Will's blog

Twenty minutes ago or so we released Miro 1.2. I was talking to Chris, Bryan, and John about Miro 1.2 yesterday at lunch (mid-release) because while there was a lot of work done on Miro 1.2, not a whole lot of it is immediately obvious to the typical Miro user. That got me thinking about writing a post that better explains what did happen and why it's important.

The Miro 1.2 release post has a list of things we worked on for Miro 1.2. Most of that list consists of things we did in a week or so. The majority of the release cycle work hours were spent on two items: switching to xulrunner 1.9 on Windows and re-architecting to further separate the "frontend" from the "backend". I want to talk a bit about those two items and why they're important.

Let's start with the xulrunner 1.9 change. Firefox 3 is based on xulrunner 1.9. Switching to xulrunner 1.9 even though it's not released yet was important because the Mozilla crew have done awesome work on improving performance in their current release cycle. Many of the performance improvements are memory-related. It definitely doesn't make Miro the most optimized thing ever, but it helps. Additionally, Nassar (who did the work) spent some time refactoring bits to make sure events were happening in the correct thread of execution and reducing some of the layers of abstraction and indirection involved. This work will make Miro on Windows more stable than it was previously.

The re-architecture work that Ben did is also really important. Previous versions of Miro had a backend and frontend that were tied together. Creating new platforms was arduous and it hampered any efforts towards building a daemonized platform or a platform that talked to MythTV or Elisa.... He made the split between the two much cleaner and at the end wrote a sample command line interface. In the process of doing that work, he did a bunch of other things that affected the entire code base: he fixed the namespace issues we had with Miro Python modules and he did some refactoring.

This opens up a lot of possibilities. It will be easier to write a daemon Miro platform that has an XMLRPC interface. It will be easier to write a slimmed down version of Miro for smaller computers like the Nokia n810. It's a good direction to be heading in.