Over the last few weeks at OSCON and the Summit, I’ve had a chance to talk with people about the platform side of the community. There’s a lot of interest and ethusiasm around this topic and hopefully we’ll see some increased activity around XULRunner soon.
In addition to talking about what we could do going forward, people had a lot of useful feedback about why the mozpad effort ran out of steam last year. Hopefully the following thoughts on mozpad will be viewed as constructive feedback and as a guide for any future efforts to build the community around the platform.
- The mozpad project had too many goals. Although many worthwhile projects were identified, the group never prioritized these efforts and instead tried to make progress on everything at the same time. For a new effort, it would probably work better to focus on just one thing first and then expand to other projects after making real progress on that initial project.
- There was not enough effort spent on bringing people into the process. The website was run on a wiki and the meetings were held in public on IRC, but many important XULRunner users never get involved. For a new effort, someone should go to each of the organizations building Mozilla-based applications and talk to them about their needs and explain why working with the community is worth their effort and how they can get involved.
- There was uncertainty about what the rest of the community felt about mozpad and XULRunner. Although there were encouraging words from Mitchell and others, people seemed reluctant to spend too much time without being given a green light to work on things. A green light wasn’t needed since anyone can contribute to an open source project, but I think there was a perception that nothing would end up being accomplished without some official approval. For a new effort, I think this perception can be fixed by clearly documenting how people outside of the core project can collaborate on Mozilla code and showing how Mozilla Corporation employees sometimes have a difficult time getting their changes accepted as well.
- The active mozpad members were too busy. The people coming to the meetings had their own products to ship (Firefox, Komodo, Songbird, AllPeers…) and they didn’t have much time to devote to fixing platform-related issues. For a new effort, we should look into ways to pool resources so that available volunteer time doesn’t become a bottleneck and so that fixing common platform issues doesn’t become the responsibility of any one organization.
davidwboswell 
Summary?
Mozilla had/has the resources to build XULrunner but failed to show the appropriate leadership required to either establish a mozpad-like effort themselves, or help develop one externally.
If Al Billings (AFAIK the most active mozpadder) had the time to travel around the world “bringing people into the project”, who was going to pay for that? Mozilla? Not under Mitchell’s leadership it seemed.
The role of any leader is to inspire, motivate and channel effort. When Mitchell first wrote that Mozilla would not be committing to XUlrunner, Mozpad was doomed then and there. Mitchell may not have thought she was killing Mozpad. Mitchell might have thought she was just making Mozilla’s stance clear. However the real effect was nothing of the sort. Instead there was the vague approach of ‘we are not going to back it, but there is nothing stopping everyone else doing so … such is the beauty of open source … bla blah’. What a load of malarkey. That crowing about open source being infinitely hackable is a delusion. The resources required to build any sort of decent software do not end when potential developers checkout an SVN dump of some open source code base. It is naive of the open source community to promote such fallacies.
MozPad wasn’t a failure. Several of the goals were achieved, even if not so visibly. Firefox includes the XULRunner runtime, WebRunner grew into the wonderfully promising Prism project. Mozilla stepped up support for the platform in a number of ways. The only way to look at it as a failure is if you had high expectations. Based on my low expectations MozPad was a success and the MozPad wiki is still a fairly useful resource documenting a lot of good ideas that we worked on – even if some are not (yet) completed.
Oh – and as you said, it “ran out of steam last year” – partially this is because the most pressing needs were addressed, even if only partially, it was enough to pacify the developer community to some extent. Many of the burning issues were addressed through better understanding and partial compromise on both sides (That is, the “developer community” and the Mozilla organization / employees)