Spent some time with DW's codebase tonight; just reading it. Most of it is ok. A bit denser and less commented than I prefer, but hey, that's code. At some point I'll be able to actually move on to the project I want to write: a journal exporter. exor674 and I are planning to collaborate on it. Should be a lot of fun.
I went a ways in the top level of code before I had a OSX 10.6 kernel panic. On reboot, I found a problem with GoogleUpdater that was clogging my system log. We will see if that solves the isse. I've used OSX for nearly 5 years now; first kernel panic.
After the kernel panic and subsequent removal of GoogleUpdater, I dug around and found a perl call graph generator I'd worked on a few months back & hacked on it. Currently it's styled for the code style *I* write, which is very structured and has a distinct form. (somewhat ada-esque I expect) The code also doesn't handle objects, packages, or unnamed subs. So I'll be working to bring it up to date; then possibly using it on the entire DW codebase to map out what statically calls who and where. Obviously dynamic calling is not going to be very analyzable in static analysis.
If anyone is interested in seeing the call graph generator code, I'd be happy to share - it's not up for public purview right now.
I went a ways in the top level of code before I had a OSX 10.6 kernel panic. On reboot, I found a problem with GoogleUpdater that was clogging my system log. We will see if that solves the isse. I've used OSX for nearly 5 years now; first kernel panic.
After the kernel panic and subsequent removal of GoogleUpdater, I dug around and found a perl call graph generator I'd worked on a few months back & hacked on it. Currently it's styled for the code style *I* write, which is very structured and has a distinct form. (somewhat ada-esque I expect) The code also doesn't handle objects, packages, or unnamed subs. So I'll be working to bring it up to date; then possibly using it on the entire DW codebase to map out what statically calls who and where. Obviously dynamic calling is not going to be very analyzable in static analysis.
If anyone is interested in seeing the call graph generator code, I'd be happy to share - it's not up for public purview right now.