Stopping Web Services

Yahoo Photos is going away and Jeremy wonders what they should do about the API.

Reading about the pending closure of the Yahoo! Photos and Yahoo! Auctions (in US, at least) services leads to an interesting question for those of us in the business of providing free APIs to our services.

What’s the right way to decommission a Web Service API?

Clearly we ought to strive to shut things down in such a way that breaks people’s code the least (code that we’ve never seen and can’t influence).

There’s some good suggestions and thoughts in the comments.

This is right up the alley of Versioning Web Services in which I pondered the problems of improving an API while not breaking existing clients.

I’m looking for thoughts on how to version a web services API or really any software that’s delivered as a service. How do you prevent changes from having adverse effects while at the same time providing improvements. Is different API versions at different URLs really best or only way to go?

That post garnered some great comments as well.

This isn’t a new problem—at some point all software has to deal with how to end the support cycle for particular versions. However, with web services the problem is complicated because clients are fixed to the upgrade cycle of the server.

For a data structure or service delivered over the public web to a disparate set of clients, this isn’t possible. I cannot time the upgrade of the server to coincide with an upgrade of the client. In many cases, I don’t even know who the client is. I certainly don’t control them. And even if I knew them all, synchronizing an upgrade would require that I coordinate the upgrade schedules of multiple clients, maintained by multiple people, with varying degrees of interest and need.