>>37
Why the obsession with being an exact clone of GNU/Linux? That to me is what makes this project unworkable. A public-domain pure POSIX/SUS implementation is an easier (and more definable) goal. The problem is that you don't have a spec to point to when someone submits trash code. And it's unavoidable, because all they can do in the first place is see that "hey it kinda acts like my GNU/Linux install". And I don't even have a GNU/Linux install.
There are also serious quality-control issues with the automatic acceptance of anonymous code patches (prog.git is especially guilty of this). Establish some kind of approval process, even if it's just you responding to each patch and rejecting ones that don't meet expectations or provide a tangible improvement. Eventually you'll need to delegate responsibility to people who understand the details of certain tool domains.
You've pretty much avoided any serious discussion of overarching design issues and technical guidelines. These things
need to be laid out before each individual tool takes its own "whatever's convenient here" approach with the same problems being encountered again and again and in the end none of the code is really even compatible.
Whenever a design question does come up, it really feels like you're just throwing darts at the words on your monitor rather than enlightening us as to the elements of a plan. Stop doing that. If you don't know something, start another thread to invite discussion about it. But however you do it, make sure you come to a conclusion that fits the plan and helps move things forward rather than stagnating and scaring off contributors.
And please stop with the OMG AnonAnon Internet Code Machine Lulz We Gon Take Down GPoLogy hype. Even if you did build a clone of GNU/Linux, the original is moving much too fast to track. There are already respectable, long-running OS projects that have a hard time competing with it on things like basic hardware and application support. What are you going to do differently?