Coghead, Datamashups.com, Teqlo: How will the everyman build mashups? | November 2, 2006
by Byron Binkley
The launch and subsequent buzz around companies like Coghead, Teqlo, and Datamashups has made me think about how end users (versus developers) will build web based mashups. The continuum of mashups from simply recombining content to legitimately new applications is fairly broad, and is dependent on how much business logic there is behind the scenes recombining the services. Programmers have the luxury of using their "logic layer" of choice. For example, looking at Yahoo!’s developer network you can get started with JavaScript, Flash, .NET, PHP, Ruby, and Python. But what will end users use to combine web services? Datamashups.com provides a couple mapping screens to bind one service to another, and Coghead shows some screen shots of what looks to be a more sophisticated visual development environment.
But is this the same old promise of end user programming to allow business users to create their own solutions? The comments on Techcrunch regarding Coghead expressed a range of polarized opinions from optimistic to loudly skeptical:
Comment 17.There are many needs where proximity to the problem and domain understanding are most important, and where the software engineering challenges are less daunting.
Comment 14. This class of application–those which promise to allow “non-developers” to develop (an obvious logical contradiction)–has been around for at least 15 years. To date not one of these products has ever made it possible for anyone to create any but the most basic business applications…The idea that you will get something new and exciting by taking this old idea and putting an HTML interface on it is to be wilfully ignorant of the past.
Comment 22.As noted above, the classic problem with any tool that tries to make development easy for non-developers is that sooner or later you end up with a tool that either doesn’t do much, or else has a learning curve that will frighten off the noobs.
Comment 28.Initially they are fun and exciting to use. However, as the user requirements evolve, you find that the tool cannot meet the requirements (usually when it comes to creating ‘business logic’).
I think there’s a key distinction between mashup building and the long string of CASE tools and RAD environments that never succeeded in leveling the playing field. Namely, end users building mashups are not expected to create the services that they recombine. With a little business logic "glue", they might be able to leverage the componentized work of developers and actually build useful new applications. And that logic layer and development environment could be significantly less sophisticated than a typical IDE. It could even be a visual programming environment like Coghead or, dare I say, Proto? Programmers aren’t going away any time soon. But I do think the mashup philosophy towards end user development presents a cooperative and collaborative approach to bridging the gap between end users and developers.
The eyebrow raising question I still have is whether there’s any reason for mashups to be built, or for that matter distributed, in a web browser.
