Mashup vs Widget?

Following the announce about the founding of “Mashery“, an interesting start-up that will create “resources” for developers, API providers and Mashup users, I was wondering if such company was targeting Widget developer or Mashup developer?
So, came the question of what is the difference between a Mashup and a widget, after all, they seems both some technology that could be used to create a new web page/service?

The actual situation:

  • a mashup, is a new web service, usually a web site, created by aggregating existing Web services. Typically,  a service tot display renting house on Google map is a mashup, taking information both from a renting site, and from Google Map, aggregate them, and displays them
  • a widget, is a “micro application”, that do a small individual task and that can be put as a component in a desktop or a web page. But some widgets can be a mashup…

So what are the real differences between the two? Today, a mashup is more a programmer affair, while widget is more an end user technology. The other actual difference is that most of the time, the newly creating service imply that there is glue between all data sources (for instance, some code to put house locations on Google map) while widget does not speak each other: the clock widget does not talk with your flickr badge widget, etc…

  But the limitation starts to fade down, and I think that this will slowly disappear, and I am quite sure that there is probably tons of widget that act as “mashup”.

  • It will be easier and easier to create new composites application, based on third party app for everybody, and not only technical people.
  • The question will be again business model? One trend in Widget industry is to provide money to developer according to widget usage, based on revenue like advertising. But the value come also from the data itself, the initial provided one. If the mashup/widget creator get some revenue from his creation based on external data source, these external data source could ask for money too….

This latest part will be an interesting challenge in the future, and I think that the value of company like Mashery will be also to be able to manage the value chain and provide fair revenue to the entire participant….

Update: waow, what a sync, I was ready to post the article when Techcrunch pointed out this new start-up, dappit. Seems to be very close to what I describe as the ability to create Mashup for end user.  But I’ve tried to use it and was not able to extract easily simple information for an HTML table, so I do not share Techrucnh enthusiasm. But this just means that there is still a lot to do, but direction is here

Technorati Tags: ,

7 thoughts on “Mashup vs Widget?”

  1. Thanks for mentioning Mashery, Tom. You’ve provided a great overview of the “what is an API”, “what is a widget”, and “what is a mashup” debate that is reminiscent of the still-unresolved “what is a blog” debate that has been a topic at conferences for the past several years. I think your diagrams will be of great assistance to me as I try to explain mashups and widgets to friends.

    Fundamentally, we at Mashery believe that people who offer web services of various different types (define ‘em as you like) will need various enabling technologies and services that other such companies will need, and that are more efficiently delivered as a service across multiple providers than they would be if developed separately by each web service provider. This will benefit both web service/API providers and the developers who build applications using the services, APIs or widgets.

    - Oren Michels, CEO, Mashery

  2. Pingback: Praxis
  3. Hello Owen,

    Thanks for the feedback…Hope to see in the near future some examples of newly created services using mashery…

  4. Tom, there is some existing research on end-user mashups since early 2005:

    http://simile.mit.edu/piggy-bank/

    The idea is that as long as people still think of data-centric mashup services delivered from web servers, users will still experience limitations on what can be mashed up. Mashup capabilities need to be moved over to the client side where each individual user chooses the data sources that suit his/her need at each particular moment.

    For a glimpse of how automatic scraping can be made available to end-users, see this video: http://people.csail.mit.edu/dfhuynh/research/papers/uist2006-augmenting-web-sites.mov

  5. Davdi, thanks for the link It’s an interesting first step, but just a first step. End user mashup based on purely extraction of data from web pages will be quite limited first, and non optimized….
    I envisage tools provided by some third party that will allows end users to do this mashup directly, with much more powerful results than direct extraction from web pages…

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>