Ina Fried says that Microsoft’s much rumored OS in the cloud is still “pie in the sky“.

So far, she is obviously right. The current iteration of Window’s Live is not much to write home about. Nor is the current offering of Google Apps enough to be considered an OS in the cloud. The Google Apps enterprise (AKA gmail, but with your unique domain) is great application, but Google Docs and Spreadsheets are weak offerings that need significant improvement.

No other major vendor is trying to offer a single destination “OS in the cloud”. However, there is plenty of recent evidence that it will be here soon.

What will users need from a Cloud OS?

They will need more than just web based copies of their current desktop tools. A good way to understand this is to compare a blog to a word processor. They are both content creation tools, but writing information within a blog can lead to network effects. As more people comment on your article, it gains more value. As more people cross-link to your article, it gains value and gets a higher ranking on search results. The same is not true of content locked within a doc file.

For the enterprise, an OS combined with applications become the vehicle for doing the following

1. Creating content ( words, pictures, presentations, video )
2. Analyzing data ( Excel )
3. Communicating with colleagues ( Email, IM )
4. Storing information
5. Finding information
6. Running domain specific applications ( Expense tracker, CRM, ERP, IDE )

As an avid fan of “Enterprise 2.0″, I still do not use Web tools to do all these things:

1. Creating content - I use Keynote, Pages, iMovie, Inkspace and GIMP to create content. The only web based content creation tool I use is WordPress.
2. Analyzing data - I still use Excel
3. Communicating with colleagues ( Email, IM ) - I’m 50% Web based here. Gmail for email. iChat and Skype when iChat doesn’t work.
4. Storing information - Also 50% here. Obviously my mac harddrives, some jump drives and then Joyent’s amazing BingoDisk.
5. Finding information - About 75% web based here. In order, Google, OS/X’s Finder, Wikipedia and Ask.com
6. Running domain specific applications ( Expense tracker, CRM, ERP, IDE ) - 50/50 here - Eclipse and Mantis

Cloud OS network effects will change how you work

Recently, Peter Rip said something like this to me:

Businesses think that their employees use IT systems to get information, but actually, it is more accurate to say that businesses use their employees to move information between their IT systems.

The point is that knowledge workers spend a huge amount of time not creating or analyzing, but instead on processing. Examples include:

• Sending emails that ask who has the latest copy of something
• Trying to find out who has the answer to a specific question
• Manually extracting reports from one system, only to have you, or someone else upload that report to a second system.

Cloud OS based productivity tools hold out the promise of improved process. Work is not a pile of concrete deliverables. Work is not collection of files. “Work” is a verb. It is a dynamic series of actions.

Cloud OS based productivity tools are particularly well suited to anything that is dynamic and on going. Facebook or MySpace are both popular not because of the collection of content amassed on those sites, but instead because their server software and their community support and on-going dynamic conversation.

In business, there is the same emphasis on the next activity, just like the next comment or the next post. It doesn’t really matter what you did yesterday. What matters are the sales the company is going to make today and into the future. Even the knowledge you acquired yesterday is only important in the limited specific set of ways that it can be used to help you make more profit in the future.

Clearly a blog, when compared to a static word doc, delivers on some elements of an improved process. A blog supports an on going conversation through comments and link backs from other blog entires.

What’s the proof that we are quickly moving towards a Cloud OS?

There are a bunch of technical things that need to be in place before people can do all the work they would want to do exclusively through a browser. By work, I mean the six points mentioned above, such as creating content. And to move from working in a Desktop OS to working in a Cloud OS, people will have to clearly feely that their routine daily processes will become significantly more efficient if they move to the Cloud.

Having said this, it is clear that the necessary technical prerequisites are starting to show up.

Open APIs First, more and more services are realizing that they have to open up an API that can be leveraged within composite applications. If part of your job is to pull information out of system A, sort it and then shove it into system B, then, it can be argued that you server the machine, rather than the other way round. The best way to fix this is with a composite application that automates as much of your work as possible. However in order to move information from A to B, both systems must have an API.

Recently, a whole raft of powerful services have suddenly realized that they need to open up an external API:

• LinkedIn has recently said it will open up an API
• Facebook users have add 65 million apps in first month of Facebook’s new platform, as Facebook continues to open up.
• Plaxo new 3.0 system has also started to focus on combining information from a broad range of services, while Plaxo also opened up their synchronization API.

Business Models the Support Integration Each of these companies now understands that their networks are both too big AND too dynamic to be threatened. If you build a crawler to download every single bit of data within LinkedIn, you still couldn’t replace. What counts for LinkedIn users is the dynamic ongoing interactions and conversations hosted on the LinkedIn service. To “steal” business away from LinkedIn a competitor would have to not only get all the data, but also convince all the users to switch to the new site overnight. That isn’t going to happen.

In addition, and critically, these companies have worked how to make money while opening up their APIs. Here, Google Maps is a terrible example. How is Google making money by giving away access to their maps? In the future, of course, Google might figure out a way. However smaller companies can not afford to take that approach. They need to charge now.

Networks become more valuable the more nodes they have. As tools to support business, any service becomes more valuable as it is hooked into more business processes. LinkedIn becomes more useful to me if I can use it as my address book within my phone, within my email client and within any other business process where I need to leverage my network of personal contacts.

How do these networks monetize their open APIs? In the case of LinkedIn and Plaxo, they charge for premium services. Facebook charges for integrated advertising. In all cases, the open API drive more users which, in turn, drive more revenue.

The Personal Server

The personal server is actually not one dedicated machine. Instead it is a collection of services that end users select, run, customize and connect over time. Get a blog at wordpress.com, get a page at Facebook, another at LinkedIn, etc.

The difference between Web 2.0 / Enterprise 2.0 and the Enterprise 1.0 team collaboration solutions is self provisioning and the realistic promise of process integration.

That process integration will be delivered by new types of web services that take advantage of the open APIs and are capable of delivering useful tools to support process integration. My company, Teqlo, focuses on exactly that. We have been heads down in development mode for the last few months, but in the near future, we will be releasing some useful tools that help at least one targeted by of user with process integration.

When will the Cloud OS Arrive?

It is not here yet, but when it arrives, I believe that it will not be delivered by one single vendor, such as Microsoft. Instead, it will be a constantly evolving heterogeneous collection of services with open APIs and tied together by integration services.