Basic plumbing

I think it is easy for a project to under appreciate the barriers for entry that may be erected by something as simple as SCM choice. It isn't difficult to understand how projects end up with aging plumbing. Given a finite bucket of effort, it will usually be allocated on the most immediate and important concerns.

Speaking for myself, I often won't even look at the source for a project if the SCM system is exotic or makes it difficult to have my own working copy (e.g., cvs, svn). The other major issue that will deter me are "odd" or "custom" build systems. I genuinely don't want to untangle build a system that relies on tcl, m4, and some obscure feature of tcsh. C projects often don't have a CI system or even a test suite -- this makes me concerned that a small patch won't be accepted because of the manual testing burden on a maintainer. There was a time when this state of affairs was the norm but there are now an enormous number of projects that are relative easy to jump into.