I am working with an Arabic font and the diacritics for one of the glyphs were badly off. I opened the font in FontForge and saw the glyph had no anchor points like the other glyphs and so I added them. I exported the font and it worked: the diacritic positioning was much better using the anchor points that were, already in the font. I didn't have to create any new lookup entries because they were already there.
I wanted to create some alternatives for the glyph so I created some new unassigned glyphs and gave them names. I copied across the outlines from the successfully modified glyph, saved, and exported the font. However new glyphs I have created are not recognising their anchor points. Why could this be? Apart from the lookup table do anchor points have to be registered elsewhere? Or could it be the new glyphs outside of the Unicode block do not have their anchors recognised? Do I have to include the new glyphs I made in a mark set?
More detail: I am editing SIL's Scheherzade 2.1. The glyph U+FDF2 (ﷲ) did not have anchor points so I added them. I created alternatives for this glyph which do not appear to be in Unicode but are widely used in fonts, namely (للّٰه) and (للَّه), that is [lam + lam + shadda + small-alif + ha] and [lam + lam + shadda + fatha + ha].