Posted (Sept. 26, 2011, 6:01 a.m.)
Seems reasonable, though it seems odd to have already pushed the change creating the variable and not just include it in this changeset. I was going to comment when I saw these two reviews that they should be merged, but I was too late. The related question that we need to resolve is: what is our naming policy for global constants? Seems like we had this discussion recently but I forgot what the outcome was and can't find anything relevant in my old email. Seems like we have other constants that are mixed-case with underscores (e.g., Full_System). Another argument is that the format of an identifier shouldn't depend on whether it's a const or not, so this should really be fullSystem or isFullSystem or something like that. This makes sense if we think that someday it won't be a constant... e.g., should it really be a per-System variable at some point? The use case of mixed-mode simulations isn't clear, but logically that's the granularity it could go to. Once we resolve the style issue and get the right name here, I'm OK with this.
