When a “Web OS” hits, it will be so much more than a browser launcher!
Everyone got in a tizzy over the Google very-early-darn-leaks-pre announcement of Chrome OS. There have been many attempts at a “launch into browser” system, and the timing / marketing hasn’t been right. The journalists have by and large shown that they do not understand any nuance and could only think to write about the external battle of the titans (Chrome OS to destroy Windows!) or the internal battle of the titans (Android vs. Chrome OS).
The news is so early, that it is mainly a no-op right now, other than showing that to get to that point, Chrome will probably be extended quite a bit (access to more native services etc). I am a believer that this has to happen (hence Gears fan), but I also do worry that it will be rushed. We don’t want to fork the Web….. but let’s wait and see.
What I find most interesting though, is that I don’t think that the world needs a computer that boots almost-solely into a browser.
Kent Beck was talking about how he is using Chrome Browser OS right now as a social experiment. Yes. We can all delete every application from our machines and maximize a browser and live in it. That isn’t what will be exciting about a true Web platform.
I have been playing with a Windows laptop, and what has interested me is how much I can get done because my world is available through the browser. Back in the day, if I was at a friends house, I would download putty.exe and telnet/ssh into my world, and it would be ugly text. Now, I can jump on a random computer and access my entire social environment (email, twitter, etc). I can get real work done. One of the reasons that we did Bespin was to extend that to coding too.
This brings me to my benchmark for when I think a web OS will be here.
If I can sidle up to a random computer and have access to everything that I can do on my own laptop, then we have made it
Ironically, I almost used to be able to do just that! In university I could login to any machine on campus and see my screen exactly as I left it. I want X back baybee! ;)
We are getting closer, yet still a ways off. I hate the management that I have to do with iTunes. I want to by the license to play music, and have my collection in the cloud, and sync’d in smart ways. I want to be able to set profiles for what is downloaded and offline-available on my laptop versus the media server vs. the ball and chain’s laptop vs. my phone.
I want my settings to follow me around. Another reason why I am so excited about the future of Weave-like services. Let me login to the browser and get my history, passwords, awesomebar setup for me. Go further, and let me “switch user” on the persons machine and have the login be an identity service like Open ID, Facebook, Weave, or what have you.
When we have an open identity system that works, with services that let you take your entire world with you hooked to your identity, then we can talk about how cool the Web platform is as an OS. We will have moved up the services stack from managing the state of windows on the screen, to higher level services.
Marry this to auto syncing of data between all of your devices, and you have a truly new world. Give me that world, not just and autoexec.bat that launches browser.exe :)
July 16th, 2009 at 10:57 pm
My intention with my experiment wasn’t to say that all we need is a browser, it was to find out what would happen if all we had was a browser. My conclusion was that it wasn’t half bad. If there’s more to the Chrome OS story than that, so much the better.
July 16th, 2009 at 11:20 pm
We’re definitely moving toward your benchmark. I’m not sure if you were just using your music management as an example, but I dropped iTunes a few months ago to go with Lala.com. I moved my entire library to the cloud, purchase web-only licenses to songs for $.10, and even wrote an Addon to control the web player from Firefox (Lala Control).
It’s a huge step in the right direction and I look forward to other paradigm-shifting web apps in the near future!
July 16th, 2009 at 11:27 pm
“buying the license” is something I’m also passionate about and I think would be a fun and elegant thing to work on. Unfortunately, the real challenges are the business ones, not the technical ones.
July 16th, 2009 at 11:29 pm
autoexec.bat ?! :-) That makes us old! Great post
July 16th, 2009 at 11:31 pm
I’ve been thinking about this as well, trying to imagine not what Chrome OS is actually going to be, but what it potentially could be. This gets especially interesting if you don’t tie your imagination down to the whole pragmatic (and responsible, and good) “apps work in app other current browsers” limitation, but rather start thinking of the capabilities of the future browser marketplace.
So here’s a labs project I’ve been thinking about: The goal is to set up a comprehensive road map from where we currently are to the totally perfect Web OS of the future, and then render this roadmap in a physical poster. My hope is that over time the poster actually serves as an effective piece of propaganda, and that we all start to build and work on the various components along the path. Worst case scenario it’s a fun thought experiment and great piece of nerd swag.
The road map could include things like:
-moving local storage for Web apps online into Weave (just calling it “storage”)
-data portability (”file open” is dead on the Web, and this is bad)
-file type associations to Web apps
-access to local devices (cameras, USB storage, etc.)
-notification
-identity
-semantics
-sharing (an OS that understand the concept of friends)
-additional UI widgets (but some of them shouldn’t make the transition)
-a decent layout model
-leveraging a centralized aesthetic (kind of like GTK icon uplifting, but applied to online apps, and with really advanced things like you specify the shape, we generate the color and texture)
For some reason I’ve always pictured this roadmap poster looking like the Final Fantasy X sphere grid :)
http://people.mozilla.com/~faaborg/files/daf/sphereGrid.png
When the browser learns “Ultima,” the desktop is dead.
July 16th, 2009 at 11:39 pm
Regardless of whether you *could* log into your environment from someone else’s machine, would you *want* to? Even if you trust that person, do you trust their security?
July 16th, 2009 at 11:53 pm
@Kent,
I didn’t mean to say that it was “only” that….. rather that you can have that right now!
July 16th, 2009 at 11:55 pm
@Peter
lala.com is very cool indeed. I tend to listen to my music more when I am not on the computer these days…. so hopefully it will expand to go with me.
July 16th, 2009 at 11:56 pm
@Aaron,
Totally agree. There is a tough future for music rights, video rights, etc. Interesting times ahead :)
July 16th, 2009 at 11:57 pm
@Alex,
Awesome thoughts. We need to riff on this at the next labs all hands ;)
July 16th, 2009 at 11:58 pm
@Anonymous
I trusted X :) A security model that makes sense needs to come along for all of this work, for sure.
July 17th, 2009 at 12:02 am
@Alex. This is an interesting idea. I think that one of the key limitation of browser based applications is that it is very hard to make them inter-operate. By that I mean that if my application wants to allow users to send an email or if my application needs access to the users address book, or etc., I have to do all the heavy lifting as part of my application and the browser is not helping.
As a majority of application move to the web and become services, the browser could evolve to reduce the barrier to interoperability and manage security and permission as part of the process. The android application mode includes some of that though the notion of activities. Aza’s vision around intent and tasks has some of that. I think that it is time for people to take a few simple use cases and make them and reality. Mozilla is uniquely positioned to make that happen.
July 17th, 2009 at 3:42 am
Dion,
I’m more negative than you re. Cr OS – I think GOOG is just trying to push people onto the web, when it should be trying to pull them in with better apps.
As to what ‘better’ apps means – well, here I agree with you. The web is creating for sharing information, and web apps currently accomplish this in a very rudimentary manner. There is a lot of scope for creating a layer of open interoperability, which would greatly enhance the user experience.
I go into more detail here: http://lmframework.com/blog/2009/07/chrome
Cheers
D.
July 17th, 2009 at 6:34 am
100 years in the future our grand children may discover that people were using things called “hard disks” and realize that life will go on if they disconnect themselves from the omni-feed.
July 17th, 2009 at 10:00 am
That mean You want eyeOS!
July 17th, 2009 at 11:00 am
@David,
I am net neutral on Chrome OS. Let’s wait and see what they actually produce. It is a ways out, and I agree with you that what is more interesting is better experiences, not the fact that it is the Web.
I hope, and think, that they will be pushing behind the browser launcher view that some people jump on, and I hope to see better experiences for users….. and part of that is privacy and identity, and taking your data with you.
Cheers,
Dion
July 17th, 2009 at 3:04 pm
(Please put a ‘u’ in “I want to by the license to play music”, if that’s what you mean.)
The solution to music files and local storage when you “sidle up to a random computer” is to carry a USB flash drive (or for those darn kids, your iPod) with you. Browse file:///media/MY_USB and you can play your OGG files in Firefox. I’ve suggested enhancing Places and merging Download manager for Firefox/Namoroka to make it more usable for managing local files, since in the new world almost every file you have locally came from the web.
I hope Mozilla community is thinking about this because in all the lightweight web OS initiatives so far (Chrome OS, Palm Web OS, that Crunchpad thingy), Firefox is at best an afterthought. Maybe that same USB flash drive with your data can have a Live USB partition that boots into Firefox on it. Is anyone building a lightweight Linux distribution based around XULRunner apps? There’s Mikaelrogers’ mega-awesome “Duke Nukem OS” but some haters act like it isn’t real. ;-)
July 23rd, 2009 at 2:57 pm
Like Alex, I’ve also been thinking about what is needed in order to make the-broswer-as-operating-system a viable future. (This is a bit long, and apologize in advance for that.)
1. Firefox needs to be able to handle almost any file type thrown at it. On top of that, it needs to be able to access the information in that file is a manner less limited than the rectangular black hole that the current type plug-ins produces. I envision a new generation of plug-ins that automatically convert content into standard HTML, allowing it to be exposed fully to the browser and the user. Thus, a PDF plugin would convert PDFs into standard text and images, using either its own algorithm or by using a third-party service (such as Google Docs), and hand off the output to the browser, which would be completely in charge of displaying the result. This would allow the user more control over all that content that previously needed non-browser applications to manipulate. I imagine Ubiquity would be very useful here.
2. Firefox needs to facilitate finding all that content that one’s file browser previously handled. This includes not just the files on the current device but also all the content one has in various “cloud” services. This would be done through specialized indexers that one could install. For example, a Google Docs indexer would index all one’s files on that service and expose them through Places as special bookmarks, automatically tagged and updated in the background. These indexers could also serve a secondary function of providing a mechanism to write to all those locations in the cloud. Thus, saving an image to Flickr, or moving an image from Flickr to Picasa, would be trivial.
3. The files that one has on one’s home machine would need to be available anywhere. Like Opera is doing, Firefox could also serve as a server, allowing one to access all those files using any browser (whether on that machine or elsewhere). Moreover, Firefox would also index all those files, making one’s browser the primary means of accessing any and all content that one has.
4. Since one is not always online, and, moreover, the web isn’t static, there needs to be a way to access pages when not connected or when those pages are no longer available. So, like Internet Explorer did ages ago, Places could allow one to choose to make an offline copy of a page when bookmarking it. This would replace the ‘Save Page As…’ functionality. Thus, one would have guaranteed access both to important documents and to articles that are not necessarily important but that you’d like to have access to in the future.
Once all those pieces are in places (in addition to the other ones that were already mentioned), I can see Firefox replacing the the whole Desktop—not just for casual browsers, but for power users and hackers too.