Let me be blunt about this: we already have quite enough on our plates already.
I, for one, have a TODO list that reaches probably 10 years or more ahead. Besides openssh, if you *do* use OpenBSD, contributing helps the project. Speaking for myself, if you do appreciate: - having binary packages that work - having a parallel make that works - having signed packages in the very near future - having kde3 and kde4 built by default then consider making a donation. A *large* part of that work is made possible, and better, by the build machines at Theo. I've been spending *a lot of time* working on parallel builds, and builds on clusters, between make, and dpb. You guys got to realize that actual builds, on real hardware, find *a lot of bugs*. In particular, missing dependencies in Makefiles and variations of these. These bugs are often *highly* timing-dependent. Having fast machines with slow disks, slow machines with fast disks, single processor oldies, multi-processor new things, 32 bit machines, 64 bit machines... every single new combination will exercise the build in a different way, and find new bugs. (yes, there are bugs that are pmap dependent, so that if you reorganize the way you handle memory, which happens "naturally" on some exotic arches, you're ways more likely to meet them) Consider this: we introduced parallel make around 2007. We're still finding concurrency issues in parts of the tree. Likewise, dpb is more than 3 years old. It helped weed out some NFS issues. It helped fix a very nasty race (and stupid) race condition in ld.so/ldconfig (that one was a completely MI error). It still helps finding hidden dependencies... 3 YEARS. Having "interesting" architectures helps finding memory allocation and alignment errors, which tend to be noticed WAAAYS earlier on sparc, but tend to affect everything... I've seen some disparaging comment about our quality process. Well, speaking from a ports tree perspective, we have to shovel after upstream... you have no idea about the smell. Put things in perspective. There are about 20 gigabytes of *compressed source code* in the ports tree. We're doing our best to make it work. But it's nowhere near perfect. Outfits with ways more resources than us like redhat, debian, or google don't manage to do that. How would we ?.. I could go on and on... I'm mostly a developer. I already make the effort to pimp myself and go to conferences to present my work. I'm already donating a huge amount of personal *time* to the project. Now you guys want us to "justify" our work some more and to spend more *time* doing paid consulting work besides the project ? Speaking for myself, do you realize how far behind I am on my todo-list already ? If I get more time that I want to spend on OpenBSD-related stuff, where should my priority be ?.. as for asking for money, well, let's put that in the BSD spirit. You guys do what you want. You're definitely NOT obligated to pay anything. But if Theo has to shut down his rack, you're going to lose *big time* on what's being done within the project. "Gambling" that some other outfit may pick it up is just that, gambling. I wouldn't put my money on it. For one thing, I don't know any other obnoxious paranoid guy who would ride my back as often as Theo does, and force me to write better code. :)
