All you need is a quick visit to John Musser's most excellent programmableweb.com or to one of the upcoming Mashup Camps (the next one is coming up in Silicon Valley in July, register here) to know that mashups are the hottest software development category going right now. Mashups, normally a kind of browser-based software that draws on multiple disparate Internet sources to arrive at some unique user experience, span the gamut from finding cheap gas (the data is superimposed on Google Maps) to discovering what local muscians are playing in your area and then sampling their music through downloadable MP3s (before wasting your money heading over the bar). See podbop.org for what I'm talking about (podbop won the first place prize in the Best Mashup Contest at the first Mashup Camp in February 2006).
Even grandmothers and 5 year olds are mashing software up by plotting bird sightings and the birthplaces of the world's explorers on Google Maps (according to programmableweb.com, Google Maps accounts for 50 percent of overall API usage in mashups).
Now, roughly a year after consumers and consumer-oriented developers started grokking the idea of mashups, big business are getting hip to the idea as well. Compared to IT projects that once took months, years, or even a decade to complete, the art of mashing up software can often produce similar results in far less time -- sometimes in just hours. Not surprisingly, some of the companies that are well-known for enterprise-class software development tools -- the sort of tools that typically enables teams of developers to re-use business logic components while collaborating on a single codebase -- are looking to make it easier for businesses to not only build mashups in the same fashion, but to do in a way that the results are both reliable and secure. Today, reliability and security are not what mashups are known for (and they'll have to be known for that if enterprises are going to consider their adoption). In fact, Mashups are pretty brittle. If one source upon which a mashup draws (for functionality like mapping or data like an address) is unavailable, the mashup usually fails.
Last year at the second Mashup Camp, IBM started to show and discuss its QEDWiki-based collaborative mashup development platform with the basic idea being that if wikis can get knowledge workers to collaborate on documents, why can't they be extended to also get developers collaborating on mashups. Yesterday, at the Enterprise 2.0 Conference in Boston, IBM formally rolled (audio here) out its Web 2.0 strategy (as a part of its Information Management portfolio which includes solutions like DB2 and Informix), an announcement that included discussion of the QEDWiki development environment and a technology IBM calls the Mashup Hub (see IBM's YouTube-based video demo below). The Hub is where re-usable components including software widgets (both internal and external), RSS feeds, SQL queries, etc. are registered for use in the collaborative environment.
Meanwhile, BEA which is mostly known for its J2EE server was here at Enterprise 2.0 showing of AquaLogic Pages. Somewhat aytpical of BEA's solutions, AquaLogic Pages does not require other BEA infrastructure (good!) and conceptually has a similar architecture to IBM's solutions where re-usable components are registered with the system and then developers can collaborate on software that uses those components. When I asked BEA's Director of Emerging Products Ajay Gandhi on video (see video above) whether he felt he was competing with IBM, his main take was that AquaLogic Pages will ship next month while IBM's solutions are still being prototyped. Via telephone, Dan Gisolfi who works in IBM's Emerging Internet Technologies area countered that even though IBM's products have yet to go "gold" yet, preview editions will be available for download from IBM's AlphaWorks site at the end of this month.
Finally, to get a better idea of how AquaLogic pages works, the video above includes a demonstation given to me by BEA's Gandhi (and below is IBM's QEDWiki video....bear in mind it was produced by IBM. I was interviewing Kim Polese when the announcement was made):