Inconsistent Names in `fort.t` File

[mandli]

I noticed while adding another file format that the ascii routines write out nstates in place of ngrids and loops over states rather than patches. This clearly works ok but I think this confuses the role of states and patches in the code but before changing it I wanted to get some feedback as to what was intended with the number of states here. Are there plans to record the states at some point as well as patches for SharpClaw?

[ketch]

No concrete plans. Of course, if we ever use multiple states per patch we'll need to update all the output routines to handle that. It's probably best to change this routine to be consistent with the others as you suggest.

[mandli]

Ok, I will write up a PR.