Nightwind’s home

Sometime during last evening (EEST), Google launched its new product: a browser. Rumors about this have been circulating ever since 2004 but have been slowly dying as nothing official was announced. Still, I believe that few were those who truly though that Google has abandoned a possibility to strike at rival Microsoft on a ground that Google seems to dominate: the web.

Microsoft’s Internet Explorer has been dominant simply due to the fact that it comes within Windows and people are simply to lazy to see alternatives even when the product their using is more detrimental to their own computers. But lately, with the emergence of new technologies and with increasing internet risk (identity theft, phishing, credit card fraud, etc), the limits and downsides of the dated Microsoft browsing engine became more than obvious and alternatives such as Opera, Firefox or even Konqueror rose up and started eating away at IE’s market share.

Today, Google added Chrome to the panel of Internet Explorer adversaries and (as Google has accustomed us to) it brings a series of novelty concepts that are sure to be the death warrant of many things we were taking for granted in web software, both on user interface level as well as the actual unseen architecture of the software.

For one, Chrome proposes a greater level of decentralization by using not only tabs but also having each visual tab as a separate process, meaning that if you open an error-generating application in a tab, you won’t lose everything.

Secondly, Chrome has a new memory-management model that offers a more balanced memory-consuption scheme than Firefox (the presentation cartoon is a serious blow against the more memory-inneficient Firefox).

Thirdly, Chrome killed off the classic File/Edit/View/Help menu we are used to in Windows applications by using a very spartan interface that leaves more room for the site itself. The most used functions from those menus were crammed in two interface buttons in a very functional way (Internet Explorer 7 also tried to offer something similar, but there you could find ALL the said menus crammed and cluttered in a very confusing way).

Even if these promises would be the sole features of the browser, it would still be much to offer, but they’re not. Better embedded always-active anti-phishing, a more efficient Javascript engine completely rewritten, using Apple’s Webkit rendering engine (without its own Javascript engine though) and many other offers that would make Chrome the most performant browser ever are there to incite us. But now that Chrome is out, do they work?

Well, I would say so and so after spending some time browsing around with it.

Before I start presenting my findings, please be advised that Chrome isn’t final now. Just as they done with Gmail before, Chrome is out in its infancy for a public test. It’s version today is 0.2, not even 1.0 and as such it should be treated with latitude.

Bad news first. Chrome’s model of tab management doesn’t work quite as advertised. Opening a page by the process of creating a new tab then pointing it to a page leads to the creation of two processes in memory and only the second one goes away when closing the tab. The other one dies much later (sometimes, as now for example I have only one tab and still 4 chrome processes running) I guess, but didn’t figure a pattern as of now.

Also, the memory consumption thingie isn’t quite as advertised. Adding up the memory used by all Chrome processes when having just 3 sites opened makes an uglier picture than a Firefox with 6 tabs opened that’s been running for 3 hours already (all you Firefox users know what I’m saying). The browser itself still runs smoothly (Firefox sometimes chokes itself) and as I boast with 3Gb of RAM I’ve noticed no overall system performance decreasing but I’m sure other will.

As far as the interface is concerned, it takes a while to get used to but I must say I like it as far as regular browsing  is concerned. However I haven’t seen anything yet that would really help web developers (Firefox’s Firebug is definitely the best in this field, so I hope Chrome will bring up something similar). Sure enough, there’s a JS debugger there and a task manager that shows information about the open processes and tabs but it’s terribly inaccurate. If you compare that with the Windows process report in the Windows task manager, you’ll see that in fact the memory consumption is about 3-5kb greater than reported in browser.

Also, in large applications such as HI5 for example, Chrome does miracles. Those who know hi5.com are aware of the fact that it’s a buggy site, very slow, with a huge amount of problems that run the risk of killing your browser. Well, visiting HI5 under Chrome has turned this experience from a stressful nightmarea into something closer to what it was meant to be. There’s still the amount of annoying details (like having hi5 telling you that you have messages of notifications that aren’t there) but nothing’s perfect.

About the overall browsing speed, I’m quite undecided. I’ll just say that Chrome’s speed is on par with Opera, sometimes it seems to be even faster (judging by empty-cache visits) but on heavier sites like BBC they are much the same.

Like I said above, however, Chrome is still an infant and to that it’s a very promising infant. Given Google’s reputation, I am positive the Chrome will end up surpassing expectations when the time comes for its final version. I also hope the the Google people won’t start alienating users by bringing more clutter to the interface (though I think this should be the last of the worries, the Googlers are masters of functional web interfaces) or make compromises for navigation speed.

When it comes to JS/web standards, well, Chrome is too much of a hype. Chrome fails with 63% score and specifically failing the Linktest. So far Opera is head of the crowd with 83%. Next comes Firefox with 71% and Chrome falls in third with its 63%, all way ahead of Internet Explorer 7’s 12%.

Still, I’m completely looking forward to see Chrome and it’s first version.