02.06.06

alphabet soup

Posted in Technology trends at 6:08 pm by ducky

There has been a lot of buzz in the past few years about Web services: XML, WDSL, REST, SOAP, WDFL, that I just don’t find very exciting. People seem to think that Life Will Be Better when everybody and their sister provides access to their data (aka web services) with {insert your favorite format and transport layer}, and you’ll get interesting combinations of data.

I admit, there is a really good example of combining two different services: Housing Maps is an absolutely brilliant combination of craigslist and Google Maps. However, I don’t think we are going to see a lot of Web-based mashups like that. Housingmaps.com is particularly interesting because there are so few mashups like it!

There is little economic incentive to provide a service that other people can mash up. It costs money to provide data, and there aren’t very many ways to recoup the cost of publishing data:

1. Sell advertisements on the page where you display the data (e.g. Google)
2. Charge for access to the information (e.g. a subscription charge)
3. Reduce the costs of providing your core business (e.g. a SOAP order-taking app is cheaper than a human entering data)
4. Get the taxpayer to pay for it (e.g. NOAA providing weather information)

Very few of these lead to more data becoming available for mashup.

1. If party A provides the data and party B displays it on a web page, party A bears the cost and party B gets the advertising revenue. No incentive for A.
2. If party A charges a subscription and party B displays it, then party B is then effectively giving away access to party A’s information, which is not in party A’s interest (unless the data is so valuable to B that B is willing for A to charge by the access). (Party B gets the advertising revenue, goodwill, and/or publicity, not party A.)
3. There is lots and lots of room for businesses to streamline their transactions with other companies, but little reason for businesses to make that information generally available. Walmart might insist that Pepsi set up a WDFL system for Walmart so they can order sodapop more easily, but I really doubt that Pepsi wants to take orders for sodapop from me. Walmart might insist that Coke set up a system where Walmart can get pricing on Coke products, but I bet Coke doesn’t want Pepsi to be able to access Coke’s prices.
4. There is quite a lot of really good information that governments and non-profits/NGOs provide, I grant you that. The U.S. Census Bureau, for example, has a lot of really juicy data. However, I haven’t noticed goverments making a point of expanding their services lately. Kind of the reverse. Furthermore, even government offices like publicity, so I think they’d still rather you looked at their Web pages instead of someone else’s.

Looking around, I don’t see a whole lot of commercial entities making data available for free. Google Maps and craigslist do, it’s true. (I’m not really sure why.) Google has a search API that they let people use if they register and only for 1000 searches per day. (I suspect that those are “free samples” to get people hooked, and that they are hoping to later get people to pay money for additional searches.)

There is a very small amount of publicly available data in WDSL form. It is small enough that the services available just don’t lend themselves to interesting mashups. I just don’t see value in combining golf handicapping, northern Ireland holiday calendar, and Canadian geocoding services.

Note that even the cool housingmaps.com mashup doesn’t use WDSL or SOAP or any of the acronyms-du-jour. As best as I can tell, housingmaps.com gets the craigslist information by scraping HTML; Google Maps has its own API.

It would be really cool if there were all kinds of web services that were freely available for anybody to mash up. But I’m not holding my breath.

Permalink