Use of GUC to determine default (sphere/spheroid) calculation for geography

[robe]

I think we had discussed the use of defining GUCs at least for some other PostGIS thing since PostgreSQL allows modules to load custom GUCs. Forget what we were thinking for before to use it for, and forget why we decided against it.

It would seem that that would be the ideal way of allowing people to control which measurement behavior they would prefer for the ST_Area, ST_Distance etc. functions and we could default this behavior to sphere as we are doing currently.

This we may need to table till postgis 2.0.

[pramsey]

We decided against it because the GUC can't be set using the "SET" syntax until after the dynamic module is loaded, and that doesn't happen until the first function is called. Could be worked around by providing the functionality via a set_blah() function instead of the SET syntax. Need to decide if we want this for real. I'm indifferent, except insofar as it makes things a little more complex to document.

[pramsey]

I think with function signatures we are fine as things are, I'm closing this one out until someone comes with their hair on fire needing it.

[strk]

See #1393 for a generic ticket about GUC