I started Tuesday morning at the conference with Cliff Lynch's keynote about Cyberinfrastructure, which I will confess mostly went completely over my head. I took notes, but the talk didn't really make much sense to me at the time (I don't have that deep a background (yet) in the area of electronic scholarly communication - which I think was the general subject of the talk) and my notes make no sense to me at all now. So, I'm going to cruise right into the first major session, which was about mashups. Now, I'm already fairly familiar with the idea of mashups (for folks reading this who don't know what a mashup is, a definition is coming a bit further down), and since I have a programming background, I'm confident that I have the skills to create one, should I ever be inclined. What I don't seem to have these days is a lot of time to keep up with what kinds of mashups are out there, so I thought this talk would give me a good overview of some of the stuff that's out there now (which it did).
Darlene Fichter, Data Library Coordinator, University of Saskatchewan
- They're like the recipes on the back of Jelly Belly jelly beans - somewhat pointless.
- Stupidest term ever.
But there's a serious side to mashups, too: Enterprise mashups at IBM:
- Lets employees create applications in 5 minutes via toolkits.
- Gives people the freedom to innovate - short development time for really small projects and widget-like apps.
The way Darlene sees mashups:
- today's playground
- even the trivial stuff shows the potential for other things
Definition of a Mashup
- A web site or app that uses content from one or more sources to create something new.
- Content is typically sourced from 3rd party via an Application Programming Interface (API) or an RSS feed.
- Need open data.
- Not necessarily free data, some can be licensed for a fee.
- Open set of services.
- Need a programmatic way to get the data.
- Need a culture of trust.
- This is the idea of small pieces loosely joined.
Mashups are building blocks - think about legos - you get a bunch of peices you can assemble into something bigger.
The wild world of mashups (note that I didn't get URLs for everything, but that's what Google is for, folks):
- Housing maps - find your next apt via google maps & craigslist housingmaps.com
- Zipcode lookup maps.huge.info/zip.htm
- Routemap for bookmobile deliveries: library route data and google maps api
- Daily Mashup: photos links and news: flickr, del.icio.us & furl
- Newsmap: a nifty visualization of what subjects are being covered where.
- Earthquake data in realtime via yahoo maps & usgs data
- Book carousel: Syndetics book covers and a library's top 20
- Darlene noted that it would be better if it was the last 20 returns, so you could see what was actually available to check out.
- Crime statistics: chicagocrime.org
- Placeopedia - google maps and wikipedia
- Digital life aggregators - there are lots of these, dunno which she showed, and I can't think of any I know of off the top of my head...
- frappr - the blogging librarians! (which I must add myself to at some point).
- Liveplasma: looking for relationships in media content (music, movies, etc.)
- BashR - flickr & wikipedia & del.icio.us
- Weatherbonk: weather data (from diff sources) & web cams
- bookburro.org - Firefox browser extension to find books in stores & libraries (some libraries are doing something similar.)
- I doubt this is the sort of thing Amazon had in mind when they opened up their API - it let's you price compare with competitors.
Facts & Figures
- 1105 mashups as of Oct 24th (via programmable web)
- There are about 2.5 new ones per day
- Maps, photos, and search are big
Typology of the Mashup
- Presentation mashups (and unfortunately, I don't remember what this means).
- Make use of client-side data.
- Have some sort of clientside software.
- Do it via server-side software.
- Make use of server-side data
Where do you start?
- Point, click, cut, paste & publish - there are sites that make building these fairly easy (though I don't remember what any of those sites are, now).
- Cloning similar source edits - build on what other folks have done already.
- Just program it - for those with mad skillz.
Your First Mashup
- Get an idea
- Sign up for a developer token and read the fine print about what you're allowed to do.
- Create your first mashup
Commuity Walk is a good place to start - there are lot of cool things you could do with this, and you don't need a lot of technical knowledge to use it.
- Tech is still in infancy
- Tools still fall short of ideal
- No universal registery for APIs
- There are scaling and dependency issues
- How much to invest?
- Will the data always be free? Will it always be available?
- Dev is quicker with languages like Ruby (Wahoo! Go Ruby - I so want an excuse to learn Ruby....)
- Do you have the right to remix? - Intellectual Property issues
- What's the provenance? - Where'd the data come from, and is that source trustworthy?
Unintended Consequences (both positive and negative)
- Somebody created a mashup of people who are interested in suspicious books via amazon wishlists and yahoo people.
- Okay, that's just scary. Really scary.
- End users can do a lot on their own (whether we want them to or not).
- Client side scripts to modify pages - this can be both good or bad, depending on what the scripts do.
- First round of prizes have been given out, but the contest is still going on.
- We're moving in the open data direction, and we want to encourage people to make use of mashups in the library world.
- Were setting up the innovation directory.
Summer Winner: Google Gadgets - John Blyberg
- Pulls library info onto the Google personal homepage
Another reminder: Competition has been re-opened, and yes, there are still prizes available (including stuff for little innovations).