www.dpingram.com

A friend just alerted me to a Forrester paper on RIA’s entitled “eBay San Dimas Marks A New Era For RIAs”. (You can buy it from Forrester here).

Here is the introduction paragraph to whet your appetite…

The much-anticipated beta version of eBay San Dimas has arrived, ushering in a new era in rich Internet applications (RIAs). Adobe Integrated Runtime (AIR, formerly known as Apollo) applications like eBay San Dimas take RIAs out of the browser and put them on the desktop. These desktop applications enable occasionally connected use, customized content views, and a branded experience that can act as a platform for closer relationships with customers. But desktop RIAs aren’t for everyone. Companies must assess whether their power users will benefit from the capabilities of AIR applications in a world in which the desktop will likely become very crowded.

What’s interesting is seeing this new class of application that is made up of both software (run locally) and services (accessed over the cloud). I love the model because it’s bang in line with the ethos behind izimi/sharenow in which I am involved. Of coures there are other examples of note, including Desksite, Maven, the eBay San Dimas project, plus a whole raft of new apps that are being developed on Adobe Apollo and Google Gears.

You’ll probably recognize the swing cycle we’ve seen over the last 15 – 18 years in software: dramatic swings from dumb terminals (all server side), through client/server, to internet (server side), back to java (it its inception, which promoted NCs that downloaded applets as needed which were run on the client), back to internet (as java initially fell out of favor), to …

History has seen different groups promoting their own points in the spectrum as the ‘correct’ best way, depending upon their own vested interests. But I think that is now passing.

What I think we see now is a maturity that causes us to take a less ‘fashion-conscious’ view, and one that sees us asking “where can we best do this bit of processing?”, and then creating systems which partition parts of the whole application where they are most appropriate. So, I love the principle of RIAs.

Its very interesting to think that the traditional Google supporters (with the concept of SaaS - software as a service) used to find themselves at war with the Microsoft supporters (and the concept of software that is installed and owned). Yet today with Apollo, Gears, and even Microsoft’s own positioning (they now speak of (”software AND services”) there may yet be a recognition that you choose the architecture that best suits each particular application on a case by case basis.

Old-school software developers are turning in their graves at the thought of opening up APIs into their code so that others can write new apps that use their data and processes. Its a scary notion, but it makes perfect sense for consumer web apps in the web 2.0 world.

Consider this, if you are developing and running an enterprise app, something that runs your business, something that your staff use in their daily processes to conduct your business, the story is very different. Often your data is considered a component of your competitive advantage so why would you want to open that up to possible competitors? You probably wouldn’t, and that’s how many MIS managers minds are still thinking.

But, in the world of the read/write web, where apps are written for consumers to use, and given to them for free its a different story. We monetize our businesses based on large numbers of user using our services over our competitors, and that is achieved by being competitive and being well known - opening up an API is a neat way to do this. Why? It’s well explained here (except from Wikinomics by Don Tapscott) with the aid of an example, Technorati:

“Perhaps the most powerful characteristic of the programmable web is that it invites collaboration by design with open standards and open application programming interfaces (API) that allow separate websites to intermingle. Startups like Flickr, 43 Things, del.icio.us, Technorati [DI> and Facebook, Yelp, Google Maps, Microsoft Maps, etc, etc] for example opened up their APIs as a way to crank out new features, attract users, and scale up their businesses quickly.

Lets just reinforce that point can we, this is a strategy to:
• crank out new features
• attract users
• scale up your businesses quickly.

If you are operating in today’s web space aren’t these the things you are trying to achieve?

“It comes down to a question of limited time and, frankly, limited creativity” says Tantek Celik, Chief Technologist at Technorati. “No matter how smart you are, and no matter how hard you work, three or four people in a start-up – or even small companies with thirty people – can only come up with so many great ideas.”

Its all based on a principle the new generation of Web start-ups learned from the open-source software community: There are always more smart people outside your enterprise boundaries than there are inside. By opening up their APIs companies create an environment for low-risk experimentation where anybody who wants to develop on top of their platforms can do so. Celik says there are potentially millions of developers who might just have the right combination of skills and insight to create something really valuable. “No need to send you a formal request,” says Celik. “They can just take those APIs and innovate. Then, if someone builds a great new service or capability, we’ll work out a commercial licensing agreement so that everybody makes money.”

Other posts from me about API and Mashups

The push towards openness continues. Local reviews site Yelp rolled out an API at the start of August, allowing other developers to build apps upon and mashup Yelp local search data.

Functions include the ability to retrieve reviews and ratings, display review info, track reviews for a certain business, display images of high-rated local businesses and look up business ratings using the business’ phone number, among other things. Examples include a map-based search for sushi restaurant reviews embedded on your site.

One of the first to mke use of this is a new app for the iPhone called iPhocal it uses the Yelp API to offer you local search on your iPhone.

The search tool works as you would expect. Type in a search term and enter a city, and you’ll receive a list of search results. Choosing one of these results will lead you to a detail page that shows the pertinent information, like address and phone number. You’ll also be able to read the comments of Yelp users that have reviewed the particular venue or location you’re looking up.

Original article at Mashable

Facebook confirmed yesterday that it has killed off its Courses feature. This was a way to enter your college classes and display your courses on your profile, created prior to the Facebook Platform. In contrast to the usual scheme of things, however, Facebook is retiring Courses in favor of better services created by 3rd party developers on the Facebook Platform:

As of today, we’re turning off our version of Courses and have decided to turn this over to the developer community and let you – our users – decide which Courses application works best for you…In many ways, our developer community is the best suited to create the applications that help people connect, track, and collaborate.

Although most developers assume the worst about APIs - that they’re simply a way to outsource research and development so the host can steal all the good ideas - this move contradicts that assumption. Instead of trying to clone the best apps in-house, Facebook is referring users to 3rd party applications like Course List and Courses ++.

Mashable

Edgeio, the aggregated classifieds website, has created a widget that lets you distribute content and enables you to sell content directly through the widget as well.

This makes it easier to distribute content because it allows customers to purchase the content via the widget, and view it immediately through the widget as well. It removes steps for paid content, and allows publishers to expand on the places their content can be seen. What’s more, is as a widget distributor, you can take a cut of the sales made. This is like earning ad revenue, but for direct sales of content, instead.

The widget can be used by any publisher, from a newspaper to an e-zine or a blog. Pick and choose which content will be sold through the widget (podcasts, text, images, etc.) and place it on your site or encourage others to place it on their site as well. It’s a good way for publishers to retain their branding while spreading full access to their content. I consider this a good way of taking what Scribd has made popular and applying it to branded, paid content distribution.

Mashable

After the Facebook API and the Pownce API comes the coRank API. This story is from mashable…

coRank, the site that lets you create your own Digg, has launched the first version of its API.

All users with coRank sites have their own API, which makes it more flexible for the creation of applications to integrate coRank data. You’ll remember that while coRank first offered a way for users to collaborate within the main site to form groups of social bookmarking and aggregation tool, coRank shifted a bit after it’s initial launch to offer a way to create your own network of sorts for these social bookmarking tools. So further integration of the API into these networks will let you do more with your site. The API can be used in the more general sense as well, tying it in with your own personal website, blog or service.

It’s important to note that this first version of the coRank API is read-only in functionality. The upcoming version, which is currently being tested, will offer writing and editing functions.

Source - http://mashable.com/2007/08/11/corank-api/

In these days of open source, crowd-sourcing, and peer production (read Wikinomics if the last two terms get your juices flowing) software developers are embracing the almost infinite development power of the crowd - that crowd being the tens and hundreds of thousands of smart developers around the world. Its rapidly being recognized that no matter how big your development team you cannot compete with the almost limitless external resource. Social media technologies, email, IM, VoIP and the like are making collaboration even easier, allowing previously unconnected developers to self-organize into powerful competitive teams.

Facebook launched its API and benefitted from the mass of new Facebook apps that drew new users, media buzz, and yet more users. Here is a great piece from Mashable about the new Pownce (limited) API:

IMified, the tool that lets you interact with web applications via your instant messaging client, has created a Pownce add-on.

The first add-on of its kind, IMified has taken advantage of the limited Pownce API offering to integrate it with its host of services. This brings you full instant messaging support for your Pownce account, enabling you to post messages, links and events. You can also get instant messenger notifications for new messages that have been posted to your Pownce account. In a sense, this also makes Pownce even more of a micro-blogging tool that resembles Twitter, which already has instant messaging integration.

In related news, Pownce has announced its upcoming public API, which will aim to address the concerns and requests brought forth by the developers community. We’ll be sure to see even more mashups as a result.

source - http://mashable.com/2007/08/12/imified-pownce/

I love Zillow. Its a truely web2.0 site (i.e. deep user involvement, UGC, social networking) that is based around residential properties. The great thing about it is, if you live in the US, YOUR home is probably on there with estimated valuation etc, so your sort of gently but involuntarily already involved. Anyone can update info about any property, but if you claim your property you then get to lock down some of that info and make your own updates.

There’s an old parallel with this (which I cant recall right now, but I’m racking my brain to remember), the way you are ‘involved’ before you’re even involved is very smart.

Zillow has raised $57 million in venture capital.

In product terms a ‘mashup’ is a website, or more usually a web-based application, comprised of two or more components from different sources, but presented to the user as a single, seamless ‘experience’ or application. Today, most developers experimenting with mashups are using consumer-centric content from the likes of eBay, Amazon, Google, and others. However, in the future, corporate developers may well combine Web service elements from a range of vendor solutions with bespoke, in-house line-of-business applications to present business users with their very own “Enterprise Mashups.”

Advantages: speed of getting an offering to market, while one team is tucked away coding it all themselves, another bunhc in another garage are cobbling it together from pre-available and open components. Speed to market has never been more important, because if you miss the wave you probably loose out to a more agile, faster competitor.

The wikipedia definition is here

Dion Hinchcliffe says…

“One of my favorite Internet stats to check out are the mashup and open API trendlines on the front page of Programmable Web. Consistently, month after month this year and right up until present day, we’ve seen the mashup stats climb intriguing steadiness. And along with hundreds of available Web APIs, we’ve also maintained a sustained rate of approximately three brand new mashup applications being released per day. Interesting growth, but is it the start of a major trend?

It’s too early to answer these questions though the broad outlines do seem to be coming together.”

OA by Dion Hinchcliffe is here