On 12/03/2016 00:11, Marvin Humphrey wrote:
There are more compiler warnings than I'd prefer, but I don't think those should block the release at this point.
On my Linux/i386 setup, the Clownfish build only complains about a couple of unused variables in the CFCPython code and an unused function in the generated callbacks. I don't get any warnings for Lucy.
But I haven't tested on OS X and on 64-bit for a while. So I just built everything on OS X and there are a couple of warnings that I didn't see before, mainly because of redefining "inline" for MSVC in the cmark code. That's my fault and I'll fix that in the 0.5 branch.
I'm also aware that there are many unchecked 64-to-32-bit conversions in the Lucy code after we switched to size_t indices in Vector and Hash. Some compilers complain loudly about that, but it's not trivial to fix. I think it would be best to use size_t indices in the Lucy codebase as well.
LICENSE, NOTICE, CONTRIBUTING.md, INSTALL.md, and README.md all looked good in a spot check -- except for one thing. The copyright notice in NOTICE still reads 2014 and should have been updated to 2016. This isn't the end of the world though, and I'm still willing to vote +1 despite it.
That's a stupid oversight on my part. It's not a big deal to prepare another release candidate, but I'm on vacation the next week. If anyone wants to build and upload an RC2 in the next days, I'd be grateful. But I'm also OK with releasing RC1 as is.
Nick
