def_T and def_p behavior changed

[chmarti1]

In version v2.2.5, when users configure the unit_temperature or the unit_pressure, it no longer affects how the defaults, def_T and def_p are interpreted. I've been getting questions from users that are already in the habit of updating def_T and def_p when the change units, so it's worth opening an issue to allow discussion.

Before v2.2.5, def_T and def_p were interpreted in the same unit system as all other temperatures and pressures, but this created confusion when users changed the system of units without changing the default temperature and pressure (Issues #68 and #29). This led to requests for a way to specify the "default state" in a way that would not change with the system of units.

In v2.2.5, there are two new configuration parameters - def_T_unit and def_p_unit - which are separate unit strings for the default temperature and the default pressure. Many users will not need to specify them, but for users that like to change the default conditions, this will be important to understand in future releases.

[jranalli]

Just making a note here that I should probably create tests around this issue to document/freeze behavior.