The Experience Mashup: Mash-up the Desktop with the Web

Mash-up is a word that one hears frequently these days. Definitions of the word are still evolving - there are data mash-ups, consumer mash-ups and business mash-ups. that are beginning to be understood by a small population of developers and vendors.

I think that we are at the beginning of the evolution of the word mash-up, and what it means. Basically, mash-up is a mixture. One can mix data from different sources - as done in data mash-ups, one can present in novel intriguing ways - as done in consumer mash-ups, and one can present it in unique way to bring out some aspect of the data which is now visually easier to recognize - as done in business mash-ups. But those are only the early mash-ups, the first-borns, if one may say so, of a species that will soon thrive and multiply.

One mash-up, that I think is just on the horizon, is the mashup of user experience. A mash-up that will enable users to experience information where they want it, regardless of where they are at this time. Whether on their desktop, or on a web site.

Experience web news when on Word. Or experience a Excel table, when viewing stock news on a financial news and stock movements. Mash-up information interfaces and tools and enable a seamless user experience of information where they want it, without worrying about which interface tool they should pick up. Not just about data, but about user experience. A mash-up that merges our private desktop world, with the outside web world, and in a manner that puts us in control.

-Sangeeta

The One Interface to Bind Them All

The Browser. Thats the one interface that people think would bind them all - all the multiple data sources. Whether the information comes from the ever expanding web, or from deep within the enterprise data ravines, or from Excel files, or from E-mail stores. If you want to reach it to more people, just put a browser on the face of the data store, and you are on! Bingo!

A large part of that is true. The browser indeed has changed the boundaries of the information repositories forever. The browser has enabled information to be served to people in self-serve modes and has created the internet revolution.

But lets not get carried away. Browsers are great for many applications, and are woefully inadequate for some. Firstly, browsers have to be told where to go. Further, browsers open up in different windows and force me to switch my context. Browsers work only on desktops. Browsers do not enable me to mix data from more than one source. One browser page per site. Browsers do not allow me to add my own stuff and store it away for future use. Browsers do not enable me to create my own format - my own view of data.

Because of this, despite a decade of the browser being born, the desktop applications are still there, and will continue to live. Further, with the growing complexity, there will be- and are - more interfaces that I would need to do my job. Interfaces that would track things for me, interfaces that would enable me to work on my documents that I store on my machine, interfaes that would bring me data wherever I am, without forcing me to change my context.

I want information to be available wherever I am. I may want it when I am creating a report in Word, or creating my family contacts in Outlook, or as I track my recent investments in a stock. And I find that using a browser to look up information is not as productive as I like it to be. No wonder, I have multiple applications on my desktop as I work, and as I flit between applications and web pages, changing my context, and cutting-pasting across windows several times during even 5 minutes. Worse is, when I am away from the office and away from my desktop.

Surely, there is a better way than a browser. The browser does what it does admirably. But does it do all the things that I want it to do? Is it possible to create a browser that is a one-interface-that-binds-them-all?

A Swiss Army Knife has many tools. There are multiple things a woman needs to do. How can one tool do them all?

-Sangeeta

Well, why not Duet?

It seems to me that it was time I spoke about some of the things that I do like about Duet from SAP and Microsoft. No, it isn't a change of heart - I still think Duet has a long way to go before it delivers on what people expect it to be - but there are several things that are right about it.

For one, Duet is right in providing a solution to the friction between the workflow-enabled-business-process SAP world and the personal colloboration based Outlook world. It helps a user perform work flow transactions from within Outlook. It helps users look up data, compare it with previous history, and finally, complete the transaction from right within Outlook, and have it reflected in SAP.

Secondly, Duet does a decent job in providing the context sensitive data elements along side the automated e-mail generated, so that the user can effortlessly do the job. Further, it intuits the users context reasonably well with its integration with Outlook objects, and succeeds in providing a reasonably intuitive delivery of information.

Thirdly, the experience is seamless, and yes, it does look pretty. Prettier than SAP screens.

If Duet could connect to more data sources besides SAP, and enable users to build their own pathways inside Outlook and enable their work-flows, it would be loved by CIOs and users alike, notwithstanding its heavy middleware and upgrade costs.

-Sangeeta

The connected 'Office": The Microsoft Version

The MS Office applications are a big part of the desktop, and for years have faced the same fate as the MS desktop. The MS Office applications have been disconnected with back-end process applications and web data stores, and users use Cntrl-C and Cntrl-V to navigate across the two worlds.

Office Business Applications, aka OBAs, is a product from Microsoft to change that. OBA provides a mechanism to connect Office to back-end Line of business applications. With OBAs, information workers can connect, interact and perform business transactions with Office as a front-end. Sounds good, does'nt it?

But of course, OBAs will only run on the recent versions of Office. OBAs are built using Microsoft Visual Studio for Office, and under the hood, are deployed on Microsoft Office Sharepoint Server, and as given in the Price Management example here and SCM example here, runs on Microsoft middleware stack, on top of an enterprise application. Considering that this is a offering from Microsoft, this was expected. Does'nt Microsoft always want to have it all?

OBAs could be built in other ways and without adding all that stack from Microsoft. Extensio has built its own set of OBAs - which we call as Extender for Excel, and Extender for Outlook, on a SOA backbone. It is free of any Microsoft middleware components, and the server can even run on Linux. And yes, it is built on older Office interfaces of Microsoft and can run on Office 2000+ installations.

Things are beginning to get interesting!

-Sangeeta

Duet is not doing as well as it should be...

Someone called it the Duet sad song!

It is now out in the open. Steve Ballmer acknowledged at the keynote for Software 2007 that Duet has not done as well as it was expected to.

Not surprising. Afterall, Duet could connect only to latest versions of SAP- AND could work only with Office 2003+ versions - AND had no development framework - AND required a whole bunch of middeware from SAP and Microsoft both - AND enabled only a few limited value packs that enabled a handful of SAP transactions though Office. With all that onerous bunch of software under its hood, and the steep upgrades it needed, Duet really did deliver too little.

So, what could happen next? Would Duet be given a quiet burial - or would it be re-packaged and re-sold as a newly christened Office Business Application aka OBA, this time around only by Microsoft, and only with Microsoft stack? Something tells me that it would be the latter.

Could it be that Microsoft just used Duet as a pilot for its own OBA run? Nasty of me.

-Sangeeta

Vista Gadgets would need Vista on the desktop to run

That is an obvious statement.

There are many of us who wonder why gadgets would not run on XP. Because there are many of us who do not want to upgrade to Vista, for lack of sufficient reason to do so. Why buy a more expensive OS when the current one suits us fine? There are others, desktop gadgets - mentioned in my earlier posts - which do not require Vista, and can run on Windows 2000+ machines.

Power to such alternatives! May their tribe grow!

- Sangeeta

The Desktop Getting Crowded, but why only Vista Desktop?

The race towards offering the best connected desktop is hotting up. Already there were quite a few players that had created desktop-resident technologies with Yahoo and Google being the heavy weights from the web space, and of course, Microsoft with its Vista Gadget offering - after all, Microsoft does own the desktop. But there are more players getting added in every month, and that too, more of the enterprise types than the consumer types.

Hyperion is a case in point. With its recent announcement of Smart Space Gadgets , Hyperion delivers, I quote - "always-on business intelligence (BI) and BPM to knowledge workers throughout the enterprise" from its BI application. Smart space gadgets run only on Vista, so I guess it can get rolled out only after the next desktop upgrade cycle. I saw BEA making a similar annoucement some time back.

But desktop connectivity can also be made available on desktops without Vista upgrades. Yahoo Widgets, for example, runs on all Windows 2000+ machines, and so does Google Desktop.

Then why this preference to Vista gadgets? Specially since Vista hasnt exactly seen a upgrade wave as yet. Incidentally, clever strategy by Microsoft - using their hegemony at the desktop to enter some other area. In this case, it appears that Microsoft wants to enter the enterprise back-end. They have done this earlier - piggy backing on some existing products to enter other areas, and with devastating effect. This time around, other vendors seem to be helping Microsoft to succeed in this upgrade-to-Vista-to-get-personalised-gadgets strategy. Interesting.

-Sangeeta

One of the Duet partners goes Solo

The term Desktop Gadget means Microsoft Gadgets? Naaaah...

While I was writing the previous post, I checked out Wikipedia, and was surprised to see that the term Desktop Gadgets currently points to Microsoft Gadgets.

Amusing, isn't it? As if Microsoft Gadgets are the only Sidebar gadgets - Pshaw!!

-Sangeeta

WhatsIsName - Widget? Gadget? Desktop Gadget? Pagelet?

There are so many terms floating around to call these little floating windows that developers are creatings for the desktop users. Not surprising, considering the fact that the terms are just beginning to gain currency, and it would be a while before the larger populace decides what to call this thingummajigs.

Newsweek called it widgets - though what they wrote about were desktop gadgets from Microsoft and Apple. Whats the difference? Not very much in the eyes of the users, maybe, but plenty in the eyes of the developer.

My take on these terms, and the difference between them:

1. Widgets: Small snippets of code-enabled-HTML that can be embedded in HTML pages by end-users themselves. The widget contains content that is hosted on some server, either owned by the end user or by some content provider - but the widget itself is displayed on some other site than the original hosted site. Typically on the "social" sites, such as Myspace, or on blogs, like this one. Widgets run on browsers only. Some people also call them web gadgets. Sigh - more names, more confusion

2. Gadgets/Desktop Gadgets: These are desktop applications that float a small pretty looking window on the users desktop. These are proper desktop applications that need to be installed on the desktop, in some cases, and do not need a browser. There are gadgets which work stand-alone, and are reasonably popular.

And then are some gadgets that require engines to run them. Typically, they have a mother application that provide the run-time environment for these pretty little windows. The mother application, called by some as the Widget Engine , has a interface that sits on the user desktop - as a sidebar, or as a toolbar. This mother application, or widget engine, hosts the pretty little windows, called Gadget (or desktop gadget) which runs within it. Google Desktop is one such example, and the MicrosoftVista gadgets are another. In case of the Google desktop, users have the download the Google Desktop on their machines. In case of Microsoft Vista, Vista comes bundled with the Vista Sidebar, and with some out-of-the-box gadgets. In both cases, other developers (other than Google and Microsoft) can build gadgets for the Sidebar, and make it available to end-users. Though the point looks obvious, but it still needs to be stated - Gadgets built for Vista would not run on other Gadget Siderbars, such as say, Google Desktop and vice versa.

There are people who call these desklets also. Yikes, another term...

3. Pagelets: These are again small code-enabled-HTML snippets that run on HTML pages, and need a browser to run. Much like widgets, but with a key difference. All the widgets, er, pagelets, are hosted on one Pagelet Server. Much like the "portals" of previous years, these pagelet server providers provide a entry page, where several widgets are hosted, each displaying its own content. Much like the previous portal pages, users can personalise their home pagelet page to suit their preferences. Needless to say, these too are available only through a browser. Best examples of these are Netvibes and fittingly called Pageflakes.

The underlying idea for all three, namely Widgets, Desktop Gadgets and pagelets is the same. To extend the idea of proving data within users' context, and go beyond providing links. All of them are designed to create a user experience that is connected, easy to use, and within the users context, a crying need for information crunchers today. They are also similar in another respect - that of getting data from mutliple places from the web, or some other data stores, and placing it within a particular context.

What the difference is, is really the context. In case of Widgets, the context is a page, owned and managed by the user. In desktop gadgets, the desktop of the user is the context, and in pagelets, the pagelet provider's portal page is the context. Does context matter?

I think it matters a great deal. But more on that later

- Sangeeta

SAP does not eat its own dogfood?

Jeff Nolan writes about SAP not using Duet itself .

I quote from Jeffs blog "How long before you see mashups entering the workplace that do things like expense management from calendar details, a solution you would have to buy from SAP/Microsoft (Duet) and implement providing you have all the pieces in place to do it, and if it were so easy why is SAP itself not running Duet internally (the answer is in the version requirements)?"

I cannot but agree with him. The kind of infrastructure and versions needed on the application, middleware, and the desktop for running Duet is way too onerous, as I pointed out here.

- Sangeeta

The Year of the Widget is Upon Us

The Year 2007 will be the year of the widget, says Newsweek.

But, of course! I had'nt expected this kind of resonance to my thoughts, but I sure feel validated.

Widgets featured in the article may fall short of going mainstream. The current crop of widgets run only on Windows Vista and on the forthcoming Apple OS. None of those are mainstream. Not yet, and not likely to be in 2007. The current lot also does not bring the desktop context to the widget, nor does it interact amongst each other.

The widget-on-the-web, such as Netvibes or Pageflake, may become mainstream though.

My feeling is, the term Widget will get redefined during the year, evolve to address the desktop connectivity issue, get re-purposed, and certainly kick up some dust - not just dust, maybe a storm.

- Sangeeta

Desktop is Desktop – and Web is Web – and never the twain shall meet! And should they?

The importance and role of the Desktop has been challenged recently by the Web, with some of the only-done-on-desktop stuff being done on the web, with on-the-web spreadsheets, on the web documents, on-the-web videos and what have you.

But would the Web replace the Desktop? No, not quite yet, may be never.

And here is why. The desktop world is about privacy and control, and the web world is about exploring, sharing and collaborating. Most of us build walls, called firewalls, around our desktop world to protect ourselves from the outside prying eyes. Our desktop world is our personal world, and the web is our social world. And yes, we would continue to want to keep them separate, or at any rate, would want to have to right to decide how much of our personal stuff we want to keep on the web.

Our information world, then, will continue to the divided into two – the personal one that is stored on our desktop, the one which is guarded fiercely from the world, with firewalls, intrusion detection software and anti-virus software and spam filters etc - and the other social one, the public information space that we inhabit and work on.

The same is true for the business world as well. Businesses will continue to barricade themselves within firewalls, put rules on who can access information when and where, and even, frame rules on what kind of web content can be viewed on their business desktops.

Does this desktop-web divide need to be bridged? Why should anyone require to bridge these two worlds? Because they are a pain un-bridged. Every travel done across these worlds, and it is done often enough, requires one to swap context, painful and error prone cut-and-paste, and is simply irksome. Clearly, this is problem which needs to be solved.

The Desktop and the Web - the time to meet each other has come!

Odd, A lone customer testimonial...

I recently read this article by Bruce Richardson of AMR Research, on how Duet is considered as one of the potent weapons to grow the SAP business by Team SAP.

The revenue potential for Duet, or applications that extend the context of the desktop applications to ERP applications is large. Specially on existing accounts. And yet, yet, there is only one customer, out the 300,000 licenses sold from 200,00o customer interactions, that has chosen to come out and talk about it.

Why this reticence from customers? After two years of the announcements and significant customer interest displayed during the launch time - and all this while customers keep asking for more desktop integration, there is just ONE lone customer who decides to speak about Duet publicly?

Odd. Distinctly odd.

In my opinion, these would be the top reasons for the customer silence:

1. The current Duet scenarios are too simple - and do not really deliver much. So, at this time, there is'nt anything to talk about.
2. In absence of development tools, the potential of desktop integration is not really well understood or realized
3. CIOs are yet to find the exact ROI for desktop integration. Sure, it improves the life of Information Worker. But how much? How many hours are saved? Does it reduce the number of steps a process takes?
4. No one has gone "live" on Duet yet. Maybe it is early days yet.

Maybe there is more to this silence. Any other opinion, anyone?

-Sangeeta