Welcome to WebmasterWorld Guest from 188.8.131.52
...which is that I think Google's going to build a Web-based thin client-type hosted environment-slash-operating system replacement. Or at least, they should, and that's only if Microsoft doesn't beat them to it.
Althought I loath the idea, she is probably right. I know I'm old school, but this whole "web as platform [web2con.com]" brow beating is more of a threat than a promise to me. It represents the end of computing as we know it and the start of something much darker. The web as platform mantra, could look like the Blade Runner future (state as mother/father) many of us have had nightmares about.
One need only look at the way Google Desktop search results obliterate the line between desktop and web to see how web-as-platorm (or Google as Platform) can quickly blur the lines between "mine and yours". If you are on a reasonably fast connection, you can not determine quickly whether those serps are on your computer or out there on the web. Think about that, the next time you type that credit card number into your computer...
One need only look at the way Google Desktop search obliterates the line between desktop and web
I don't see how it obliterates it. It's still very much a desktop application, it's not a thin client. Sure it has an HTML interface but that's not a defining merger of Internet and Desktop. My Television has an HTML inferface, but that's for network use, there's no link between it and the Internet.
Or am I misinterperating your point?
I will NEVER have anything from my computer/website on the web
I'm not sure the Web (in the form of HTML over HTTP)is the best way for this kind of thin client system anyway. Dedicated applications on the host machine are a better solution with data being passed over HTTP.
That said, Hotmail has proved very successful as an HTML over HTTP application, although some people use Outlook to access Hotmail (and Gmail) storage which i personally find a better solution.
I believe there are definite advantages to distributing different types of computing tasks to specially tailored servers which are specifically designed for that purpose. It's more efficient as a whole.
With most of the major tasks done away from your desktop, your workstation can the concentrate on its job which is displaying the data.
A couple of examples where I use the thin client model today...
Email. This is one of the most popular thin client systems in use today. I've already mentioned Hotmail. However a lot of people (and corporations especially) use a thin client model for email;. IMAP is one example of this. I personally use a remote server that is tailored to perform spam filtering. This elivates the processing load on my desktop machine, and when I connect with IMAP, it also reduces the need for storage on my desktop machine.
Broadcasting Rather than clogging my machine up with Radio card, I listen to radio through my desktop media player streamed over the Internet. I would love to do the same with television broadcasts to my media centre.
Remote Storage On a similar note, I use a dedicated MP3 server which holds digital copies of my CD collection. Although this is located internally, it's the same kind of concept, it's a dedicated machine which reduces the storage demand on each desktop computer or media centre. A similar concept is used for remote backup across the Internet.
I know I'm old school, but this whole "web as platform" brow beating is more of a threat than a promise to me. It represents the end of computing as we know it and the start of something much darker.
AOL, and before them CS, and others before have tried to capture the audience. If anyone has come close it's AOL. (There are millions of "family computer user types" who don't know anything exists beyond the swirling "A").
MicroSoft could more easily open up client server versions of the office programs that "everyone" already knows and uses and roll it in for MSN.
There will still be people that will swear by their favorite non-MS solution, but the market share they already have makes them the obvious ones to pursue this environment.
You can bet dollars to donuts, (sorry --- I don't know if that idiom makes sense anymore now that donuts can cost a dollar or more each), but anyway --- you can bet MS will develop anything like this 6 months to a year AFTER anyone else does it --- it's their way... as soon as someone opens a new category of computing, they give away a MS-Category-Killer v1.0 and keep giving it away... then when there's no competition, they will own 100% end-to-end computing and user data for the entire planet. (I got carried away, sorry).
Back to the G story --- the Ajax talk is intersting, but in the end it's HTML which does not make for an OS. If G did develop a client OS coupled with a 'terminal/device' with; removable storage; server data file management (including "delete"); data synchronization, (when CLIENT requests it) ...and you've got something.
But I can definitely see more and more of what you do with your not so thin client happening online, through your browser.
I can't see how a rich user experience could be delivered to the user using a browser as a browser is currently defined. Maybe the XAML stuff will help with this, but it is hardly a standard is it? ActiveX well maybe but what about the *nix & Mac people?
Heaven knows how I could go off and create an article with my webtop word processor. It would be like going back to the dark ages. Printing from your browser...lol that's a laugh. I haven't printed a HTML document yet on the internet that didn't go wrong in some way.
The browser should have become the 21st century operating system, but it hasn't. Why? Perhaps because the world #1 has no real competition.
We tried to put together a rich client type application (a wifi access point placement tool) for the web. But, with current browser technologies it just wasn't a goer.
Though desktops will never ever become obsolete, what Sun wanted to do will and is coming true via Google. We know now for a fact that there will never ever be one device that suits all tasks and preferences. The proliferation of various devices we use today for info exchange is here to stay.
What Google will do finally is help us integrate all of these via the web.
I dont now whether anyone else has seen this, but today Google is the BEST at distributed computer applications and remote server management. I know of many a bank that will be willing to adopt this technology. The future is Google without a doubt.
The spinoffs from the technology that helps them serve search results is enormous. I see big corporations struggle every day with simple 3 tier apps. How about farming Googles systems for better performance and efficiency?.
This is where google is headed.
I currently conceive of a new application service concept on average every couple of days (OK, maybe not that frequently ;) - and whilst I can develop, prototype and host these on my own servers there is the ever daunting prospect of how I would actually cope if one of my ideas ever "took off" big time. By being able to develop and host applications in the "Googlesphere"; I can concentrate on the application logic and user interface whilst "outsourcing" the scalability and resilience of the large scale web farm necessary to run such applications to Google.
By "Application Service concept", i'm talking about new ideas that are only made possible by the real time information sharing capabilities of an "always on" Internet platform. I'm not talking about "web based" versions of traditional software programs such as a word processor or spreadsheet.
Of course the platform would have to provide some core services, just as any operating system does; but these could be extended to offer single sign-on (a user could "login" to one of my applications using their Google Account) for example. Database service is another obvious one.
We know that Google are building the infrastructure that could ultimately support this - Googler's can already run their own applications on the platform (labs.google.com/papers/mapreduce-osdi04.pdf) for example.
Come on Google - open it up..... :)
I see the problem exactly the opposite way to you. To me the server end isn't where the problem lies. If you want highly scalable backends the likes of Sun & IBM either are building or have built large arrays of computing power that anybody with a few bucks can plug into.
The client end, the browser has not moved on much in ten years. What has the last ten years of development done for the browser? Not much really. A bit of standard munching CSS & new [X]HTML standards but thats about it.
Why can't I write a rich user experience through a browser? Perhaps because the market leader has a vested interest in the browser not becoming a platform. Worse still if the rich browsing experience helps people write web based apps that circumvent my office monopoly.
Microsoft is far more vulnerable than most people imagine. If Windows/Office lose their lustre they are history. Those two combined are the only two major division making money for them.
I beg to differ...from what I see Google is developing a web services based thin-client OS that people access from anywhere, any device, at anytime to find what they are looking for..
this is the promise that Larry Ellison was talking about 5+ years back...(to paraphrase him and his extreme distaste for the MS platform):
..."I don't need no stinking desktop OS...I store everything I need in the network (internet) and everything I need can be found in the network (internet).."
So Google is, in fact, creating a network-based OS, if you will...not the standard tired desktop OS...
Thin client computing...all you need really is a browser and all the functionality and interactivity is online in the app (that is GOOGLE.COM and all it's associated apps it is developing out)
The problem it seems; is that we have media types who pick up on Google developing an "Operating System"; and the only context within which they can think about it is in terms of a replacement for the only thing they know of as an "Operating System" - i.e. Windows XP etc.
Yeah, that's technically a way you can look at it, but I think it's easy to get carried away with this. What Ebay or Amazon or Yahoo provide you with are certain web based applications, and I definitely see the web moving more and more towards providing more web based applications, such as google search, yahoo/hotmail/gmail, but that's not something I'd call a thin client per se, a thin client has nothing in it to speak of.
The day I see a consumer migration to this type of Ellison vision is going to happen sometime around the time consumers in the USA stop moving their bodies around in 1-2 ton objects made out of steel and plastic and start using only public transportation.
Obviously Google, Yahoo, or any other company that is trying to make itself into a portal of some type, will try to offer as many services as it can to lock in its users, but that's a far cry from thin client computing.
If you've ever taken a close look at privacy laws regarding your emails stored elsewhere than your machine, you might find yourself slightly disturbed. Or not. The same laws would apply to any data you stored there. Those are the now virtually non-existent laws protecting your online behavior from unwarranted scrutiny, that most, but not all, ISP's hand over to almost any agency that requests them, without even the benefit of a court order or suboena.
Consumers are buying faster and faster pcs with bigger and bigger hard drives, the amounts of data they store isn't practical to support in terms of transfer speeds in a WAN type setting, what Ellison is really talking about is controlled corporate environments, where data is stored on central file servers, then moves over extremely high speed connections to the thin clients on that network. That vision is becoming very close to realizable now, and I think more and more smart corporations are going to start moving to some version of that, or to slim clients. But those are internally controlled networks, not public, very slow connections relatively speaking. For $60 a month you can get around 4 mb a second, on a decent new lan you get 1 gigabit, some 10 gigabit. Those lans transfer data as fast as an internal harddrive request roughly. That's really what Ellison is referring to, that's what Sun sells, or wants to sell, that has nothing to do with WAN thin client computing.
Very few consumers are going to give up the speed they are paying for to give all their data to some third party, MS tried this very thing and it was totally rejected by the market, correctly so.