Fork me on GitHub

When it comes to API discovery, as an industry we haven't been able to find a satisfactory technological solution yet. While I often feel the right approach hasn't emerged yet, I think we are just overlooking "good enough" solutions, because we are waiting for the holy grail of API discovery.

I can't imagine that indexing, search and discovery of the myriad of web APIs out there is that much harder than indexing, search and discovery of the billions of HTML documents available online. Sure, when you are talking about programmatic interfaces, you need a little more precision, but I think us technologists are caught up our own beliefs that APIs should be perfect.

It would make sense that we adopt some evolution of the common sitemaps format, retrofit it to be JSON, accommodate open data catalogs, and allow for various API definitions in Swagger, I/O Docs or even API blueprint. Sure these formats won't have the precision of an evolution of the precious WSDL, or some actually agreed upon standard, but it will get us over the hump we are in.

ProgrammableWeb is 8 years old now, and in 2013, we still don't have any next step, let lone an actual usable solution for API discovery? I just have a hard time believing this is a technological problem, that it is more the stubbornness of the leaders in the space to just take any meaningful step towards API discovery, and lead.

There isn't going to be any money in API discovery, so this isn't something a single startup can emerge to solve. It is something we will all have to discuss and play with until we can find something will get us to the next step, together.




comments powered by Disqus
Google+