IT Industry Analysis

Professors David Hsu and Kevin Werbach of the Wharton School of the University of Pennsylvania have a good analysis of Google's strategy behind Chrome, available as podcast. This analysis goes further than the typical analysis such as "the browser war is back" and "Chrome is a direct assault on Microsoft's Internet Explorer", which you could read in the hastily written IT press articles ...

This podcast can be found at http://knowledge.wharton.upenn.edu/article.cfm?articleid=2045. There is also a transcript version available.

There are a lot of comments on Microsoft's launch of Live Search Cashback. For those of you who have missed this news, Microsoft will offer real hard cash back to users of this Live Search feature - at least if you buy products during that session.

TechCrunch has a very good description of this new feature, but we were particularly pleased with their analysis of what this means to the search engine market and to Google, and what their first user experiences were.

Of course, this move would have a huge impact on the search market if Google would not respond, but don't expect Google to sit back and relax ...

There have been rumors about Google's grand entry into the mobile market for months. Some were expecting a Google-branded mobile phone, the gPhone, but it turned out to be a open source-based mobile application platform that should be available on various mobile (smart)phones from e.g. Motorola, Samsung, HTC and LG by mid-2008. Actually, Google will not just deliver the platform called Android but it will also its application suite (Google Maps, search, Gmail, calendar, chat, ...) to that mobile platform to conquer the mobile market. Because Android is an acquisition that Google made in 2005, it is interesting to look at the acquisitions that Google made since then but also the patents that Google has submitted. On the acquisition front, take a look at what Google can do with Jaiku, the Twitter-like microblogging and social networking tool, or the mobile social network Zingku. The acquisition of Grand Central has gone a little bit unnoticed by the analysts, but may turn out to be extremely important for Google's Gmail and Gtalk success). On the patent side, the text message micro-payment solution could turn out to be the most interesting of all.

While the Android platform announcement was a bit of an anti-climax for some analysts who had expected to see a real mobile device running Google's killer applications, we think this is a very significant move in Google's Web domination campaign. However, a few questions and issues remain unanswered:

• Given that the Open Handset Alliance relies heavily on Linux and Java, why was Sun not a part of the Android announcement ? Could this have something to do with the Apache license that Google uses for its open source platform ?


• Given that this will be Linux-based, how will this influence the battle between the Microsoft's Windows Mobile, Symbian, Apple IPhone and Mobile Linux platforms ? Om Malik has an interesting discussion about this, including a confidential smartphone market share slide by Symbian. Will Android impose a landslide in market share for Linux ? Will open source developers start developing a compatibility layer for Google's apps for Windows Mobile, Symbian and IPhone based on what they see in the Android source ? Why wouldn't these clever open source developers produce a version for high-definition television sets, a market that Google definitely not wants to miss, but is pretty quiet about.


• Will the strong bond between Google's applications and the Android platform - even though it is open source - not give rise to antitrust discussions about monopolies, consumer choice, ... ?

Will definitely be continued ...

During our seminars on service-oriented architectures (again scheduled in 2008 because of their huge success), there was a lot of discussion about how this SOA solution will interact with its users. In other words, will the trend towards flexible architectures also influence the way we deliver applications in the future ?

The answer is probably yes: the services of the SOA will be combined with other services from external partners and external data sources to deliver the information that the user needs. These so-called mashups are still highly experimental and overhyped, but there is a big chance that reliable, robust and secure "enterprise mashup services" will be the buzzword of 2008.

Anyone who wants to track the success of mashups should take a look at ProgrammableWeb, where you can find out what's new and interesting with mashups, new applications, new APIs, and about the Web 2.0 as a Platform. For the moment, there are not a lot of very useful enterprise mashups, even in the category "Enterprise Mashups", among the more than 2500 mashups that are listed there.

When you try out some of the mashups at ProgrammableWeb, you will see some of the problems and challenges facing mashups:

• Reliability and robustness: when a service is unavailable, every mashup that relies on this service, gets broken - when we browsed, I think 10% of the mashups had this problem


• Data quality and governance: the quality of a mashup is only as good as the quality of the services it uses - and some are really bad


• Security and privacy: although there are no real cases of public mashups that use private, secure services known to us, it will be a big challenge to implement single sign-on and federation in such a loosely coupled environment.


• Dion Hinchcliffe has many more challenges in his Enterprise 2.0 weblog entry.

I.T. Works will continue to track this evolution in application delivery, and we will probably organise a seminar about "enterprise mashups" in 2008. Particularly the "how to build an industry-strength mashup", the impact that such tools may have on business analysis, and the use of tools such as lower-level Microsoft Popfly or Yahoo Pipes, or higher-level Serena Business Mashup Composer or Openkapow's Robomaker would be very useful parts of such a seminar.

Hence, we are interested to hear from your experiences with mashup applications, API's, tools, etc. Just email us at mashupseminars@itworks.be.

At their IT Forum in Edinburgh, Forrester CEO George Colony told lots of European chief information officers, IT directors and senior IT managers that the name "IT" should be changed to "business technology", or "BT", as technology is now vital for running a business. Colony asked chief executives to do a "scream test" in their company: seeing how long it takes before employees begin to scream when you pull the technology plug. Not because they can no longer access "Second Life" or their personal email, but because the company is loosing money...

Colony expects a transition from information technology to business technology, and advises CEOs and corporate boards to learn more about this. Of course, they should not really understand what SOA, CRM, ERP, ... is, but how this "business technology" can generate revenue for the business.

Not so long ago, HP came up with similar conclusions about this paradigm shift, probably inspired by its acquisition of the excellent products of Mercury and Systinet.

More info about Forrester's IT Forum at Computerworld UK. Colony wrote his ideas down last August in this Forrester report.

Linspire, the company Microsoft sued a few years ago for using the "Lindows" name, has joined Novell/Suse and Xandros in Microsoft's "IP Protection" clause, (according to Microsoft) saving its customers from potential Microsoft patent-infringement lawsuits.

Sources: Mary Joe Foley adds some details about the agreement

As long as Microsoft refuses to show proof for its claim that Linux and open-source software violates 235 of Microsoft’s patents, nobody really believes their statement. On the other hand, the more Linux distro vendors join the "IP Protection" club, the more fear, uncertainty and doubt (FUD) Microsoft is creating that there really is something going on. Will Red Hat be the next company to seek "IP protection" or is Microsoft trying to align all Linux players against Red Hat ?

Meanwhile, Microsoft recruited Tom Hanrahan, previously director of engineering at the Linux Foundation. He will become director of Linux Interoperability, and will make sure Windows and SuSE Linux work well in virtualization, directory and identity integration (Active Directory and eDirectory), and on systems management. Is this the foreplay to the acquisition of Novell ?

Microsoft has just announced that it will offer an entry-level version of XP and Office, as well as Windows Live Mail and Math 3.0, for only 3 USD. Microsoft wants to make deals with governments to provide hardware with this software to students.

Officially, Microsoft wants to get itself an extra 1 billion users by 2015, and wants to tackle piracy in developing countries. For example, why would students in China buy an illegal copy for 1 USD when they can buy a legitimate copy for 3 USD ?

Here is more information:

• New York Times article
• An interview with the relevant Microsoft VP Will Poole
• Negroponte's One Laptop Per Child (OLPC) Project, which may come under pressure by the Windows/Office bundle office. OLPC is a 150 USD PC based on Linux

WebMethods, one of the last pure play SOA/BPM vendors just got acquired by Software AG.

No surprise, I was talking with Rick van der Lans at our seminar "Van Procesmodel tot Dynamische Applicatie" about who would buy webMethods and Tibco. These companies have great SOA/BPM/BAM stuff that many others (including IBM, BEA, Oracle, Sun, SAP, Software AG, HP, ...) are lacking or are not very strong in... By the way, the seminar that we organised at that moment emphasized on the intersection of SOA and BPM, exactly what is demonstrated by the acquisition of WebMethods by Software AG.

Beth Gold-Bernstein of ebizQ has a great analysis about this acquisition, the implication on Software AG's deal with Fujitsu, the strengths and weaknesses of the Software AG/WebMethods combination.

Personally, I think this acquisition increases the chances that Tibco gets acquired sooner rather than later by one of the big SOA/BPM guns. Furthermore, it makes Software AG as an acquisition target all the more interesting.

Years ago, Microsoft tried to be the guardian of our personal data with Microsoft Passport.   Passport was conceived as a single sign-on for  the internet.  You were supposed to  use Passport to log in into your bank's website, Ebay, the social networks you were member of...  The intention was noble-minded in that respect that it allowed users to divulge their personal data selectively depending on the relationship with the diverse site owners. 

But it didn't work, for reasons that Microsoft since then has acknowledged (see the "Seven Laws of Identity" by MS's digital identity guru Kim Cameron).  Users did not grasp what a technology company had to do with the relationship with their banker or auction site (users probably even didn't like the idea of one centralised identity).

Windows Cardspace is built in into Vista and the latest version of the .NET framework.   Unlike Passport,  it  is an enabler instead of an actor.  It's a framework that provides a consistent interface for users to manage their multiple identities.    On an auction site, you might want to have your bank assert that you can pay for the item you bid on.  When commenting on a blogwith your own blog url, you want to have your blog software assert that you, the  commenter, are really  the person behind the url you pretend to represent.  All of this becomes possible with the single, consistent user interface of  Windows Cardspace, in which all kinds of identity providers can plug in. 

Cardspace is an open specification (Windows Cardspace, previously called "InfoCard",  being the MS implementation).  Several parties are working on their own implementation, see for example this (cross-platform)  Firefox plugin.  It's a perfect illustration of Microsoft as "the company that helps you to organize your own information".  As opposed to "the company that wants to organise all the world's information".  Guess who that is...

Relevant links:

• "What is Microsoft Infocard": an excellent introductory write-up
• Cardspace : MS resource pages
• Kim Cameron's Identity weblog

Several months after the famous Microsoft-Novell cooperation announcement, the first positive signs of the agreement are popping up:

• Novell has adapted its award-wining Zenworks desktop management suite to run natively on Active Directory. It is remarkable that Novell plays down its own metadirectory software such as Novell Identity Manager, but it is certainly a good thing small and medium-sized businesses.
• Novell has developed an open source version of Microsoft's InfoCard identity technology that will run on both Linux and Mac. Actually, Novell was already working on this long before the Microsoft deal, so the source code is available and was presented at Novell's Brainshare.
  A more technical description of the InfoCard selector for OS X and Linux based on the Brainshare presentation can be found at the excellent "Between the Lines" blog.

Of course, not everybody is happy with the Microsoft-Novell agreement. Open source guru Bruno Perens wrote an extensive blog entry "What They Aren't Telling You" about the deal. Also, Richard Stallman (who will present on the goals and philosophy of the Free Software Movement at our local University next week), lashed out at the deal in his presentation made at the fifth international GPLv3 conference in Tokyo.