Redefine VEGAMAG and OBMAG so astropy can tell them apart

[pllim]

synphot_refactor/synphot/units.py

Lines 52 to 55 in e51d760

	OBMAG = u.def_unit(
	'obmag', u.mag, format={'generic': 'OBMAG', 'console': 'OBMAG'})
	VEGAMAG = u.def_unit(
	'vegamag', u.mag, format={'generic': 'VEGAMAG', 'console': 'VEGAMAG'})

As stated in astropy/astropy#13158 (comment) , the new astropy.modeling.models.Schechter1D takes magnitude as parameter but currently it cannot tell VEGAMAG and OBMAG apart. @mhvk advised the following:

---

[The] problem is that they now are both just aliases to u.mag, so the units system cannot know what to do. I think best might be to treat it a bit like ST and AB, and define units VEGA and OB - which both being units without a known number, and then define VEGAMAG = u.mag(VEGA) and OBMAB = u.mag(OB). Then one can no longer just mix and match those magnitude systems:

>>> from astropy import units as u
>>> VEGA = u.def_unit('VEGA')
>>> OB = u.def_unit('OB')
>>> VEGAMAG = u.mag(VEGA)
>>> OBMAG = u.mag(OB)
>>> (5 * VEGAMAG - 2.5 * VEGAMAG).to(u.one)
<Quantity 0.1>
>>> (5 * VEGAMAG - 2.5 * OBMAG).to(u.one)
UnitConversionError: ("'VEGA / OB' and '' (dimensionless) are not convertible",
'Did you perhaps subtract magnitudes so the unit got lost?')

Note that of course the user can still shoot themselves in the foot by forgetting what band a magnitude is in...

---

To go with this patch, relevant documentation and tests might also need updating. And this might be a breaking change, so should not be in a bugfix release.