Monday, November 23, 2009

Chrome OS

A topic has recently come up on our forums regarding our reaction to Chrome OS and the possibility of REALbasic supporting it.

Chome's hidden agenda is to improve WebKit adoption in order to force Microsoft's hand into better supporting modern web technologies such as CSS3 and HTML5. Google doesn't plan to make money from Chrome, they intend to improve web standards and compliance so they can in turn write better web apps that run on all platforms. It is those web apps such as Google Docs where they can make money. Remember, Google is a business, they're not simply going to devote so many resources to Chrome OS just to help people. There is always a bigger agenda.

Now, I'm all for this tactic. Microsoft really does need to get their act together, and competition from Mozilla, Apple, and Google has already forced them into putting a ton more resources into IE. Will it work? I'm not sure, there are already signs that Microsoft is again trying to fight standards with IE9 just as they did in the Netscape days.

Chrome OS is an interesting concept, but I'm not betting on it. Apple tried the webapp-only model on the iPhone and failed horribly. Why would Chrome OS be any different? For example, Macs are frequently bashed for not playing games well. Chrome OS won't be able to play games at all (flash doesn't count, I'm talking about real games) and does anybody expect that gamepad, joystick or G15 keyboard to work? That's just an example, there are a TON of things users will be unable to accomplish with Chrome OS. Enough that I believe Chrome OS will remain nothing more than an interesting concept.

Now, I'm a power user for sure, so my view could easily be skewed. But it does not make sense for REAL to put significant resources into coding for an operating system that will likely have an adoption rate lower of that of Ubuntu. The only consumers who will get this will be the ones who recieved it on their netbooks, and only if said netbook will have it pre-installed. That said, since everything runs on the web, there isn't much we'd need to do. REALbasic can already compile code for Linux (most popular web server) and has an included CGI template, so go write a web app right now.


MGV said...

The way I'd prefer to see real basic do a web app is as another platform.

I would envisage this as being a split of the platform between client and server. So you would write the code as if it were for a standard platform. It would run server side, but the web page delivered would be the user interface.

So it would look like any other rb app, except perhaps that the event model may be limited by the events that the web browser could supply to the user interface elements drawn on it.

In other words - your standard rb app just gets ported to the server. The app generates the user interface not as a local application, but as a javascript enabled web page, with the user interface drawn on it.

Actions on the user interface would be sent by the javascript web page back to the server and presented as standard events on your code on the server.

Now that would be a platform worth supporting.....


Anonymous said...


Thanks for the post. I agree Google has a large web investment at stake and will do just about anything to promote it. However, saying that the ulterior motive is to prod MSFT into support standards falls short of what Google's ultimate goal maybe.

It's entirely probably that Google sees netbooks as a major platform. In fact, it's clear that SaaS is growing and both Apple, MSFT, and Google all have clear roadmaps regarding cloud computing.

Sure getting MSFT to support more web standards is an issue, but it's not the primary reason. Software as as Service, cloud computing, and web apps are the ultimate goal. That's my story and I'm sticking to it.

Eric said...

I see Chrome doing well in the consumer market. Cheaper hardware, access all your documents wherever you are, you don't lose stuff if your hardware gets stolen, (hopefully) no viruses, no need to back up as that is done for you... it will have a strong following, if marketed properly, especially with older people who will find it much easier to use.

I don't know much about RB CGI, as I write databases with fat client front-ends, but I do feel that there will be a market for a development tool that is easier to use than current web development tools, linked to some great templates so you can be up & running quickly.

Geoff Perlman said...

I'm not sure whether Chrome OS will do well or not. However, if it does, it will need to deliver more than just being an OS on a netbook. I don't see consumers flocking to it because they can access their documents from anywhere. That's a nice feature but I don't think it's nice enough to get consumers to switch to it. And the older generation is much more resistant to change so I don't see them flocking to it either.

Chrome OS could turn out to be a success but it's not clear to me yet what market segment will make it a success. If you look back at successful products, they often become a success due to a different market segment than was originally considered.

Unknown said...

As an user, I have absolutely no interest in the Chrome OS. From what I have read everything on the 'cloud'. None of my data will be stored on the local disk drive. That is unacceptable to me. There is simply no way I am going to give my data up to be stored on some system that could be anywhere on the planet. Not. Going To. Happen.

Unknown said...

I am more interested in Google Native Client that would allow regular applications to run in the browser.

Pharaoh Atem said...

Technically, it should still be possible to run REALbasic apps on Chrome OS, because GTK+ is still there, and though a strange convoluted way, you can access the root filesystem through a GTK+ file picker. I bet all the tools needed to extract stuff isn't there though.

Anyways, the stuff is still there to run local applications, given that Chrome itself requires a lot of what REALbasic applications require.

Unknown said...

regarding the inability to run games on ChromeOS one
ought to note that there are currently two startups with support form the games software industry that are aiming to bring gaming to the "Cloud". Thismight seem a technically impossible task but one has to bear in mind that in many developed countries broadbands speeds are reaching the point where such a browser based gaming architecture becomes viable.

Thers serious money behind this and the gaming industries see this as having many benefits to them market wise:

1) it eliminates piracy
2) it encourages casual gaming and impulsive gamers who dont want to go through the pain of installing a new game
3) it opens up high quality graphics gaming to PC/laptop owners who dont have and have no desire to buy a dedicated graphics card or a gaming PC, ( or even a console )

4) it makes software updates, distribution and marketing a lot simpler due to centralisation.

5) it allows for many new small startups to use a revenue model that allows their games to get out there and generate income without the need for packagind and distribution model of "boxed" games. See this as the equivalent of what has happened with the iphone app store, except now theres no need to download the software - ( or EVEN to "own" it - you just "rent" the title )

Because of all of the above i actually think google are on to a winner with ChromeOS.

The only downside i can see is that it - of course is highly dependent on a reliable ( and 99% "always up: ) broadband connection.

But thinking specifically of gaming, thats an area that can tolerate some downtime imo.

I've recently toyed with the idea of writing a REALbasic < javascript module for my in house Squirt converter ( a friend is already working on an Objective C converter for it) If you think RS can benefit from this code feel free to get in touch :)

The EAM-Dude said...

Doesnt Chrome support Google native-plugin standard? I think so. You should compiler your games with their C-compiler to a native plugin and that will run in Chrome.

Dave said...

Oracle did what you are suggesting several years ago in order to port their fat client applications to web-based. It works pretty well (although they had to invest a lot of effort to get the browser applet to bundle stuff sent to the server. The original implementation got bogged down in net traffic). It is a complex platform and even Oracle is trying to move to Java-based apps instead.