A fascinating post on Ars Technica on the Twtiter implementation of OAuth. It confirms a concern I had. Since the OPML Editor is open source, and it supports Twitter's implementation of OAuth, I didn't see what good it did to pass around "secrets" in the source code. But I did what I was asked to do. I wasn't watching the developer mail list closely so I didn't see the discussions Ryan Paul mentions. The more I read about what Twitter is doing these days, the more I see the shift away from developers and toward internal development. When people "argue" with me or have a "difference of opinion" about OAuth, they seem to only look at the technical issues. As if every developer could pour infinite resources down every hole. In practice, you only have so many hours in a day, and worse, you have fewer and fewer remaining hours in your life. So you have to choose carefully where you spend your time. There are lots of ways for platforms to fail, and it can be hard to parse when the platform is failing, but the platform vendor is prospering. Sometimes that growth comes from eating the developers, which seems to me very much like what Twitter is doing now. So the OAuth transition will be the first of many hurdles for developers. I choose not to jump over hurdles for platform vendors. I learned a long time ago that the only platform that actually works is one with no vendor. The more the vendor controls the developers, the less I want to be part of the community. At the same time, I was running out of clever ideas for projects to do with Twitter. They haven't added any functionality that I'm interested in in a long time. I was at the point last year where I felt that everything interesting that could be done with Twitter had already been done. That every nook and cranny of their API had been explored. It seems they agree, more or less, because their approach is to try to capture the developer growth from two or three years ago. At the same time, no new territory is opening up. That, and the new controls, say the Twitter coral reef is no longer attracting much new life. With the caveat that that's how it seems to me, and I've been wrong many times before. |