ownership of parisc-kernel code
PA-RISC Linux development is going well, but some things have changed from a few months ago. We have more developers than ever before, and because of that, it looks like some people are stepping on other people's toes.
I think there's some areas in which the parisc-linux developers could do a bit better, and a lot of it comes from the fact that there's no understood boundaries of who owns what code.
So, I'm going to suggest that I implement the following plan. If people have objections or suggestions, I'd appreciate feedback in the next 24 hours.
What we'd do is document which people are owners of which chunk of the parisc kernel code. In the end, those owners are ultimately responsible for that chunk of code.
How they handle incoming patches for those areas, or handle changes that affect multiple areas is up to them. If they decide that they want to read every little patch that goes through, great. If they want to assign subsections to others, fine. But they are responsible in the end.
Here's my suggested division:
Interruptions and Interrupts: Philipp Rumpf Virtual Memory: Philipp Rumpf Syscalls: Matthew Wilcox IO Peripherals/inventory: Alex deVries Boot code: Helge Deller Everything else: Alex deVries
In the end, this division is up to me, and I'll be happy to re-assign people, settle disputes (*sigh*), etc. So if you have a gripe about a certain section owner not doing their job, I'm the second person you should complain to.
There are some simple things I insist all developers abide to:
- ALL changes need to be documented properly in the CVS comments
- Significant changes discussed on the the parisc-linux mailing list
- All changes *MUST* compile and produce a bootable kernel for the default
configuration, unless there's prior warning[1]
- Be civil on the list. Anyone who commits code does it because they genuinely want parisc-linux to work, we're all in the same boat together.
Also, another note about the mailing list: this list was intended to be used by the developers, and there's no limit of technical depth. If you want to mail patches, mail patches. People who don't want this forwarded to them will just have to deal with it.
[1] By prior warning, I mean a mail to the list that says something like "I'm working on support for the XUAZ SCSI controller, this might break XYZ800 support temporarily". Sometimes you need to take one step back to go two forward.
- Alex
Signature · 3 lines
Alex deVries Vice President of Engineering The Puffin Group