Mashups Finding Their Place in Enterprise IT | Blogs | ITBusinessEdge.com
The article links to interesting posts by Mike Kavis and others. I have to agree with Mike's view that Mashups should not be about "on-the-fly" applications. Rather, emphasis should be given to how they make enterprise data accessible to the average user, so that they can use it the way they see fit and compose new things.
The next point in discussion is the traditional Enterprise Portal and it's potential conflict with Mashups, which is a valid concern these days. As a co-developer of the only Open Source Mashup Platform for the last 2+ years, I agree that the traditional Portal has to go. But the concept of A Portal can not be removed from the equation. Why? Portals are a presentation layer while Mashups primarily deal with service (data) composition. It's true that a presentation component is involved but it can never serve the purpose of a portal like interface.
But. It's clear that traditional Enterprise Portals do not complement the power given to end users by Mashups. So the ideal Enterprise Portal of the future should be a platform that allows both IT and end users to collaboratively create the presentation layer. Apache Shindig, Google Gadget and Open Social Specifications should be technologies to look out for. These allow Enterprise Portals to become platforms such as iGoogle, Facebook, Hi5 and Orkut. The user experience is very similar to a portal, yet open to contributions by end users, not just the people running it.
So what do I see in the future?
As the piece points out, most of the mashup discussion has focused on building on-the-fly applications when, in fact, the real enterprise potential lies in accessing the data. Kavis decided to use a tool – he talks specifically about Jackbe's Presto or WSO2's Mashup Server – to “present various data services in a secured and governed fashion.” How they consume it, he adds, is up to them:
The article links to interesting posts by Mike Kavis and others. I have to agree with Mike's view that Mashups should not be about "on-the-fly" applications. Rather, emphasis should be given to how they make enterprise data accessible to the average user, so that they can use it the way they see fit and compose new things.
The next point in discussion is the traditional Enterprise Portal and it's potential conflict with Mashups, which is a valid concern these days. As a co-developer of the only Open Source Mashup Platform for the last 2+ years, I agree that the traditional Portal has to go. But the concept of A Portal can not be removed from the equation. Why? Portals are a presentation layer while Mashups primarily deal with service (data) composition. It's true that a presentation component is involved but it can never serve the purpose of a portal like interface.
But. It's clear that traditional Enterprise Portals do not complement the power given to end users by Mashups. So the ideal Enterprise Portal of the future should be a platform that allows both IT and end users to collaboratively create the presentation layer. Apache Shindig, Google Gadget and Open Social Specifications should be technologies to look out for. These allow Enterprise Portals to become platforms such as iGoogle, Facebook, Hi5 and Orkut. The user experience is very similar to a portal, yet open to contributions by end users, not just the people running it.
So what do I see in the future?
- Enterprises will open their data services via published APIs to end users. These will be secured where necessary.
- End users will use these APIs to compose Mashups.
- The future Enterpirse Portal will allow users to contribute Gadgets/Apps to it (similar to iGoogle, Facebook et al). Others who deem these Gadgets useful will add them to their individual profiles/personalized pages.
