They seem to be using their own javascript engine, so it might need separate testing because of that. Although if it becomes popular enough, most javascript libraries and frameworks will surely support it.

There are actually very few inconsistencies in browsers implementations of the JavaScript language, it's the DOM that's the problem. If all they're doing is swapping out the engine for a JavaScript 1.5+ compatible engine, then we should be fine.

You can bet they will. The V8 engine is unlikely to present much of an obstacle. We have a large JS codebase and the Opera, WebKit (Safari), Spidermonkey (Gecko/Firebird), Rhino (Spidermonkey port to Java) and JScript (IE6/7) engines are no hassle to support.

The hard part is DOM interaction and styling. But as they're using WebKit and are putting a lot of effort into compatibility - this should be a breeze.

The really hard work is support for IE6. 1990s JS assumptions, very bad and inconsistent DOM behaviour, etc.

New, open JS engines with a different take on executing JS (Tamarin, V8) are wonderful news for JS developers everywhere. And consumers too, of course.

I know the portion of IE6 on the web is still around 25%.. but do you know what portion of your customers is actually using IE6 still?

I wonder at what point it is justified to return a 'please update your browser, preferrably to one of these: [list of reasonable browsers]', just because of the effort it takes you to keep IE6 compatibility.

At a small startup with a huge and complicated JS base, we've done just that. There's no way we can justify the man hours required to support IE6.

any link to V8 project? Comic strip says its just a module in Chrome and can be used extensively for other projects i.e. that means games/cool desktop/mobile apps

I think the article isn't serious. It's obviously meant as a joke. So lighten up guys :)

Very nice tutorial. It's good how he points out the useful libraries for web development with lisp.

facebook has become a useless spam machine. Doesn't work for me anymore.

NIH is the mother of invention :)

I agree. Google simply followed the the semantics of http. They had similar problems with the the Google Web Accelerator before. But it's really hard to blame google for the mistake of other programmers.

It really depends on the quality of people you have. This is a really nice way to run a company. But for it to work, you need to get employees that care and are willing to do their fair share for the company.

Obviously. But in a software company, specially a startup, if you don't have good people you were doomed before you started.

On the other hand, the whole democracy thing and the employees having stake on the company may even take care of or at least counter balance having bad people.

This new feature addresses what I think is the biggest complaint about EC2, as seen on their discussion boards. People used to complain about the lack of static IPs, now that's no longer a problem. Once they get around to rolling out this feature to the public, the only legit complaint that people would have for them is the base cost when starting out.

When your traffic is zero or very low, as in the case of when you're just starting out, a conventional VPS host is still a lot cheaper. But the good thing with AWS is that it's easy to port your apps to run on their platform.

"But the good thing with AWS is that it's easy to port your apps to run on their platform."

Once you have mountable volumes, you can pretty much run the LAMP-style stack of your choice unmodified, right?

Pretty much. You essentially get your persistent read-write /var filesystem. You can keep the read-only parts of your system in your boot image. Unix has a long heritage of working in such an environment. It's a natural fit.

The next low-hanging fruit is auto-restarting instances, so you don't have to have two instances just to babysit each other.

Way better than lolcode :)

I just use linux on my servers and OS X on my desktop. Works very well for me and I'm sure it also does for other people.