JG> now that we have cu_statusf, should we use that instead of JG> CMSetStatusWithChars even in places where we don't need any fancy JG> formatting? Yes please! JG> It helps to maintain consistency, and (although this is a bit of a JG> minor point), it is a substantially shorter function name. It's not a minor point to me :) JG> Paying more attention to how we free things now that we know JG> virConnectClose is fine when given a NULL, wouldn't the same hold JG> for classname? Yes. Most of the stuff I write free()'s all the dynamic strings on exit (error or not) to make sure we don't skip one due to a broken exit sequence. JG> Optionally, we can go with "not relevant to the patch" and I'll JG> try and clean up this type of thing wherever I can find it as a JG> separate patch. Sounds good to me. -- Dan Smith IBM Linux Technology Center Open Hypervisor Team email: danms@us.ibm.com