User>Driven

Have you noticed a tendency for successful Open Source projects to become less free over time? I sure have. Recently JBoss adopted MySQL's strategy of only offering support for it's licensed Enterprise version. And, of course, MySQL itself has just been bought by big, bad Oracle.

I usually get a kick out of the tech comic Geek And Poke. I thought this one and this other one were particularly apropos. 

Google Gears is here and it promises to change webapps forever. I've been blogging about the fact that lack of offline access makes many webapps unusable as every day tools for a while. And I confess I thought Google's browser-centric approach to application development was their Achilles heel. By creating and open-sourcing a set of javascript extensions that allow webapps to operate when offline, Google has changed all of that in one deft move.

Google Gears was inspired by a user (naturally) of Google Reader, a Google developer who wanted to read feeds while on the company shuttle that had a "flakey" net connection. I laughed when I read this because it sounded just like my issue with getting connected on the train every day.

It uses a local instance of SQLite together with catching and multithreading features to capture what you do while disconnected and synchronize with the webserver when you are connected again. You can download the beta version of the plug-in for IE6+ and Firefox 1.5+, but individual webapps must be modified to call the new javascript extensions before they can use the plugin to operate offline.

Adobe is listed as a development partner for Gears. Their Apollo team was developing parallel technology (right down to using SQLite according to Michele Turner, Adobe's VP of Product Management) but will now apparently focus on making Apollo a way for Flash developers to make use of Gears. This really takes the wind out of Apollo's sails but those who believe in the Flash/Flex approach to RIA development will have even broader support for enabling offline activity in their apps with Google under the hood. Apollo can also handle Ajax apps, but Ajax developers may be more comfortable with an open source IDE such as Dojo. (Dojo has also announced support for Gears.)

Google Reader is the first app to be Geared up. Look for a review shortly.

There’s a battle for the handset being waged now among RIA development environments. I predict the winner of this battle will win on the desktop as well. Java on the handset has a commanding lead today, according to David Berlind of ZDNet, but it faces new and formidable foes in Adobe Apollo and Microsoft Silverlight.

Rich Internet Applications (RIAs) provide a more interactive, responsive and desktop-like user experience compared with traditional browser-based apps using html or even the new crop of javascript-driven Ajax applications like Gmail. Flash apps, for example, can run inside or outside of the browser, can quickly exchange data with the server without requiring a page load, and can support complex interactive forms (read commerce) and animations (read games).

A key advantage of RIA development environments like Apollo, though, is that they wrap Flash apps (and Ajax and other existing application types) in a layer that allows them to operate without a connection to the network and then allows them to sync up with the server when the network is available. This is critical for mobile apps where your signal may not always be available (like on parts of my commute home by train). Imagine trying to make a purchase or play a multi-player strategy game when your connection kicks out. Did your credit card number or your latest brilliant move go through?

I contend this ability to keep working when disconnected is also critical for desktop apps and that dominance in mobile app development tools will lead to dominance on the desktop as well. PCs are now mobile devices, too. Everywhere you go on trains and planes and in the backseats of cabs you see people on their laptops typing away. Laptop prices have come down so much, people just aren’t buying stationary desktops anymore. I personally think of my daily train ride as part of my work day. The only trouble is, most of us are stuck working on our laptops disconnected.

Worse, desktop apps that make no use of the Internet are quickly becoming a minority as Microsoft rolls out Office Live, Google apps gains traction, and people switch in droves to web-based email apps. A perfect example is Quicken – one of the most popular apps on the planet – which began as an offline checkbook balancing and check-printing utility and now downloads balances from your bank and credit cards and pays bills electronically.

Mobile app development is all the rage, but frankly, my laptop spends less time connected than my Crackberry every day and I use it more for the kinds of things RIAs are good at. Moreover, RIAs are often self-downloading and self-updating like traditional webapps, making them much easier for Joe consumer (or Joe IT guy) to install and manage than traditional shrink-wrapped software. All of the advantages RIAs bring to phones play as well or even better on laptops, in my opinion.

So once one or more of these RIAs becomes fruitful and multiplies on mobile devices, enabling more fun, more interactivity and more commerce in that realm, I predict we will see them more and more on these other, lager mobile devices people use every day. The decision to use your laptop vs. your mobile will become less about the availability of a network signal than about whether you are standing or sitting and how much keyboard and screen real estate you need for a task.

The kicker for application providers, though, will be when they realize that not only can they leverage the skills their people learned for mobile device development on desktops, but that there is a demand for access to the same app on both platforms. Gmail is already on the web and on my Crackberry, for example, as is my access to Exchange email. How about letting me pay my bills on my mobile, Intuit? And can I have it in a familiar interface like Quicken?

My dream is that using the same development environment on multiple platforms will make seamless the use of these apps across those platforms. Compare this with the cranky, hard to maintain and hard to use integration between Exchange, Blackberry Enterprise Server and the four different clients I have to use to access content from the Exchange server. (Yes, four: 1 is Outlook 2003 on my laptop, 2 is Blackberry’s email client, 3 is Outlook webmail via IE – which is relatively full-featured - when I am using a PC other than my laptop, 4 is Outlook webmail via Safari or Firefox on my iMac – which is not very full-featured.)

Where can we see the beginnings of this trend? Check out Allurent. (Full disclosure: Allurent is a partner of my employer, ATG.) Their Flash-based eCommerce UIs “deliver rich interactions throughout the shopping lifecycle, from browsing to choosing to buying.” Imagine if you could shop your favorite online clothing store and use a rich UI that let you zoom in on individual items, look at close ups of fabric swatches, watch videos of models wearing the clothes, browse outfit configurators, and place your final order – all on your cell phone or laptop on your way home on the train or in your carpool.

I give the industry 12-18 months to start offering rich internet UIs for games and eCommerce sites that can operate with intermittent connections to the net. Virtual Ubiquity has announced their intention to release an Apollo-based word processor that can work online and off. Anyone know of examples out there already?

Over the last 6 months I've been asking IT managers whether they prefer webapps or classic desktop apps for deployment to their employees. Most have said they prefer webapps but they use whatever does the job best. I like this pragmatic answer. It tells me these IT folks are sensitive to the requirements of the business (and hopefully of the users).

Ryan Stewart at ZDNet has a good list of reasons for when to use each platform. I think he glosses over the need for offline access in his comments at the end, but the list seemed otherwise solid. 

Michael Fitzgerald has an article in the New York Times about Web 2.0 word processors that talks about the shift to online apps that started with the late '90s idea of the network computer and now seems to be catching on with wide adoption of broadband.

He talks about these online Office substitutes being mostly for people who don't want to pay for or manage heavy-duty traditional apps like those coming out of Redmond. He also touches on the ability to share documents stored online (though without providing any examples of where that is useful).

He shows some healthy skepticism of Mitch Kapor's claims of unsettling the major players with a new online application paradigm, but he cites only inertia as his reason.

My main concern with online productivity apps is access. They are both more and less accessible than their traditional offline counterparts. They are more accessible in that you can get to them from any web accessible computer (and sometimes other connected devices). But while net access is becoming more ubiquitous, it is not universal. I spend two hours most workdays on the train disconnected. I work with my laptop - and it's prime productivity time because there are no office-mates or children around to distract me - but I couldn't use an online word processor during those times.

This is why I think Adobe's new Apollo technology is so interesting. Rafe Needleman on Cnet's Webware blog says the coolest thing about Apollos is that it allows you to create connected apps that run outside the browser. Maybe he's just more connected than I am, but I think the coolest thing about it is it lets you make apps that work both online and off and that synch up automatically when you reconnect. Check out this demo at Demo where Mike Downing shows an ebay app that doesn't miss a beat when he dramatically pulls the network plug on his laptop. zdnet also has a good backgrounder on Apollo if you are curious. Adobe also has a wiki page on Apollo with a good selection of info for developers.

People are saying Apollo is Adobe's move from the browser (with Flash) to the desktop, but it's more than that. Making RIAs work offline is bringing Web 2.0 to the offline world. Even if Apollo doesn't reach the moon, I see good things coming from this idea.