tony Hsieh came into
  > conflict about the company’s future. Moritz, the sources say, wanted Zappos to
  > sell while Hsieh wanted to remain independent.

http://www.pehub.com/45388/zappos-ceo-wanted-to-stay-indepen...

Also, this job for HW engineer makes it much more obvious:

http://www.facebook.com/careers/positions.php?id=672

Funny, I can't find them linked from the careers site, just as ads in the right hand gutter.

What I think is interesting is that they're beyond what general recommendation engines can provide; it sounds like they're down to making domain specific optimizations. I wonder how many of these techniques will be generally applicable.

Actually, I think this is a pretty realistic - and appealing - view of startups, but it's also very much a description of what the experience of a VERY small startup is.

I think it's important to note that there are different stages of a startup's lifecycle. It's been my experience that venture funded startups aren't nearly as risky as people make them out to be.

My guess is that the typical venture funded startup has a lifespan of at least 4 years. That's probably close to how long the typical college grad will want/should stay at their first job.

Bottom line: join a startup, don't start one.

Re: Apple, that's not the story I hear from apple veterans from the mid-90's. I'm not sure how apple reduced bureaucracy (or if it did), but given some of the stories I've heard about the pre-Steve Apple, it was an awful place.

I'm a little confused by how Google plans to go back and add further optimization steps. From what I gathered from the video, they skip any IR step, and compile the javascript source straight to machine code. Lars Bak said a couple of times that the amount of code they generate was enormous.

If they skip building an IR, does that mean they intend to optimize the javascript representation itself to get performance? I'm a novice here, so I'm confused about where would they do things like unrolling loops and inlining methods if not against the IR?

"Founder and CEO Mick Mountz experienced the challenges of existing material handling systems firsthand while working at online grocer Webvan."

I wonder how many other dotcom flameouts went on to rethink the expansiveness of their plans, and started solving individual problems that caused their failure. Clearly, webvan in 2009 would stand a better chance than webvan 1999.

Don't confuse continuing to operate with continuing to innovate.

A few technical individuals can move the needle more than you'd imagine, even at a big co.

This isn't the case with the MBA. You have no idea how little it takes to push it over the edge.

There's a good chance that running anything on the external display is what's pushing it over the edge. If you look around, the most common complaint is that flash video (you know, non-HD youtube) is enough to kick the machine. In my case, it's Entourage.

Yea, I don't think people quite understand what the Macbook Air issue is. As a fellow MBA owner, I can vouch for this behavior. The system will become completely unusable while the 'kernel_task' takes over the system trying to cool it down.

I've experienced the exact same thing running a 24" LCD, firefox, and Entourage. Hardly an uncommon or extreme setup, but it's enough to throw the system off kilter. And when kernel_task goes into "cool this thing off NOW" mode, the system is unusable for 4-5 minutes.

The MBA is a beautiful monster.

Funny- That's exactly my story as well. I've met quite a few Boston escapees out here in SF who all tell the same story. Loved Cambridge; hated the winter.