Thursday, February 21, 2008

Microsoft on open APIs: New tune or blowing more smoke?

It's a radical departure, this news from Microsoft that openness between its products and the rest of the universe is more than a hollow platitude. To take Microsoft at its Word, given this release, is to open an era of an entirely new Microsoft. But is it?

For Microsoft's admitting today that it needs to be more open, should meaningfully support standards, should allow developers access to APIs, should make its documents easy to share, should build new interoperability-friendly APIs for its major products -- this is also admitting that it has tried to thwart or undo these very avenues to progress in the past.

This has to be more then trying to appease the European regulators.

So Microsoft's declaration to embrace openness is at the same time a mea culpa, that it has been employing dirty tricks to seal its products off, and has tried to punish those that would seek to make productivity from the interchange between Microsoft's products and the rest of the universe. And it has been doing it all as a convicted monopolist.

And so we are now supposed to move on. Please see this as a new era of openness, we were wrong (while the price was right), and now we're going to play nice. We love open source, and standards, and API access, and third-party developers. Sorry, so sorry, let's move forward.

So they say.

And to pair this new air of openness with the acid hostile takeover bid for Yahoo offers a few potential insights, if not underscores deep internal conflicts in Redmond.

It is odd that the road to openness demands the takeover of a foe, nee friend. One week we need to outbid the world to grab a company distinctly different from our own (for the engineers), and the next week we need to better love the sharing between people and technology (for the engineers).

What? These market movers strike me as at crosscurrents, at best, or more likely some kind of bi-polar method to software development and deployment amid an advertising-crazed upset of the monopoly apple cart.

And yet, too, the odd pairing makes sense, if you're a cynic.

Because by seeking to buy Yahoo, Microsoft is also making some major mea culpas. By trying to buy Yahoo, Microsoft is in effect agreeing that open source is essential to any competitive, world-class datacenter or cloud fabric of the future. And so then Microsoft needs to change its tune about open source. Sorry, we goofed. Now we'll just buy an open source cloud infrastructure.

If by seeking to buy Yahoo, Microsoft aims to broaden its reach as a purveyor of IT functions as services (beyond the Live stuff to date), then openness and interoperability between its products and Yahoo's services is a must. And again, Microsoft needs to change its tune about openness and interoperability. Sorry, we were wrong about the Web. It's not going away. Our interfaces won't necessarily be the gateway to the Internet assets you need and seek.

If by seeking to buy Yahoo, Microsoft aims to use industry standards to integrate its Windows Everywhere arsenal with the rest of Yahoo -- and most enterprises or portals, for that matter -- then Microsoft needs to change its tune about is depth of use and commitment to open industry standards. Sorry, we should have played nice all along. Who wants proprietary formats, anyway?

For Microsoft to make the Yahoo merger work, the announcements today have to work too. By buying and integrating Yahoo into Microsoft, in essence, forces Microsoft to be integrated with a much larger slice of the real world. Mea culpa, mea culpa, mea culpa.

But what happens after the Yahoo merger, assuming they integrate it all (or at least make it interoperable, presumably leveraging the very standards they so long disdained)? What's to say that everything that Microsoft is pledging to do today, they decide not to do in two years?

Is there no going back? Or is this the means to the ever-larger platform play that Microsoft could not attain -- of extending one monopoly to another? Is this the Web-as-platform acquisition that leads only to increased advertising revenues at the cost of the demise of the original monopoly and those juicy license fees?

I do like what I'm hearing, I just wish it was easier to believe that Microsoft by its very nature has changed along with the need for it to change.

It seems easy for Redmond to cop these mea culpas. It's not as easy for me to forget what made them necessary.

Informatica offers Data Migration Suite; Wipro to leverage its power

Informatica Corp. is addressing the crucial issue of data access by packaging its data integration software as Informatica Data Migration Suite, with an eye toward providing a solution for companies involved in mergers and acquisitions, consolidation, and outsourcing.

With the shifting landscape of business today, as well as such innovations as service-oriented architecture (SOA), reliable data access is more important than ever to provide the organizational flexibility necessary to respond to internal demands and external pressures.

Despite this, most Global 2000 companies report that their data migration has come in late or over budget. According to a study by Bloor Research, commissioned by Informatica, some 84 percent of projects fell short of expectation and resulted in cost overruns averaging 30 percent. This can have a serious impact on an area expected to see budgets exceeding $8 billion in the next few years.

Informatica, Redwood City, Calif., says its new suite, which provides an independent software platform designed for data migration, will include the company's PowerExhange, PowerCenter, Data Explorer, and Data Quality. The company says the new offerings is designed to ensure the success of data migration projects.

One company has already jumped on the new platform, with the announcement that Wipro Technologies in Mountain View, Calif., will use it to provide data migration to its customers worldwide. The Informatica offering will underpin Wipro's Data Migration Shared Services.

Wipro, which provides integrated business, technology, and process solutions on a global basis, will use the Data Migration Suite to automate and streamline the process. This includes creating all data mappings required to migrate data from and to any system, discovering data quality issues at their source, and cleansing and converting the data as part of the overall migration.

The primary targets for the offering will be data migrations in support of new applications including ERP /CRM implementations and upgrades, legacy system modernization, application instance consolidation, mergers and acquisitions, and outsourcing.

Tuesday, February 19, 2008

Bungee Connect beta goes public, adds oomph to development and deployment as a service

Bungee Labs, Orem, Utah, continues its march toward "platform as a service" (PaaS) with today's announcement that it has opened Bungee Connect as a public beta, inviting all developers to, in Bungee's words, "get inspired, get started, and get involved."

Bungee Connect is an end-to-end environment that allows developers to build desktop-like applications from multiple Web services and databases and then instantly deploy them on Bungee's multi-tenant grid infrastructure.

Fellow ZDNet blogger Ryan Stewart has a god piece on Bungee's coming out. As does Dan Farber.

Bungee is dangling the carrot of a reduced time to market -- as much as 80 percent -- from testing, deploying, and hosting in a single, on-demand platform. With utility-like pricing, Bungee is also offering reduced cost. The company claims that businesses can expect to pay $2-5 per user per month for a heavily used business productivity application. However, all applications will be hosted for free during the public beta.

To provide a reference application, Bungee is also releasing WideLens, a calendar application that solves integration problems between Microsoft Exchange,, Google Calendar, Facebook, MySQL, and iCalendar.

WideLens is intended to provide examples of integrating multiple databases and Web services, end-user authentication, and dynamic user interface presentation patterns. The integration accommodates a variety of protocols -- WebDav, gData, SOAP , REST, and the MySQL client libraries.

A demonstration video is available from the Bungee Web site.

One major feature of Bungee Connect is that developers can access all applications through one of the major Web browsers -- Internet Explorer, Firefox, and Safari -- with no software downloads, installs, or plug-ins.

I've been following the exploits of Bungee since it first unveiled Bungee Connect last April at the Web 2.0 Expo. Back then, I saw it's potential to expand the universe of Web services for developers:

" . . . the real innovation is how the Bungee Connect model provides an incubator, and — in essence — a business development partner to the developer so that they do not just create an application — they can create a business. That's because the cost for the use of the tools, testing, and then hosting is free, and the subscription cost for the at-scale hosting only kicks in based on the use of the application by end users. Low use means low costs, and high use means a predictable measure of the proceeds goes to the development and hosting service."

One thing we can take away from this announcement is that PaaS is now more than just pie-in-the-sky concept or a one-off product. It's gaining traction, and is offering companies a low-cost -- or scalable cost -- route to business development.

David Mitchell, Bungee founder and CEO, sees PaaS as the wave of the future. He said in a blog post on the Bungee site:

"In our view a platform includes all the systems and environments comprising the end-to-end life cycle of developing, testing, deploying, and hosting Web applications. Naturally, this platform must also be cloud based, a platform as a service,

At Bungee Labs, we believe a transformation larger than SaaS is emerging, where end-to-end development, deployment and hosting platforms as provisioned as services over Web."

Mitchell's blog post echoes what I said in an update last July:

The net effect of these trends and examples is that the time, cost and risk of going from design to full production are deeply compressed. We are entering a period on unmatched applications, services, and media creativity.

I expect we'll be seeing more application lifecycle as a service approaches -- ones that bring enterprise-calibre development, test, deployment, and hosting together. Amazon is moving in this direction. We should also expect the same from Google and Microsoft.

Indeed any cloud computing environ that wishes to support a vibrant community ought to offer what Bungee Labs is providing now. And do hope that standards rule the day so that mashups remain straightforward.

The issue of code portability also needs to be addressed. I probably won;t make sense for these hosts to make leaving easy, but it should nonetheless be quite possible and well understood.