Researching 166 APIs for My Development Team
01 Nov 2011
I try to pay attention to how APIs are transforming as many industries as I can. I usually carve out at least an hour each day, to evalute a specific industry, the APIs in it, trying to understand what each API delivers and how the industry works, while also identifying where the opportunities lie.
Right now I am looking at two areas or industries of APIs:
Even though I want to be familiar with both these areas, and will write stories on everything I find, my primary goal is to provide the Mimeo development team with as much information, about the best document and photo storage APIs available for integrating into our systems.
There are 46 storage APIs and 120 photo APIs in the ProgrammableWeb directory. I went through all of the APIs and pulled out the APIs that were RESTful, had potential for print integration and were popular or were a stable, quality service. I found 19 storage APIs and 12 photo APIs I felt were a fit.
After looking at 166 APIs and pulling out 31 APIs that offered value, you walk away with a pretty good sense of what makes a good API deployment. Anyone who is launching an API should spend time looking at at least 50 other APIs before building their own. API areas should simple, straight the point and employ the essential API building blocks.
I’d say the biggest problem with each API is: What the hell does this API do?
Every API should put a really simple explanation of what it does in 150 words or less. Pretend you just have one tweet to use. What value does your API deliver? Make it count!
There are so many other things I consider when evaluating an API, these are a few of my immediate concerns while trying to quickly assess a lot of APIs. Think about it, I’m only dealing with 166 APIs out of 4000 APIs. What will this look like when we have 100K + APIs?