Wave is Dead

Given that I've called Wave's doom from day one, I feel I have the right to comment once more on this silly undertaking.

First, even Wave's start was problematic. Google used the Wave announcement to suck the air out of an event Microsoft was holding on the same day, which was at best childish behavior, and personally I consider it unfair business practice.

Then, conceptually, Wave was a complete dud. Wave didn't offer even one interesting concept – seeing each others' keypresses is not a concept. That so many people got their knickers in such a twist over this shows just how banal and superficial the software world is, and why some of our best minds leave it in disgust to sell beer.

Technically, Wave was even more flawed. It threw out everything we've learned over the past decades in building internet apps, and replaced it with an ill-specified, unimplementable set of protocols that sent pieces of an extended form (!) of XML around, statefully and non-idempotently. Web-nativeness was an afterthought, and offline use wasn't even a consideration. That anyone took this crap seriously (hello Novell) is another sign of the shallowness that hampers progress in our field.

Every cloud has a silver lining. If there's one good thing about Wave it's that it showed that people long for new approaches to computing. Wave was like a Rorschach test, in which everybody saw what they wanted to see. In the end there was not much to see, and what was there was hideous. But still Wave showed that there's a possibility of introducing new concepts and having them adopted. If only someone found an interesting concept.

Long live the internet!

Does it get uncooler?

First Facebook didn't want anyone else to have the word "book" in the name of his or her online community and now it doesn't want anyone to have the word "face" either.

Thankfully, people acting this way are building castles on sand.

There is a small but crucial group of people that specifically want Unix. These people are the "hackers" and should not be discounted. Almost every good feature in computer operating systems today, including most features in DOS, Windows, and Windows/NT, came from the mind of one hacker or another. Typically, the work was not commissioned by a company. It was done as a research project and then productized. Without these people, we make no forward progress. These are the sort of people that made NetBSD run all of those different applications. This sort of talent, energy, and enthusiasm can not be bought. — The Sourceware Operating System Proposal

Paul Graham on trends for the future

Paul Graham's talk at the Business of Software, 2009:



Here are some rough notes I took:

Bet On:

Innovation

People just don't build shiny metal things anymore.

Software

``Yes you can still bet on software.'' Biotech too, suspicious of cleantech, government may stop funding.

Efficient Markets

``This is a big one.'' (Example: Airbnb, efficient market for accommodation.)

Measurement

``You make what you measure.'' (HP)

United States

Only three risks: Bandits. Government. Invaders. ``Unless somebody is stealing the money, people will build new stuff and make money.''

Silicon Valley

``The Californian budget crisis is just two sets of politicians playing chicken.''

Small Companies

``We're getting this world that's higher resolution. Instead of bosses you have customers. More efficient. Small companies live from making money, big companies live from economies of scale.''

Economic Inequality

Moore's Law, Sort of

``Multi-core. Weird shit you'll have to do. There's a gap to be spanned, and somebody will make a lot of money spanning it.''

Things on Screens

``After dinner, there's 4 or 5 people typing into their laptops. People are not gonna start going out to local theater. Right now I have a suntan from my monitor.''

Server-based Apps

``It's gonna be complicated. Does an iPhone app count as a server-based app? Things will live on servers by default. It's not necessarily simply a web app.''

Super-good Customer Service

``It's gonna be easier for customers to switch. It's gonna be easier for people to find out if you have good service. Err on the side of super-good customer service. Your customers will increasingly design your product. The reason to do good service is selfish.''

Apparently Frivolous Stuff

``Bet on it. Facebook seems like the ultimate frivolous app.''

``Twitter is a new protocol: a nondeterministic messaging protocol. You send a message and don't know who receives it.''

Programming Languages

``Whatever the next popular language is, I would advise on using it.''

``Write your apps in lots of different languages.''

``Don't look down on scripting languages.''

Open Source

``Definitely worth betting on.''

``How many examples of companies can you find that have gone too far in the direction of open source.''

``Open source [your products] to the point where it seems you're going too far.''

``The limiting edge of open source is design. It's great for implementation, but you can't get design done that way. Software has given us a lot more scope for design. Now when I go up to my oven, it has instructions. Somebody has given those idiots designing ovens microprocessors. I would give a buy recommendation on AAPL.''

``Apple cares about the iPhone the way Google cares about search. Android is a hedge, originally against Microsoft. Who else is there? Palm? RIM? No. Mobile devices will win, and Apple will win in mobile devices.''

Design

``It's why the iPhone wins.''

``When I was young it said "made in", today it says "designed in"''.

Real-time

``It's not bogus in the way Web 2.0 was. Web 2.0 meant whatever was happening at the time. It's the computing equivalent of the switch from dial-up to always-on connection. All the existing protocols are based on this early [dial-up] model.''

``Wave is important, basically because it's the equivalent of Etherpad. If you make the convex hull around Twitter and Wave, and can think of something in there, go for it.''

Venture Funding

``They need you.''

Founders

``Founders will more and more have the upper hand. More and more founders will program. Programmers can learn to do business. You just make something people want, and charge them for it. They should have an O'Reilly book for business. It would be really short.''

Don't Bet On:

Credentials Granted by Institutions

``Admissions offices are bad. They don't check later how people they accepted or rejected did. Yet another artefact of an illiquid market.''

Business School

``Unsuited to the way things are done now. Business schools are the West Point of industrial capitalism. They trained the officer corps of that, not entrepreneurs. Great if you want to work for Procter & Gamble in 1965.''

Government

``Elected president with middle name Hussein. I'm impressed. But the guys in the engine room are the same people.''

Copyright

``With the copyright holders, it's gonna be an unbelievable fight, but they will ultimately lose. It will be so bloody, like the civil war.''

Restricted Flow of Information

``Everything's getting more liquid. Like hyperdrive.''

(Applause)

Facebook lol

Facebook's "global domination plan" has been here all this time & staring at us in the face- it's Facebook pages! — Facebook's killer app is here. No, it's NOT places.

At a party on Sunday I asked who of the guests had a Facebook account. There was one who had one, but he said he didn't check it regularly. Others said they knew people who have one.

I guess it's time again to quote Perlis:

Because of its vitality, the computing field is always in desperate need of new cliches: Banality soothes our nerves.

And here's a nice Hugh:

F-35 helmet

"Painting's washed up. Who'll do anything better than that propeller? Tell me, can you do that?" — Marcel Duchamp, on his visit to the Paris Aviation Show, 1912

Clients are a Lose, Right?

Almost ten years ago, Paul Graham said that clients are a lose.

Web apps have two big plusses over clients:

• As long as you stick to HTML 3.2, it will run on any computer.
• No manual installation and updates needed.

I think the second is the biggie, while the first is driving me increasingly nuts.

Simple web apps are write once, run everywhere badly. This is extremely pronounced on mobile devices, where interaction is so limited.

And once you go to richer web apps, the development process is simply insane, although tools, such as GWT, that abstract over all the different buggy execution platforms help.

All in all, I think the fight between the web app and the native client is far from over. Personally, I'd much rather write, say, a Blackberry app than a web app next time.

There's nothing like good old GUI programming to a Smalltalk72-style WIMP API.

WikiLeaks and the Afghan War:

The image we have is of an unidentified individual or small group working to get a “shocking truth” out to the public, only the truth is not shocking — it is what was known all along in excruciating detail. Who would want to detail a truth that is already known, with access to all this documentation and the ability to transmit it unimpeded?

The Collateral Damage of "Journalism"

I've always cut "journalists" a lot of slack – it seemed to me that I couldn't criticize them, because after all they were doing something really important.

But I've changed my outlook on this issue, because I've discovered that what goes today for journalism is actually harmful.

The tipping point was when a friend of mine who's managing a lot of people said:

Employees of all races, cultures, and creeds have no problem whatsoever working and having fun together, contrary to what the press would have us believe.

And I'm not talking about the designated rags, such as the Post, I'm talking about the "respected" rags, such as the Times.

Because they're backed by big money and thus widely circulated, they reach not only the people stupid enough to pay for them, they also cause collateral damage to innocent bystanders by sheer volume: today one is somewhat surprised to not find oneself in a cultural war at work, because "journalists" are constantly hammering us with faux "news" from the war of terror (among other things), copied and pasted directly from the prop-agenda-ists' feeds.

"Journalism": Good riddance.

Apple’s App Store Director Sells His Own Fart Apps

I'm actually not surprised.

"Simulate the experience of urinating for a long time," iWiz’s app description reads in iTunes. "Convince your friends that you’ll never stop. IWiz allows you to simulate urination: faster, slower or just a trickle."

An Apple spokeswoman said Shoemaker was hired partly because of his background as a developer. ... "His experience and perspective as a developer is one of the valuable things he brings to Apple’s developer relations team."

(via fefe)

Marijuana Legalization in California, Policy Perspectives

The title is wonderful, and it's funny to watch an intelligent suit from the BLAND Corporation with teh ugly slides talk about sensimilla equivalents.

The atheistic intertubes

Atheist seems to be a common answer when netizens are asked about their religion.

For me, atheism is basically nihilism.

For me, the existence of reality itself is a wonder, and proof of a kind of meta-level.

And I think God is a good name for the meta-level.

So I think I'm an agnostic. And I hope that all the people that answer atheist mean agnostic, and just don't know the different definitions.

Because what would it mean to be atheist? We believes in nothing?

There's even an old philosophical school that defends that thesis, and simply declares atheism stupid (but I forgot its name). I feel vindicated.