tedjdziuba

> It's typically difficult, if not impossible, to scale up and down your engineering team in the way you can spawn up new cloud servers. We're trying to change that by having a reliable source of contract engineers who can help us grow non-essential components of our codebase.

Sounds like an awful place to work. Treating your most valuable asset as a commodity? Let me know how that works out for you.

Gladwell is to science what homeopathy is to medicine.

"The idea of replacing spark plugs - a technology that has changed little since their invention 150 years ago - with lasers is not a new one."

There's a reason that spark plug technology has not changed in 150 years: it works. The mechanism of a spark plug (and many mechanisms inside of an internal combustion engine) are so simple that they leave little room for mechanical failure.

Everyone likes to make the joke about "If Microsoft built cars", but there is some amount of truth to it - automobile engineers value simplicity far more than computer programmers do, as it has a direct effect on reliability. The reliability of a car is far more important than the reliability of an online chat application.

Spark plugs as they are now solve the problem, and are cheap. You can buy one for about $5 where I live. That one plug will last years. If by chance it does get dirty and start to fail, I can pull it out and clean it with a wire brush.

The "fuel efficiency gain" from a laser plug needs to be substantial to justify the complexity and expense. Leave it to journalists to ignore the practical side of a story and make it all 'science'.

No, presumably if you're in a highrise, you are not running a retail business and your employees can work from home. It's an impact, but not a material one.

A better analogy would be setting up a retail shop on a busy street where every now and then, the police completely close off the street, and when you ask them why they've done it, they tell you that it's none of your concern.

Yet, people still continue to open shops on this street, that is, until the customers start going somewhere else.

They are scaling in the cloud, at least.

Reddit being down is not news.

I stopped reading at "static assets in memcached, telling nginx how to get them".

Lrn2CDN, or Lrn2LinuxPageCache

In Soviet Russia, regex starts you.

Sigh. Those who abandon Unix principles deserve whatever they get.

So if the elephant were dead by the hand of a poor African, it would be okay in your eyes? Income seems like an arbitrary factor to determine whether or not it is morally acceptable to kill an animal.

All of AWS, really. My experiences with EC2 were abysmal: unpredictable CPU and disk performance, uselessly slow disk I/O, instances dropping off the face of the earth at least once a week.

Reddit's recent downtime is a good indicator of how sloppy AWS as a whole really is.

Thanks for your reply. If you're not going to buy your own metal, then with that much cash, you're probably going to hire an army of systems administrators to be oncall for your customers, eh?

With a round that big, it looks like Dotcloud is going to buy their own hardware and become a full bore hosting company, which is fantastic, if it can convince people to move out of the EC2 slum.

Either that, or the founders took too much money and diluted themselves beyond reason.

Me three.

Damn, I was hoping to sign up and start paying you money, but since you're not launched yet, chances are I will forget about this service by the time you do launch.

I applied to YC and at one point during the pitch, pg said to me, in his best R. Lee Ermey, Full Metal Jacket voice, "Hell, I like you. You can come over to my house and fuck my sister!"

True story.

Yeah, that's it. Lease everything, don't own any assets. Let me know how that works out for you.

Seems like a lot of effort to avoid using cron.

Why would you want to do that? This isn't MongoDB, the real world cares about durability.

> We could make some speculation about the disks possibly losing writes when Postgres flushed commits to disk, but we have no proof to determine what happened.

If you read between the lines, this says that EBS lies about the result of fsync(), which is horrifying.

tl;dr: I'm vested. Peace, bitches!

Parse Error.

www.webpagetest.org is what you are looking for.

"...The Aristocrats!"

"How I Run a Website with Swap On"

Five Ideas with Microscopic Target Markets

Getting further along in the process of re-inventing SQL, I see.

Shenanigans. The author says that MySQL/Tokyo Tyrant don't work because of "death by I/O". Right, well, if your data set fits in memory, which I assume it does because you're relying on Redis, and the Redis VM feature is ghastly, then your problem isn't with "I/O", it's with MySQL not doing the right thing.

Buy some SSD drives, try out Postgres, but if you bet your business on something like Redis, you deserve whatever you get.

Of course there is. It's called trolling, and it gets attention. For example, the modern news media is built on trolling, as is most of 37signals's fame.

Cool story, bro.