You may have noticed the WSO2 Mashup Server link has been retired from our menus, and seen that the WSO2 Mashup Server page on wso2.com directs you to the WSO2 Application Server product page. Curious as to what’s going on? Here’s the whole story, from inception to the present and looking towards the future.
Where the WSO2 Mashup Server led the way
I joined WSO2 back in 2006 as Director of Architecture for Mashup Technologies to explore ways to make the emerging stack of WS-* specifications approachable and efficient for the average Web developer. The result was the WSO2 Mashup Server, which introduced a number of valuable ideas and features:
- Make dealing with XML payloads easier.
- Interface with other systems that Web developers are interested in, particularly feeds, data sources, and other web pages.
- Build a try-it functionality allowing developers to point at a WSDL and get a usable form-based user interface to explore a Web service (or to provide a default user interface for any service.)
- Make the results available in forms a Web developer cares about: web pages, gadget portals, feeds, instant messages, email messages.
- Seed an ecosystem of reusable mashups by building community features into a multi-tenant environment, where each mashup exposes services that can be reused and recombined.
- Host a public site (mooshup.com) for the mashup community.
The resulting product, the WSO2 Mashup Server, broke new ground and gained a lot of interest in the community, proving the value of many of these ideas. Here’s what the 1.x version looked like back then:
These core features and ideas have over time influenced the WSO2 Carbon platform. As more of these ideas have been incorporated broadly into the platform, the layer unique to the Mashup Server has become increasingly small. Here are some ways the Mashup Server informed WSO2 Carbon platform evolution:
- Multi-tenancy. The early multi-tenancy (really more of multi-user than full isolation) in the Mashup Server allowed many users to register, author and share their own mashups with others, has evolved into a full multi-tenant architecture across the Carbon system, and has been a core feature cloud-enabling the WSO2 Stratos Platform-as-a-Service.
- Social enterprise. Enabling community features like tags, ratings, comments, granular feeds and search embedded those capacities into the underlying WSO2 Governance Registry. They remain a key part of our governance capabilities and continue to evolve through initiatives such as the WSO2 API Manager’s API Store interface.
- Try-it. Try-it for SOAP services has been integrated into all our products that focus on exposing services. I personally think we have lost a bit of the “default user interface” focus over time and hope to push us back to regain and extend that aspect of developer experimentation, but an increasing preference for RESTful services which can be readily explored through simple tools like Curl is making that less urgent.
- Gadgets. The Google gadget dashboard and gadget generators made their first appearance as a component of the WSO2 Mashup Server, but were fairly quickly spun out into a separate product.
- WSO2 Carbon. It’s my view that WSO2 Mashup Server became in large part the straw that broke the back of the camel of a suite of related, but separately developed, products. With many capabilities and shared components between WSO2 Mashup Server, WSO2 Data Services Server, WSO2 Governance Registry, WSO2 Gadget Server, coordinated development and releases across these products became untenable and helped motivate the hard but incredibly valuable work of moving towards the world’s first fully componentized middleware platform.
- WSO2 StratosLive. The multi-user publicly hosted WSO2 Mashup Server branded as mooshup.com site became redundant as the whole WSO2 Carbon platform emerged through WSO2 StratosLive as a solid public middleware PaaS encompassing the whole range of WSO2 products. Mooshup.com was retired quite a while back when StratosLive came online.
Where the WSO2 Mashup Server missed the boat
It’s worthwhile to review some of the areas where the Mashup Server failed to reach the mainstream. As a SOAP-centric and XML-centric model, it lost some relevancy as RESTful services and JSON have dominated the API ecosystem targeted at Web developers. The Mashup Server’s focuses on APIs didn’t provide an easy environment for developing Web Applications – many Web Apps I built on as mashups were comprised of static HTML pages, AJAX and XML, without dynamic HTML creation and relying completely on AJAX to invoke any kind of server-side processing. Not always the most straightfoward solution.
Vice President of Business Development