[besson3c]

I really like this article about why Google decided to get into the OS business after sitting out for several years (despite the fact that it had been talked about for over 5 years now):

http://www.technologyreview.com/web/23140/page1/


This guy brings up several really interesting points. For starters, I constantly underestimate the popularity of apps like Facebook and MySpace. Of this current generation of computer users, I would definitely say that these apps have surpassed MS Office as being the most familiar territory for them. This carries several ramifications.

For starters, the interfaces that these users will be most familiar with are not the traditional Windows OS interfaces, but the more platform agnostic and non-standardized web interfaces. When you throw in all of the different phones and netbooks that are accessing the web, it's got me thinking about the best way to design web apps. Maybe the best approach is not to mimic Desktop apps, but to mimic Facebook and the like?

Secondly, as far as Google and Microsoft are concerned, this presents opportunities for anybody that is interested in competing with Microsoft. Windows is designed to be a desktop OS that delivers desktop apps. This is part of Microsoft's DNA. Windows does little to integrate with the cloud, although it looks like they want to go this direction with their next version of Office. If these cloud apps do remain the most prominent apps around (be it Facebook, Google apps, or whatever), maybe it makes sense to accept that these interfaces are going to be platform agnostic, and work in integrating the OS to these apps?

The OS is supposed to be an unintrusive enabler to run apps and get work done, so maybe holding our breath waiting for the cloud to integrate with all of the various OSes the reverse should/will happen? Perhaps one day in addition to getting dock notifications for the number of unread messages in your OS X Mail mailbox we get Facebook notifications, or whatever else?

With the advent of phones and netbooks and the like, it seems like trying to sort of model everything after desktop apps is a horse that has already left the barn. Until Microsoft really positions themselves as being a leader here, I do see a great opportunity for competition. Microsoft's earnings lately have been down, perhaps I'm not the only to assess this? Is this the beginning of the end, or at least the chipping away of the MS empire?

Any comments on the article? A lot to discuss there...

[besson3c]

Also, if leveraging the familiarity of something like Facebook is the best way to visually communicate with an interface, what does this do to the whole academic study and understanding of usability? After all, there are no clear rules/standards with these interfaces.

[iMOTOR]

I think the biggest barrier for cloud computing right now is residential upload speeds. There are some people in the US who still get less than 1mb upstream.

As far as Facebook goes, Facebook is great for Facebook, personally I think it would make a horrible OS.

[besson3c]

iMotor: just in case you misunderstood me, I never claimed that Facebook or the cloud should *be* our operating systems. We'll always have local operating systems, but what drives computer sales and computing as a whole are killer applications (one of the last ones being the internet itself). An operating system by itself is not a killer app. Therefore, the computing world revolves around killer apps, not the reverse.

I'm not suggesting that Apple or Microsoft make their OSes mimic Facebook necessarily, because they are suited for such different purposes, but all of the apps that we run and things we do might start mimicing these interfaces over time. If you want to leverage familiarity, these is the direction you go.

I would also suggest that the official "Mac way" is dying.

For years loyal Mac users have pushed software vendors to support the Mac better. Now with loyal Mac users having to pressure makers of these big cloud apps in addition to the usual desktop vendors, I just not see how the "Mac way" will survive - it doesn't have enough leverage. I'm sure parts of the Mac experience will be salvaged, but I don't think we can count on future versions of Facebook or Google apps or something adopting Mac conventions. For starters, they can't adopt things like the OS X Dictionary, for instance, because of the separation that exists between the web and OS, but even for little things like the close window widget on the left hand side... Why should they adopt Apple's way of doing this? Their current interface seems to work well for them, whatever conventions those entail, I don't think people are refusing to use these apps because they don't behave like Mac desktop apps, and I don't see this changing.

The Mac way will live as long as people use OS X, but it will be pretty irrelevant in terms of how apps are designed.

[besson3c]

In fact, I'm sure I'm not the first to come up with this anointment, but this is the official besson3c anointment:

Social networking apps are a killer app.

[iMOTOR]

Originally Posted by besson3c 

If you want to leverage familiarity, these is the direction you go.

Hasn’t Apple been doing this for years? Mail, iTunes, Address Book, iCal, Keynote, Pages, iWeb, iPhoto, all more or less have a homogenized UI. To me, OS X Unified UI is as good as it gets.

[besson3c]

iMotor: yeah, but how does that relate?

[iMOTOR]

You just spoke about leveraging familiarity, Apple does this very well. Apple even mentioned in their adverts “If you know iTunes, you know iLife”.

[kylef]

Save the trees, download (and upload).

[besson3c]

iMotor: I agree, but I still don't see how that relates to this discussion?

[Railroader]

Welcome to 2005.

[Wiskedjak]

Originally Posted by besson3c 

Also, if leveraging the familiarity of something like Facebook is the best way to visually communicate with an interface, what does this do to the whole academic study and understanding of usability? After all, there are no clear rules/standards with these interfaces.

As a usability specialist, I'm leveraging Facebook all the time. Not the interface design of Facebook, mind you, but the aspects that make it popular. Namely, the social aspects of it. People like to share and see what other people they know are doing. When it makes sense, I'm always incorporating social elements into my software design recommendations.

Also, usability is as much about consistency as it is about ease-of-use. If an interface has low ease-of-use but is consistent with other interfaces that the user is familiar with, I'll recommend that interface over one that is easier to use but that the user isn't familiar with. I will, however, also recommend changes to the low ease-of-use interface to try to transition it over time to become easier to use.

[OldManMac]

I'm always intrigued when people talk about "cloud" computing, especially here in the U. S. We don't have the infrastructure in place to effectively use it. As someone already pointed out, broadband speeds are spotty, at best, and far too many people don't even have broadband in their homes. The U. S. actually lags behind many other industrialized countries in connection speeds, and in the mish mash of connection technologies. I also happen to believe that "cloud" computing is way over sold, as to it's rate of adoption. For those of us who use it, for things like Mobile Me, and storage, it's great, but I suspect that is still a tiny number, compared to how many actually have computers in their home. I also suspect that it will remain that way for some time. One of the most common questions I get when talking to customers about "cloud" computing is, "Are you really sure my data is safe there?" There is still a great deal of apprehension about security of one's data anywhere other than in the building where their computer is located. There is also the issue of "ownership." I love my Mac, and I'm not going to rely on someone else's computer system to be able to instantaneously deliver an app to me, whenever I want it. I may be over generalizing, but Americans like to own physical things, and I want my iWork disc in my home, or a downloaded piece of software on an external drive, as a backup.

[krx]

I sold a domain name a while back to one of these companies - it was an acronym for "Global Load Balancing Software" - and it was while researching the company that I "discovered" the concept of "cloud computing." So I'm no expert on the subject. But I have a different understanding of "the cloud" than this.

Cloud computing it seems to me is any online activity that has its locus of operation on the internet, not on local hardware. Creating a Word doc, writing an email, running a statistical app, and reading my RSS feeds are not forms of cloud computing since the locus of the operation is an application on my HDD. Sending an email or receiving a feed is not cloud computing because the internet is being used as just a conduit for an activity that occurs on local computers. But buying something on Amazon, using Facebook, and writing a note on this doeum are all forms of cloud computing - imho. All these activities require is an OS and a browser since the app or service being used is "in the cloud."

I think it is overly restrictive to think of cloud computing primarily in terms of business or corporate purposes. Data storage on a massive scale is only one aspect of it. Looked at more broadly, in terms of the locus of computing operations/activities for all users, and especially the masses, cloud computing has virtually exploded onto the scene in the last few years. With market forces moving inexorably toward efficiency of operations and maximum market share, it seems certain that cloud computing is not just here to stay but the way of the future.

[nonhuman]

Originally Posted by krx 

I sold a domain name a while back to one of these companies - it was an acronym for "Global Load Balancing Software" - and it was while researching the company that I "discovered" the concept of "cloud computing." So I'm no expert on the subject. But I have a different understanding of "the cloud" than this.

Cloud computing it seems to me is any online activity that has its locus of operation on the internet, not on local hardware. Creating a Word doc, writing an email, running a statistical app, and reading my RSS feeds are not forms of cloud computing since the locus of the operation is an application on my HDD. Sending an email or receiving a feed is not cloud computing because the internet is being used as just a conduit for an activity that occurs on local computers. But buying something on Amazon, using Facebook, and writing a note on this doeum are all forms of cloud computing - imho. All these activities require is an OS and a browser since the app or service being used is "in the cloud."

I think it is overly restrictive to think of cloud computing primarily in terms of business or corporate purposes. Data storage on a massive scale is only one aspect of it. Looked at more broadly, in terms of the locus of computing operations/activities for all users, and especially the masses, cloud computing has virtually exploded onto the scene in the last few years. With market forces moving inexorably toward efficiency of operations and maximum market share, it seems certain that cloud computing is not just here to stay but the way of the future.

Yes, the concept of 'cloud computing' is really a misnomer born of marketing people misunderstanding programmers' labeling conventions. It's been typical when diagramming networks to display the Internet as a cloud, mostly because it would be impractical and ridiculous to try and draw any sort of accurate representation of it (really it's just a massively complex network of an essentially uncountable number of nodes all connected in different ways). So 'cloud' came to refer to the nebulous nature of the Internet.

The vast majority of Internet applications, however, could not in any way reasonably be described as 'cloud computing'. Most of them are no different than local computing, except that the computer doing the work happens to be in a data center somewhere else in the world. Even when you're talking about a cluster of computers, it's not 'cloud computing', it's just outsourcing your work to a different location where someone with more raw hardware than you is willing to do the work for you for a price.

Google's apps, however, start to approach the idea behind 'cloud computing'. Google has built a global network of distributed, parallel, redundant data centers. When working with a Google app it's actually possible that location ceases to be a meaningful as to the computation going on as they are using the computing power of many data centers in many locations, and your data, to some extent, lives on all of them. The only really accurate location you could ascribe to Google's network is Earth: it's basically one planet-wide computer.

[krx]

Originally Posted by nonhuman 

Yes, the concept of 'cloud computing' is really a misnomer born of marketing people misunderstanding programmers' labeling conventions. It's been typical when diagramming networks to display the Internet as a cloud, mostly because it would be impractical and ridiculous to try and draw any sort of accurate representation of it (really it's just a massively complex network of an essentially uncountable number of nodes all connected in different ways). So 'cloud' came to refer to the nebulous nature of the Internet.

The vast majority of Internet applications, however, could not in any way reasonably be described as 'cloud computing'. Most of them are no different than local computing, except that the computer doing the work happens to be in a data center somewhere else in the world. Even when you're talking about a cluster of computers, it's not 'cloud computing', it's just outsourcing your work to a different location where someone with more raw hardware than you is willing to do the work for you for a price.

Google's apps, however, start to approach the idea behind 'cloud computing'. Google has built a global network of distributed, parallel, redundant data centers. When working with a Google app it's actually possible that location ceases to be a meaningful as to the computation going on as they are using the computing power of many data centers in many locations, and your data, to some extent, lives on all of them. The only really accurate location you could ascribe to Google's network is Earth: it's basically one planet-wide computer.

Interesting. So the distinction between "cloud computing" and "the internet" is one of kind and degree? I was thinking of it as "internet-based apps/services." You're suggesting it is this but on a considerably wider/larger scale, involving "a global network of distributed, parallel, redundant data centers" - is that right?

[nonhuman]

Originally Posted by krx 

Interesting. So the distinction between "cloud computing" and "the internet" is one of kind and degree. I was thinking of it as "internet-based apps/services." You're suggesting it is this but on a considerably wider/larger scale, involving "a global network of distributed, parallel, redundant data centers" - is that right?

Basically, yes.

We could simply define it as 'internet apps', but I think that would be missing the point. There is certainly something to be said for your documents living somewhere that makes them easily accessible from any internet-connected location, but 'cloud computing', to me, suggests more than simply outsourcing your data storage and maybe some of your processing to another computer in a data center in Texas. The real advantage of 'cloud computing' comes when we decentralize those capabilities so that maybe there's some servers in a data center in Texas participating, but if that data center is destroyed your data is still out there redundantly stored on perhaps even thousands of other computers all over the world (and beyond?).

Think about the distributed data crunching projects like SETI@home. Now imagine extending that kind of capability to include every single Internet-connected computer on Earth That, I think, is the real promise of 'cloud computing'. Sure it'll still be used for things like hosting web sites, and collaborating on word processing documents, but the real power of it comes from a more complete and efficient leveraging of our planet's computing resources.