They'll take a few webservers and their associated middle tier boxes out of service on their front-end loadbalancers, wait for all the sessions to migrate/fail over to others, upgrade them and put them back into service. The loadbalancers will be smart enough to do affinity (put this customer onto this pool of servers if possible). At any one time after the roll-out begins, x% of the customer base will be on the new code where x->100 if things go well, or if something unexpected happens, (100-x)% of customers will never even see it, and the rest (x << 100) will see only a glitch before the loadbalancers swing them back across onto the old code.

In apps like this the policy is to only add columns to tables and new tables, never to remove anything, so the database can be upgraded hot, and the old code can continue to run on it, only the new code will see any new columns/tables.

At Yahoo we drop columns from the previous version the next time we upgrade the database, e.g. if you stopped using a 1.0 column in version 1.1, then you drop that column as part of the upgrade going to 1.2. The key thing is always being able to roll back if things go seriously wrong; hardware is much, much cheaper than downtime.

Quote:

   In fact, the most effective anti-piracy software development strategy is the simplest one of all:

   1. Have a great freaking product.
   2. Charge a fair price for it. 

Is he joking? The World of Goo is a great freaking product. And The World of Goo is fairly priced at $20. And yet the piracy rate is still at 82% or so. So where does Jeff base his "in fact" anti-piracy claim on?

That guy sounds like a passive aggressive, condescending dick. He may have some nuggets buried in there, but they're difficult to find when he drops phrases like "while you get here at 9, and mentally check in at about noon".

I worked in an environment where the business owners subtly questioned the worth of all of their employees once and I'll never do it again. I hope that I'll treat people differently if I'm ever in a similar position.

For a moderate-sized upgrade, the description given by gaius is pretty accurate. Working at Yahoo, I've been around for a couple pretty huge property changes, and then the deployment process is very different.

Basically, hardware is much cheaper than downtime, and very big web companies have lots more money. So we don't swap out old servers gradually: instead we buy and set up an entirely new set of servers, deploy to them several weeks in advance of the planned launch, and run QA against these production-level boxes. Then, when we are ready to "launch", all we're really doing is a relatively low-risk DNS change: all the potentially tricky deployment issues having been ironed out beforehand.

After a couple of weeks/months of operation on the new hardware, if there have been no major problems, the old boxes are decommissioned -- either re-imaged and put back into service to expand capacity, or more often taken out of service entirely (I've no idea what we do with old boxes when we stop using them, funnily enough).

Q: What’s an anagram of “Banach-Tarski”? A: Banach-Tarski Banach-Tarski

I'm bothered by Jeff's attitude that a game made by a bigger company is somehow more deserving of piracy.

Very true for me personally. I built a small company with a partner that was doing low-six figures annually. I bought him out and was running the show. And on a weekly basis, I was scared to death that it was luck, a fluke, that I couldn't replicate it if it all fell apart.

It made me shake in my boots, be overly cautious. What if it was just dumb luck? What if the wheels came off? When I set down to put my second company together, and it succeeded, it was like a huge weight off my shoulders. It's like - okay, you drop me in a foreign country without speaking the local language, clothes on my back, and $20 in my pocket, and I'll be back within a couple years.

I can't say exactly what it is, but I think the idea of having something you don't "deserve" (or couldn't get back) is fundamentally really scary to people. You see the same pattern with people dating someone they perceive out of their league, and you especially see it with corrupt political leaders. For me, not knowing if I'd made it or gotten lucky would haunt me on perhaps a weekly basis until I did it again.

And now? If everything melted down, I'd go hang out on the beach and read some books for a bit, then jump back in the saddle after I got sick of reading or beachlaying. It's a much better feeling than wondering if it was dumb luck.

You see, Jeff Atwood prefers talking about software to actually making software, by a long, long shot. Most of his writings are nothing more than summaries of other people's work, sometimes padded with some theorizing and sweeping generalizations of his own. His writing style may make him seem like an expert, but he isn't.

I'm now starting to understand why the YC motto is "make something people want", rather than "make something useful".

> Sure, the numbers have dropped off since then. So what?

I'm guessing the problem has something to do with the 33 million in VC they took.

> It would be more fair - to them and to us - to call them a mixed success

Perhaps you could if they don't take any money but they did, and now they have to figure out how to get that 5-10 times return on the 33 million they took.

That's the reason people consider Cuil to be a failure.

"Charge a fair price for it." seems to be a meaningless platitude in this instance, sort of a soup-bone to the "I'm only pirating because it is overpriced" apologetics. Isn't the whole point of the anecdote that World of Goo is, umm, more than fairly priced? (Seriously, people are pirating a $20 pass to pure gaming joy which requires a machine minimally ten times as expensive to run. Probably a hundred times for many of the pirates, since they are gaming enthusiasts.)

Want a recipe for being pirated less? Make something that people wouldn't want to get for free. crickets chirp Actually, that probably won't even help -- I think a large portion of piracy is essentially pathological. Bingo Card Creator -- an which assists primary school English teachers who are, on average, not likely to be reading Chinese pirate forums -- got spikes of 10k+ downloads when the new cracks were released.

It even had a description translated into Chinese... which was hilarious. (My brother translated for me: "Apparently it's like a gambling game for English or some shit.")

Honestly its like putting a shiny object in front of that bird with the unhealthy fascination for shiny objects whose name I am forgetting at the moment. If your code compiles, it will be pirated. You could make a Windows forms calculator with only a plus button, with an equals sign available with a CD Key, and people would pirate that. (I've often thought of trying it.)

Did anyone else notice how Orwellian the UK has been getting lately? Not just this story in particular, but the main theme of the stories coming out about U.K.

Feynman's strength (well, one of them) was that he insisted on a simple example to follow during an experiment. This is an incredibly powerful technique for understanding complex calculations, and no doubt many hackers here use it when doing a code walk-through.

The problem was that the mathematicians fell into the trap Feynman laid and gave overly simple descriptions. Feynman seduced them into trying to give physical descriptions when it was, in truth, important to think about the details.

Banach-Tarski is a specific example of that, as many of you will know. I think Feynman knew full well what he was doing - just one of the games he played.

The technique has its limitations too. Restricting yourself to consider only the physical loses the power of abstraction. Many of you know that too, although perhaps you don't think of it that way. When you extract a method you're using abstraction - the method may not represent something physical. Feynman's technique can limit you there.

The story comes from "Surely You're Joking, Mr Feynam", in the chapter called "A Different Box of Tools."

http://www.multitran.ru/c/m.exe?a=DisplayParaSent&fname=...

Feynman makes a big deal about having unusual tools in the box - they let you solve problems that others find intractable. That's one real lesson. Acquire a wide variety tools, and know how to use them. Functional, Object Oriented, Imperative, Logic, Database, etc. They all have their place, and knowing them, really knowing them, gives you enormous flexibility.

Never stop learning new tricks.

If you have to ask YC for an idea of what car you want to buy, don't spend $50k for a car. Buy a decent used car for $6 to $10k, or a new car for $15 to $20k (I bought a Honda Fit; it's nice), bank the rest of the money, and wait.

I won't say you shouldn't blow $50k on a car -- I have friends who drive autocross and are really into cars -- but if you don't already have a short list of performance cars that you lust for you should just drive a compact Honda or Toyota for a while and be done with it.

What's wrong with the two-seater, again? Can't you just rent or borrow a car when you have to pick people up at the airport? You can rent a lot of cars for $50k. ;)