Welcome to the new and shiny LLVM Blog. We've noticed that while the greater LLVM community active and thriving, that it is very difficult for people who aren't hacking on it to follow interesting developments. A related problem is that the LLVM web page is really old and crufty, and various bits of it were true of some (long ago) time, but which aren't true anymore. This blog is an attempt to fix this problem, by providing a way for people to percolate interesting news out to the world, news that includes a date on it!
My goal is that this blog will talk about the motivation and design of various subsystems in LLVM, including things that don't fit well in either code or permanent documentation, such as "why" we did things a certain way. Beyond the LLVM optimizer and code generator itself, I hope that this will eventually showcase ways that people are using LLVM, as well as details about what is going on in the various sub-projects of LLVM like Clang, DragonEgg, etc.
A final aspect of this Blog is that we intend to tag posts well and have tag-specific RSS feeds show up on the various subproject web pages. For example, here are some RSS feeds we have so far: [meta] [optimization] [testing] [Clang]. I'd like to have a tag for each subproject as well as (e.g.) a llvm-2.7 tag for features new in the 2.7 release.
One decision so far is that we aren't going to allows comments on posts. We don't have the time or inclination to moderate comment spam, and we already have a great forum for discussions: various mailing lists. If you have comments on a post, ideas for future posts, or (better yet!) would like to contribute to this blog, please bring up your ideas on the relevant mailing lists.