Bruce Lawson's personal site

On Google Chrome

I’m a Web Evangelist for Opera, but this post does not represent the official Opera position.

It comes as no surprise that Google was working on its own browser. When your entire multi-billion dollar business is on the web, you’re inevitably going to want some control over the mechanism by which people view it.

I imagine that Chrome was built as a Google Docs viewer. Google owns search, so can serve up adverts when people are doing that. But most people spend a lot of the day typing into Word, Excel etcetera, and not looking at Google’s adverts. Similarly, most people don’t use the vast majority of the features of Microsoft Office, so don’t need to pay the large sum of money that the Office suite costs. If those people can be persuaded to use Google Docs, Google can simultaneously deprive Microsoft of some revenue while showing discreet ads in the document.

Google Chrome has Gears built in, so people can work on the document when they’re off the web, and save shortcuts to documents on their desktop, just like a traditional desktop application. It’s also why Chrome is relatively lacking in chrome and browser controls: it feels less like the web and more like the desktop.

It seems to me that Chrome is designed to compete with Microsoft Windows as an Operating System and Office as an application: Microsoft’s biggest revenue earners (as far as I know).

It surprised me that it didn’t choose Gecko (the engine that powers Firefox), having invested so much in making Firefox fly. But I guess it pays not to keep all your eggs in one basket. Certainly, it doesn’t want to kill Firefox; it still makes a lot of money from people searching through it. And browsers aren’t its prey.

Competition between browsers can only help a consumer, although it can hurt developers if it’s like the Microsoft/ Netscape browser wars. But Google chose Webkit, which powers Apple’s Safari, so it should be pretty standards-compliant and therefore be not too onerous to test, and therefore little of an overhead for businesses with websites or those who make them.

This competition will also help us at Opera. Firstly, on those services Google provides (like Maps, Docs and Analytics) we’ve had compatibility problems as Google didn’t test with Opera properly. Those days are over: it can’t be evil and deny entry to a competitor (which we are, in the browser market). Google now advocates testing across all browsers:

Internet users have an increasing number of choices for web browsers today, including Firefox, Safari, Opera, and now Google Chrome. Sometimes web pages look and work differently in each browser, so it’s important to test your site across all of them to ensure all your visitors can enjoy the experience you’ve designed. (My emphasis)

There are a lot of features of Chrome borrowed from other browsers. Google acknowledges the influence of Firefox and WebKit, but not all the things it borrowed from Opera, which is a shame, but we’re used to it: Ben Ward wrote on Jon Hicks’ blog

I think I’m just sour at their presentation of Opera’s features as their own. The whole comic reads like the speed dial and omnibar are ingenious ideas that Google thought up to revolutionise your web experience. Then whilst Mozilla and WebKit get special thanks in the final panel, Opera gets nothing.

I don’t even use Opera, but in an industry where we preach so much about attribution and respect for the work of others, I’m not happy at the way Google have just danced in like this.

I replied “if there’s one thing that my three months at Opera has taught me (apart from how to drink like a Norwegian) it’s that Opera is constantly overlooked when it comes to doling out the props for its innovations or initiatives.” (That could be why we’re even overlooked in the Chrome user agent string Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US) AppleWebKit/525.13 (KHTML, like Gecko) Chrome/0.X.Y.Z Safari/525.13, thereby identifying itself to web servers as 1) Mozilla 2) WebKit, 3) KHTML, 4) Gecko, 5) Safari and 6) itself.)

The best thing about Chrome to me is the exposure it’s getting. It was even on the TV news last night. Hopefully, people who always thought that the big blue “e” and the Internet were synonymous will now realise that there is choice in the browser market, and they will make that choice rather than go with the pre-installed browser. Some will choose Chrome, some will choose Firefox and some will choose Opera, and that’s how it should be.

What’s your take, dear chums?

Browser tests (added Sunday 7 Sept)

Techradar tested Chrome against Safari and Opera, concluding the overall winner is Opera, saying

We’re big fans of Apple hardware and software, but Safari for Windows is half-arsed. It’s great on the Mac, but we can’t think of a reason why Windows users would want it – especially now Chrome offers essentially the same rendering engine with a better interface, lower memory usage and better performance.

So Chrome wins? Not quite. It definitely has the edge when it comes to JavaScript performance, but speed isn’t everything – if it was, our cars wouldn’t have doors, roofs, stereos or air-conditioning. Creature comforts are important to most of us, and on that front Opera is the only browser here that really delivers.

(Last Updated on )

Buy "Calling For The Moon", my debut album of songs I wrote while living in Thailand, India, Turkey. (Only £2, on Bandcamp.)

15 Responses to “ On Google Chrome ”

Comment by Dave

I’ve avoided doing lots of site testing in Safari because I know it’s an edge browser in my domain space. But Chrome changes that I think, it’ll be more popular on Windows than Safari is, so there’s no excuse for not testing against Webkit. Sigh; I have enough trouble getting things to work across version of IE, let alone IE & FF and now WebKit.

Comment by Dan

“apart from how to drink like a Norwegian”

I didn’t know you were cutting back, Bruce?

Seriously, it’ll take off if people at places like Facebook start putting ‘We recommend Google Chrome for all our crappy widgets’ on their sites – which is what I’m betting Google is hoping to see happen. Social not-working may baffle many of us, but I know people who update their facebook page, every time they sit down to eat a bowl of soup. Browsers may stand or fall on how well they support some daft zombie-poker-thing or a word game.

Comment by Isofarro

Hey Bruce,

I didn’t know Opera had done the start page thing, and it certainly be nice for Google to credit Opera for leading the way. But… there’s very little innovation happening that doesn’t use any existing ideas. And in Chrome, there’s not many new ideas, but rather a combination of a number of existing good ideas. To list all the ideas used… we’d still miss a large number. But, it is still a credit to Opera that Chrome has taken one of its ideas. Well done guys.

My take. My first impression was: This isn’t a browser. This is a window manager – like Fluxbox, KDE, GNOME. The ‘tabs at-the-top’ allow multiple applications to run in parallel.

The JavaScript engine has been rebuilt from scratch for speed/performance and with big applications in mind. Looking back in time, Java hit a tipping point as a way of building desktop applications when it’s speed started nearing that of C++ programs. Now, there’s a whole host of desktop applications written in Java – bittorrent clients, IDEs, word processors. I see the same thing happening with JavaScript running on V8. (Although I feel that getting V8 to do both JavaScript and Python would be better)

I can visualise a lightweight operating system (Linux + X) that on startup launches Chrome. And within Chrome you manage the rest of your machine, much like we do with the Windows or OSX GUI right now.

Every/most applications is just a program that runs in a Chrome tab. We already have Google Docs, Calender, Reader, Mail, spreadsheets, presentation creators running as web based applications, there’s no reason why other similarly text-related applications can’t go the same way. With Flash, video/photo/audio editing is possible (good enough for amateurs/tinkers).

Chrome becomes a platform. That platform can fill the gap on ‘secondary devices’, like an ebook reader, slate tablets, mobile phones, kiosk interfaces, ticket machines, information displays, widget containers (desktop, as well as TV) – places where a full operating system and a very rich GUI interface is not needed, or not justified in terms of cost/performance/resource.

I’d like to see Chrome fill the gap Mozilla hasn’t succeeded in plugging with XulRunner. Konfabulator/Yahoo Widgets started life running on the Mozilla platform. But each widget had its own Mozilla platform runtime instance – multi-process runtime would have worked so much better, but it just hasn’t happened. Firefox 3 was initially planned to be a XulRunner application, but that’s still a vision.

The Iphone application SDK, as per Steve Jobs was HTML, CSS and JavaScript. Chrome offers that plus plugins.

Hopefully Chrome can be opened up so that developers could write desktop applications in JavaScript that run on the Chrome platform: Todo lists, aggregators, password managers, PIM, email, calendars, text editors, messaging clients, IRC clients, music player, video player, Rich document editor/annotater, blog/website editor, a command shell, ftp client, photo gallery, desktop search, file explorer, desktop configuration applets — and stil be a capable enough browser – the typical tools people generally use on their computers.

The ‘each tab is a separate process’ evolves browsers from a Win3.1 era to a Win95/2000 era. Multiprocessing improves the stability of the browser platform, enough to make it a platform for everyday applications, much like a stable and mature operating system.

At the moment, Chrome is just a browser, a bare bones browser (with a list of very nice, but not new, features). But a multi-process browser. It offers potential we haven’t seen before in mainstream open source browsers. It might just another Amaya, or the next Firefox.

Comment by Gareth Brown

I have to say i have always been a bit of a swing voter. At the end of the day all i want from a browser is stability, best use of resource, and most of all speed.

Three months ago i was a webkit user, currently im a FF3 user, but from what i have seen of Chrome here in the office all i have to say is Wow.

The thing is, all current browsers have a little something that someone doesn’t like. I’m a big fan of the DAD TEST, which means if my dad finds something for his computer that stops him Effing and Jeffing at it, it generally a massive Pass. (Currently for my dad it is Opera).

Personally, i can’t wait for Chrome it to be available on Mac, and i hope it makes all the other browser companies work even harder to evolve (I’m looking at you MS).

Comment by david Storey

Insofarro: There is more than just one Opera idea they have taken 😉

For your comment about Chrome as a platform. It uses web technologies. They should never be allowed to make it a platform. Web Standards should be the platform. The other way around it becomes a Google web, just like MS in the early days tried to make it a MS web. Opera already use Web Standards as a platform. In a basic way we created widgets and sent the spec to the W3C to standardise it so there could be one way of making mini applications using Web Standards. We also run on all the devices you mention, acting as the application layer, so companies can write user interfaces and applications for devices using web technologies. These use Open Web Standards, instead of pushing Opera as the platform. If they wanted they could switch out Opera nad use another rendering engine if we were not performing as they’d like us to.

When using web technologies, the web should always be the platform, keeping it as open as possible, so we don’t get sucked into a single vendor world. Google have started on the right footing by using an existing standards compliant engine, lets hope they don’t extend and embrace the web. They have Google Gears, which is not a standards, but at least—although not ideal as Google control it—it works across browser, or will when other browsers support it.

I’m looking forward for them to take the open web seriously by removing their browser sniffing that blocks us and other browsers from their services. They have started on the right foot recently by removing the sniffing on Docs and in the process of removing it on spreadsheets, but there is quite a lot more. I’ve seen services where they block Safari too, which they really can’t get away with if they accept their own browser but not one based on the same engine.

Comment by Isofarro

Hi David, good set of comments, something I want to think about… But a web development point I do want to respond to (as an individual – I do not speak for my employers):

“I’m looking forward for them to take the open web seriously by removing their browser sniffing that blocks us and other browsers from their services. They have started on the right foot recently by removing the sniffing on Docs and in the process of removing it on spreadsheets, but there is quite a lot more. I’ve seen services where they block Safari too, which they really can’t get away with if they accept their own browser but not one based on the same engine.”

(I sense, perhaps incorrectly, that there’s the underlying implication that browser sniffing is being done for non-technical reasons)

Google has loads of engineers and loads of people with Phds and other weird leet-speak-like credentials. People who can do funny things with the two thousandth digit of pi. But it seems like there’s not much depth in web development.

From what I can see, the Google web development approach is to write everything in Java, and use GWT which ‘abstracts’ the HTML, CSS and JavaScript away from them. In effect, GWT is their web developer.

Unfortunately that web developer can’t, or doesn’t have the experience to deal with cross-browser issues – the code is only as good as the compiler can generate. And that leads to browser sniffing as a mitigating approach – easy solution when you don’t understand the implications and impact of it.

I see the same issue over and over where an engineer (as in ‘not a web developer’) or a piece of software tries to emulate the skills and experience of a web developer. They don’t have a good understanding of concepts like rendering differences and browser quirks, progressive enhancement, accessibility. And browser sniffing is used to paper over those shortcomings.

So many smart people can’t all be that stupid to think that browser sniffing is a tolerable practice. I’d hope.

And I did absolutely love Opera’s Swedish Chef response to Microsoft’s deliberate blacklisting of the Opera user-agent string via browser sniffing a few years back. Hands down, that was the best response ever.


Comment by brothercake

Totes man, that was totally my first thought when I read the chrome spec – “lotta stuff borrowed from opera there”.

Actually that was also my firt thought when Firefox 3 came out.

And Firefox 2.


Comment by David Storey

Isofarro: You’d be correct that a lot of the browser sniffing is not for a technical reason. For example go to Google groups and try to upload a new profile photo. It blocks Opera and Safari. Now try masking as Firefox. It works perfectly.

i’m not sure how much they use GWT over handwriting code, but that wouldn’t be an excuse as I’ve worked with the GWT guys to help them make it work in Opera, and they even report Opera issues when they find them. They even wanted to try to get it working on all sorts of devices like the DS.

Leave a Reply

HTML: You can use these tags: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong> . To display code, manually escape it.