Thin Client-Thick Server: We have deployed the TC-TS at one of our other offices, which has 6 users. The usage pattern is quite different, it being an administrative office- there is more use of OpenOffice and Tally (a popular Indian accounting package). We did manage to get Tally returning on Linux using Wine, but are having problems with printing. We did have a few teething problems – needed to upgrade the LAN to 100 Mbps and the server. Within our office, we are moving users to a dual CPU system. There is still the occasional crash, and we need to get to the bottom of why it happens. Have also done a few demos, but still no definitive beta customer. Need to make the TC-TS into a CD-installable product. Also, need to try out different Windows applications on Wine (especially Foxpro, VB-apps, Microsoft SQL). In our R&D effort, need to do (a) a graphical booting process for the TC (b) Local Apps on the TC to reduce the bandwidth (c) understand the X protocols better to see where we can optimise network traffic (d) Wireless LAN support.

Digital Dashboard: We deployed the weblogs platform internally for about 15 people in the company. People have begun blogging, but most posts are either to the private or the group blog. There is something missing in the flow here, and we need to make it easier for people to do it, and create the collaborative space. We’ve also studied Traction and Scopeware. Traction has some interesting concepts. Our next step here is put together a DD in a browser which encompasses an aggregation for much of the user’s personal information (links to recent mails, IM buddies, frequently accessed files, bookmarks, calendar, contacts, etc.) – the idea being that the respective application would be launched from the DD. The user gets a snapshot of his “home directory” and personal information in one page. This becomes the first layer, adding to the other layer of enterprise blogs which we’ve already done. The part which needs thinking is how to fit this together into a scalable architecture, like that of the Information Refinery. We’ve also spent some time studying XSLT and the XML format of OpenOffice Calc. The idea of the latter work is to see if we can use OO Calc as a computation engine without actually invoking OO.

Enterprise Software: We’ve been working to put together a model and a plan (and a team). The Digital Dashboard (personal and executive, which would comprise enterprise events) is the first step. Next is to put together a business process model for the basic applications. EAI and B2BAI come in later. More practically, we need to continue to work on building the Client Information System which integrates information from accounting (Tally), marketing (spreadsheet) and support (MySQL database). We’ve also put together a working platform comprising Rational Rose, ArcStyler, Weblogic and JBoss. It is a mix of commercial and open source tools – the former give us an idea of the best tools, while the latter is what we need to actually use. One idea I need to think through is that of an “Enterprise Emulator” – like a SimCity which simulates an enterprise ecosystem. This would be useful for us as a testbed for the software we develop.

MailServ: We have decided to begin work on version 4.0 of MailServ, so as to get it ready by the end of the year. Need to put together the feature list of what we want. They key features our current 3.2 version has are: Mail Server, Proxy Server, IM (through Jabber), basic Firewall, Anti-Virus software, Global Address Book (via LDAP). Some initial ideas: enhance the Proxy Server capabilities, add Collaboration (Calendar, Contacts, Group Scheduling) such that the backend storage is in XML and can later be integrated into the Digital Dashboard, Project Management, stronger Firewall capabilities and perhaps, VPN support.

BlogStreet: We are at about 10,000 blogs. Need to think if we want to deploy additional resources on this. Ideally, I’d like to see this work with blogs within the enterprise, but we are not there yet.

My Blog: We need to do some clean-up on some mess-up which we have done with the database (result is that some posts now have incorrect categories and the search results show duplicates). Going ahead, need to get Related Stories for each post that I do, thus creating a chain of thoughts and stories. There are a few other ideas also: for example, it would be nice to add a “blog” page consisting of the headlines for the major publications; make Search more effective – I use it quite a lot when writing my Tech Talks and am unhappy with the limited control htdig gives me (I at least need to get relevant posts in reverse chronological order – today, it seems quite random!). More broadly, how can I do “value-added aggregation” to build up a niche in the blogosphere. There needs to be something innovative which I should think of. Overall, am quite happy with the posts, etc. I have now blogged for pretty much all of the past 100 days with an average of 5-6 posts daily. Now, to think of how we can raise the standard of Emergic.org. Perhaps, we need to add a discussion forum like Slashdot for discussing topics relevant to technology in emerging markets. There’s still too much of just me talking, though of late the comments are increasing.

In an InfoWorld interview, Microsoft's Jeff Raikes outlines the future role of Office. Raikes is group vice president of Productivity and Business Services. He says:

If you step back and look at the Windows server, what's valuable from a knowledge worker or information worker standpoint? Historically, the Windows server, or previously the Novell server, was all about file and print services. It was about how to share resources for the knowledge worker, for the information worker. What SharePoint Team Services does is it builds on that concept, taking it to the next logical step.

So what we're doing with the .Net servers is we're going to be integrating the SharePoint Team Services capability as a foundation for this kind of collaborative routine work. For example, with Office 11 look at what it can do for things like meetings. Meetings today are relatively underserved in terms of software technology. I mean there [are] a lot of things you need to do in terms of preparing for a meeting -- how you set the agenda, how you share the documents, and all of that. Now you can make that fundamentally a part of the Office client world. Today, take [for example] a document where you have a set of people collaborating. It's reasonably effective to send it around in e-mail; that's common for people to do. But especially in an environment like InfoWorld, I would think it would be nice if you just posted the documents that you're working on -- at least [a document you're] collaboratively working on -- right there on the SharePoint Team Services site.

He talks quite a lot in the interview about SharePoint Team Services and Groove (in the context of collaboration) - maybe we need to take a closer look at both in more detail.

The Free-Trade Fix is a New York Times magazine story on globalisation:

Globalization is meant to signify integration and unity -- yet it has proved, in its way, to be no less polarizing than the cold-war divisions it has supplanted. The lines between globalization's supporters and its critics run not only between countries but also through them, as people struggle to come to terms with the defining economic force shaping the planet today. The two sides in the discussion -- a shouting match, really -- describe what seem to be two completely different forces. Is the globe being knit together by the Nikes and Microsofts and Citigroups in a dynamic new system that will eventually lift the have-nots of the world up from medieval misery? Or are ordinary people now victims of ruthless corporate domination, as the Nikes and Microsofts and Citigroups roll over the poor in nation after nation in search of new profits?

Tina Rosenberg frames the trade debate nicely:

The architects of globalization are right that international economic integration is not only good for the poor; it is essential. To embrace self-sufficiency or to deride growth, as some protesters do, is to glamorize poverty. No nation has ever developed over the long term without trade. East Asia is the most recent example. Since the mid-1970's, Japan, Korea, Taiwan, China and their neighbors have lifted 300 million people out of poverty, chiefly through trade.

But the protesters are also right -- no nation has ever developed over the long term under the rules being imposed today on third-world countries by the institutions controlling globalization. The United States, Germany, France and Japan all became wealthy and powerful nations behind the barriers of protectionism. East Asia built its export industry by protecting its markets and banks from foreign competition and requiring investors to buy local products and build local know-how. These are all practices discouraged or made illegal by the rules of trade today.

ZDNet's story on Sun has an interesting comment on the desktop:

Displacing Windows on the desktop, however, will take more than appealing to the bean counters. Many companies have Office macros and applications that are not easily accommodated by an open source alternative.

I can speculate that the alternative desktop will center around Java, Mozilla, and Linux, with OpenOffice/StarOffice as an alternative to Microsoft Office and Linux (or even Solaris) as the operating system, some user interface conventions, and industry standard hardware.

The basic elements of an alternative desktop exist. The challenge is in creating an integrated platform that gains the kind of community development support and governance that Linux has garnered.

There is another story on the Linux desktop potential in ZDnet. It mentions about Verizon saving USD 6 million in system costs but that also resulted in additional software conversion costs. On MS-Office and Star/OpenOffice, a Microsoft director is quoted as saying: "StarOffice or its open-source sibling OpenOffice are 'good enough' for basic tasks but are harder to use than Microsoft Office. Microsoft's studies of the 11 most frequently used operations in Microsoft Office took on average 2.5 less time than in StarOffice." [Would be good to get to take a look at the study - which are these 11 operations?]

The Real Opportunity for the Linux Desktop

One has to look at two distinctive markets for a Linux Desktop: one is for existing Windows users (primarily in the world's developed markets) and one for the new users (in the emerging markets). The first segment is going to be hard to target, and that is what everyone (including Sun and Red Hat) seem to be after. That is a mistake. It is not easy to switch people using Windows, especially when the financial incentive is small compared to the total salary of a person.

The second segment is the one to target: catch the users in the cradle, as it were. To do this, the Linux Desktop should additionally have the following (when targeted at enterprises)
- a Thin Client, to bring down the cost of hardware and move computing to the server, thus simplifying desktop administration
- a Digital Dashboad, to aggregate all information so that becomes the "virtual desktop", making it easier for new users (who may even be first-time computer users) to connect to the other applications

There is another segment - a stand-alone Linux Thick Desktop, which can be used to bring down the software cost. This is useful (and the thing that Sun and Red Hat seem to be focused on), but I think it will not succeed because it only gives one advantage (software cost reduction) compared to three benefits of the other approach based on a Thin Client (hardware cost reduction, software cost reduction, and ease of manageability).

Windows is still much easier and friendlier to use than Linux on the desktop. Rather than "disappointing" users we need to "delight" users. This is where Linux needs to look at new markets for the desktop - the next millions of users for whom the first taste of computing could be Linux.

USA Today wrote about Linux's increasing popularity a couple weeks ago (story has some good examples of companies switching to Linux):

In recent months, companies and government agencies, from Wall Street to Hollywood and from Europe to Asia, have embraced Linux, adding to its legitimacy. Corporate backers of Linux, which is free, multiply as they stake out territory around the fastest-maturing operating system ever.

Already popular as an inexpensive driver of Web pages, e-mail systems and computer servers on the edges of corporate networks, Linux has begun seeping into corporate data centers where serious computing takes place.

The McKinsey Quarterly writes on the software business (the article is focused on turning around software companies):

Software is a winner-takes-all business in which three factors—the need for compatible technology in networked environments, high switching costs, and increasing returns to scale—unite to ensure that only a small number of players in most market segments survive in the long run. SAP, the German provider of enterprise-resource-planning (ERP) software, illustrates the phenomenon: by the mid-1990s, it was such a clear leader in its market segment that it was able to claim, "We spend more on R&D than our competitors have revenues." This winner-takes-all phenomenon favors companies as they race toward market leadership but exacerbates the difficulties of companies that fall on hard times. Switching costs, for example, are high because software is expensive to integrate into corporate IT systems; this hurts the losers because once a network has been configured around new software and employees have been trained to use it, the company isn’t likely to switch to a vendor that might not be around to provide upgrades and service in years to come.

The article talks about the need to come up with new category killers, which are "products that will claim a market share of at least 30 percent."

Information’s flow is, like that of time, inexorable and continuous. Each day, we generate more information now than what would be generated in a lifetime a few years ago. This is truly an “Information” Technology revolution that we are going through. For many of us, in our roles as knowledge workers, much of our day is spent processing and creating information.

Information abounds in its pervasiveness all around us: newspapers, magazines, television, radio, hoardings, emails, documents and websites envelope us. The problem and challenge is how to best manage the information that is there around us. Kevin Werbach puts it in context: “The issue is how to manage information without managing it. We want the right information to get to the right person at the right time, but there's usually no way to know those things ahead of time. We also want to leverage the Web, which is centralized and has links that can break and go in only one direction, while engaging in bidirectional freeform interactions like Weblogging.” According to Werbach, the next great business software application, alongside the word processor, spreadsheet, browser, and email/calendar/address book, will be the Information Router.

My belief is that the combination of weblogs and RSS is what can dramatically amplify our ability to process information. Weblogs are personal journals, which have links, quotes and comments – either by one individual or a community. RSS stands for Rich Site Summary, an XML format for syndicating information from a website or a web page. Blogs can work as an extension of our memory, while RSS can pull information from various other sites (or even entities like sensors). They form the foundation of what I call the “Information Refinery”, which is akin to the Information Router mentioned by Werbach.
Much of what we do as information/knowledge workers is process information. We get it from multiple resources, we assimilate it, we route it to others, we translate it into different formats. When we look at our lives today, we have information scattered all over the place: multiple email accounts each with a multitude of folders, files on our hard disk and on the server, calendaring software (like Outlook or Evolution), our address book, the documents we browse, the searches that we do on Google, the business cards that we accumulate (each with its own story), hard-copies of notes which we may make in our paper-based notebook. In the enterprise, we can think of business processes as rules for routing information, as tracing the flow of events through various “filters” and “actions”.

In essence, we have information “ores” which need to be refined. We rely on our memory to manage much of the information. Some of us use PDAs. In recent times, Google has become our second memory – for information which is available out on the Internet. But so far, there has not been an easy mechanism to manage much of the other information that we use individually or share with others in a group. This is where weblogs and RSS can come in. This is at the heart of the Information Refinery. Whether it is news items or blog posts or enterprise events, the information refinery should be handle to handle all of them. Taken together, they can create a unified information portal (a “digital dashboard”, or more like a “digital smorgasbord”) – a single screen as a window to the information world.

Tomorrow: Blogs and RSS (continued)