Eric Snow

> for a benchmark suite, I think we'd want to upgrade these dependencies as infrequently as possible to keep benchmarking results comparable with one another +1 > It's possible there...

It would be worth updating pyperformance though.

We can merge this once the tests are passing. Do you know what the problem is?

Ah, right. I'll take a look at that real quick and then merge it.

+1 IIRC, my intention with tags was for them to be an extension to benchmark groups in the manifest. However, I ran out of time to take it all the...

Another thing we could do with tags/groups is split up a results file based on them.

Is this done now?

Also "list -b TAG" isn't working.

Also, if the "default" group includes a name that does not exist then it is ignored, whether or not "-NAME". I'd expect it to fail like it does for any...

@kmod, the main problem from https://github.com/pyston/python-macrobenchmarks/pull/10 is due to how a custom group named "all" is getting ignored. I'm probably going to just disallow such a group.