Five Things Google’s Chrome OS Will Do for Your Netbook
wired.com
wired.com
I've heard this a lot and I simply don't buy it. Assuming that the apps are Javascript-based (which all of Google Apps are today), the overhead of the interpreted language is going to slow things down as compared to something compiled down to the metal.
It is possible that Google will go the route of using its NativeClient, but that'd be a pretty big departure.
Welcome to the Java vs. C debate, previously the C vs. assembler debate, previously the assembler versus hand-coding machine code debate. Computers these days, even netbooks, are fast. Very, very breathtakingly fast.
There has also been a significant amount of work done in VMs, much related to Java, which has delivered tremendous performance boosts to interpreted languages. Google has a very talented VM team working on Chrome's JS execution engine (SpiderMonkey or something, can't remember off the top of my head).
But really, after you get past Wolfenstein done in Javascript (http://www.nihilogic.dk/labs/wolf/ ) how more does the typical user actually need? After that it isn't a matter of CPU cycles, its the programmers coming up with non-painful ways to develop software to use them.
First, SpiderMonkey is the next-gen JS engine for FireFox. V8 is the Google JS engine.
Second, Flash Player 10 is as good or better than any JS engine (with the possible exception of V8), and many Flash games bog on my netbook.
Just one example, but, as more capable web apps come online, the bar will rise, and more cycles will be called for.
You're right that this is similar to the Java vs. C (or C++) debate. And, as with that debate, the answer is that, for many application domains, native code has a significant performance advantage.
Javascript doesn't compile down to any sort of bytecode. It is just a chunk of raw source. Anything of the techniques that you'd apply (like JIT), would occur after a compile on the client.
That seems like a pretty fundamental difference to me.
And playing a 17 (!!) year old game doesn't strike me as a particularly convincing argument. You're effectively saying that javascript is some 2^(17/1.5) ~= 2000x slower than compiled C.
I'm all for high-level languages. But it's going to take some optimization to get there.
Things that come to mind:
- work when you don't have a network connection
- store a sizeable amount of media to entertain you while on an airplane
- use 'skype' and several hundred thousand open source packages
It does boot slower, but I find it hard to benchmark software that hasn't been released against software that does exist.
I wouldn't be sad one bit of google threw their weight behind linux adoption, for them to go and launch yet another OS to fragment the market is not what I was hoping for.
Yes article, I am looking at you: "it should be faster", "Google is promising boot times measured in seconds", "If Google can come up", "we might just have the ultimate", "Google may use these"
That is: should, promising, if, might, may. Not a single bit of concrete data at all, just pure speculation.
I know Google has released quite a few solid products, but aren't we giving them too much credit on something yet to be delivered? Google may know how to run servers and write web-apps, but why are we assuming they know anything about how to build a decent client OS?
If Microsoft was to release anything like this, with just as little data, the entire tech press would be up in arms about how this was going to fail.
I guess I'm getting about as tired about the Google-hype as the iPhone-hype. I use both, but that doesn't mean I believe in magic before I see it.
For good reason. They have ample precedent of huge Microsoft failures. Google, not so much (yet).
I believe Google will pull this off, because it entails repackaging and refining software that's already well on its way.
I wouldn't believe Microsoft will pull this off, for 2 reasons: (1) they don't want to threaten their Windows cash cow. (2) business partners (Intel) don't want netbooks to become an acceptable replacement for full-fledged laptops.
None of the reasons is technical.
The average person rarely understands the difference between an operating system, a browser and a search engine. And they're almost all pretty skeptical of FOSS, even after you explain it.
But most know, like and trust Google. And they know Google is fast and free. So when the Best Buy flunkie says it does the internet super-fast and the battery lasts longer, they're inclined to believe. There are no new concepts for them to tackle.
The up-side of a Google Operating System vs any other Linux distro doesn't really go to the user. It goes to the netbook vendor, who now has a far easier time moving Windows-less product and thus a stronger position when dealing with Microsoft.
Also, OS fragmentation is the logical end result of the web. The OS is still relevant, but it's not the main draw. So there's a lot of room for innovation, specialization and competition.
It's A Good Thing.