I believe in contributing positively to the community, and not just by giving them binaries for free, but by giving them the fruits of my knowledge and skills through source code they can download and extend / modify if need be. It's people helping people, giving someone else a new raw material they can receive and alter as needed. I think these things are very important.
This is all very noble, I suppose, but not something that addresses the
genuine needs of "the community" as it is today. There are vanishingly few competent OS/2 developers any more, and those who remain are unlikely to need or want yet another toolkit or library. Repositories such as Hobbes and Netlabs are littered with this stuff - virtually all of which goes unused.
Rather, what the community at large really needs are well crafted
applications targeted at the end-user (yes, "free binaries"). I'm talking about competently executed, fairly bug-free, attractive, easy to use software
products - not just a mess of code you throw over the fence and hope someone can use and/or fix.
As an example: in another thread, someone suggested porting libmtp. That's all very nice, but why bother if there's nothing that can make use of it? OTOH, porting libmtp as part of a larger project to create an attractive filemanager-like GUI for accessing the contents of your Android phone would be hugely appreciated because there's nothing out there like that right now.
Please understand that I don't want to discourage you from sharing "the fruits of [your] knowledge and skills" - we need all the assistance we can get. It's just that your efforts will bring the greatest benefits to the community and the greatest psychic rewards to you if they're aligned with the real needs of said community.