I noticed this morning that MS is longer providing full content in their RSS feeds, and I commented on that here. Since then, I was told that the MSDN site supports both mod-gzip and conditional-get - and that they still had bandwidth issues. I was skeptical, to say the least. So, I fired up a VW image and tried the following:
HttpClientModel get: 'http://blogs.msdn.com/MainFeed.aspx'.
I stepped through the request in a debugger, so that I could see exactly what happened. Sure enough, the request asked for compressed content - but did not receive it. I got back a textual response with the feed's contents. As I stepped through, my code cached the necessary information for a future request (i.e., the requisite information for conditional-get). Then I immediately executed the same request. Still not compressed content, but another surprise - the exact same response (i.e., same content) - not a 304. So, color me unimpressed. The problem isn't with RSS, and we don't need some snazzy updated version of nntp. What we need is for someone to provide a cluestick to the people behind msdn blog feeds. To see a large number of people who are unclear on the concept, read the comment stream from Scoble's post