Recently in Services Category

OK, so it isn't Rock 'n Roll. So sue me. But just as rockers can have fun doing what they love, so can library coders. And this Thursday and Friday, that's exactly what a bunch of us will be doing in Seattle -- having fun. We're sorry you couldn't make it, but we have done this kind of thing before and before, and will do so again (a VALA Bootcamp already scheduled and who knows where/when next).

It's two days of library mashup goodness, fueled by our suite of APIs as well as many others from libraryland and beyond. Out of the two days, we hope to do presentations of options for no more than a quarter of the time, then code like mad, and regroup toward the end to show-off what we've accomplished. Like I said, we're sorry you couldn't make it. But if you're nearby, come. We have slots left and we'd be happy to make your acquaintance. As for the rest of you, if you feel some serious mojo seeping out of Seattle, at least you'll know why.

I just finished reading the article "Using the OCLC WorldCat APIs" in Python Magazine by Mark Matienzo (fondly known as "anarchivist" in his online persona), and I have to say he did an excellent job of describing not only how to use our APIs effectively, but also providing a very accurate and thorough historical context. I really don't think even an OCLC employee could have done any better.

But of course the real value of his piece is his "worldcat" Python module and his explanation and demonstration on how to use it. If you have access to the issue (or buy it for $7), there is even a temporary, limited access key you can use to test his code against our services.

This piece is required reading for any Python programmers wishing to take advantage of our APIs, and highly recommended for anyone else wanting to use our APIs, given his straightforward descriptions of the services and his examples on how he mashed them up along with other APIs (such as Google Maps) to create a useful service. This is exactly why we are offering these services, so it warms the cockles of our CPU to see it happening.

[full version of article A Web Services Taxonomy (PDF 84k)]

A Web Service, according to a standard definition, is "a software system designed to support interoperable machine-to-machine interaction over a network." 1

To put it another way, a Web Service is some useful service offered (usually) on the Internet, designed as a sort of building block you can use any way you want.

So, for example, Google Maps, a free service that dynamically draws maps of any location and locates addresses, has been used by thousands of people to build new services such as crime-report maps and real-estate listing tools,

Another way to wrap your mind around Web Services is to consider a range of well-known ones and what they do. That's what the chart below does, with services such as Paypal, Google, Twitter, and Sabre, the airline-reservations system. (click on chart to see full-size):

This chart represents a taxonomy, or classification, of Web Services, constructed by characterizing all services according to two factors:

1. Data quality: from simple/commodity to complex/unique


2. Transaction level: from basic lookup to real-world transaction.

Based on this, I suggest that the Services with the most usage, customer value, and/or revenues typically have more complex/unique data, and/or are more transactional.

See also the above chart in full size, or the full article (PDF 84k).