This week's progress

Earlier this week I encountered a "show stopping bug": when running with coroutines enabled, the simulator seemed to stop immediately without actually doing much of anything. I figured this out after running the benchmarks and getting pretty much exactly the same result (timing) for all of the queues. This was suspicious to say the least. I mailed Kurt who could reproduce the bug and fixed the bug. I updated my svn rep today and will be running the benchmarks again tomorrow. Some of the things that make my life more interesting:

• My windows machine has an svn repo but can't compile, nor benchmark
• My xubuntu machine can compile but doesn't have an svn repo (it seems to have been lost in transit) and also can't be used for benchmarking because it's a virtual machine
• My fedora machine can compile and benchmark, but is always controlled remotely which makes it inappropriate for development (although I'm normally a vim guy, I must admit I prefer eclipse for java development). Due to ridiculously unreliable WoL I can only access this machine during the weekend.

So you can imagine the juggling I do... I'll probably fix the Xubuntu machine for svn tomorrow, and might even spend some more time on getting WoL working on the fedora machine (although I'm reluctant to spend another afternoon on that problem).