One of the concerns that I had from my brief look at the server implementations on offer the other night was the lack of support for caching mechanisms.
Thinking about Robert's request for example scenarios, it seems sensible that EVERY alto flow starts with a request to the IRD. That is, the application always retrieves it, but it is probably cached locally. Caching is going to be key to the scalable performance of this protocol. Implementing that caching in clients is a reasonable starting point, but what I'm seeing right now is that clients are caching on their own. There is no guidance being given by servers. Doing this means that you can't take advantage of the existing caching infrastructure of the web. It's really quite easy. All you need to do is provide Cache-Control headers. --Martin _______________________________________________ alto mailing list [email protected] https://www.ietf.org/mailman/listinfo/alto
