Two of the most interesting/innovating/exciting open source projects that I've watched in the last little while are "git," the distributed source control management tool, and "Awesome," a very... niche window management program for Linux. When most people think about open source they probably think about something like the GNOME, or Firefox, or to a lesser extent Pidgin. These second group of programs are the big projects that lots of people pat attention to, projects that are targeted at non-developer users.
They also kind of suck.
I mean, I use parts of all of these programs, and I don't mean to diminish what exceptional accomplishments these projects represent, but they're all boring and stale, and I don't think that they're particularly innovative. Read the bottom post on this page for more on this. Has GNOME ever developed a feature that Apple and Microsoft has rushed to catch up with? Has Firefox, (since tabbed browsing) done anything other than play catch-up with Opera and WebKit/Safari/Chrome? And Pidgin is designed to be a copycat.
When I said, "they suck" I guess what I meant is that they're not particularly innovative, they're great pieces of software, but they're not game changing. Now before anyone says, "ha! that's the downfall of open source, lets dig a little deeper into Awesome and git," which I think are different.
While I'd used SCMs before git, and was already a pretty big open source advocate, when I learned about git, and started to really "grok" the program, I said to myself, "this is something that's really revolutionary." Git is a program which stores changes ("revisions") of a group of files (a directory "tree"); the revolutionary part is that it makes merging different and divergent "change histories" almost trivial. With merging histories easy, it becomes easy to have many different development branches. This makes it easy to have people working in parallel on a project, and for developers to experiment without fear of ruining things. Because you can always revert to a previous revision. It also helps that the system is designed to allow people to work offline, and give everyone a full history of the revision. Also it's fast so all of these operations, which can be rather complex, are blindingly fast.
Maybe there's nothing stunningly original in git, but pragmatically git can and has made a huge impact on work-flows, and I think could really impact the way people work--not just programmers but anyone whose doing something in a plain-text format. It's already change the way I write and work, and I'm just using a fraction of what it can do.
And awesome? Again, the idea of a tiling window manager isn't particularly new, there are even a number of different contemporary options, but I can't help but think that given some recent trends among so called "power users" (hot key usage, a renewed interest in command line interfaces,) that there's something sort of innovative and different about Awesome.
What strikes me a innovative about these two programs is that unlike the first group they're all written explicitly for other programmers and ubergeeks, whereas GNOME and Firefox and Pidgin are written for "entry-level users." I'm not discouraging the production of open source software targeted at less technical users, but I think that heretofore no one has been really clear about what "usability" really means. What we've gotten has been software that's designed to operate "as expected," which while often sufficient for usable, is never sufficient for innovative.
So where does innovation happen? In projects that people do for fun and for themselves, because if you write a program, like Awesome, mostly for yourself and if you're friends you don't have to worry about having a UI or configuration interface that's "easy to understand," you're more free to develop innovative features. Also I think innovation can also happen in projects where developers have a budget to be able to afford a close working relationship with usability labs so that the developers and designers can innovate and still end up with an easy/intuitive interface.
Also, I think the web provides the option for some innovation as web-based software is easier to develop rapidly than desktop software, so interface and feature changes are "less expensive" comparatively.
In any case, I think it's important to think about where innovation happens and what in open source serves to encourage or discourage innovation. Thoughts are, as always, very welcome.
Onward and Upward!