Invalid xml:ids are appearing on elements in lowfat.
For example:
MAT 1:1
or
Words within those appear to have valid xml:ids.
The issue being xml:ids can only start with a name start character, which excludes all digits.
Acts 1:1
Checking with BaseX, there are 98,248 invalid xml:ids in the lowfat database. For some reason the n that prepends the xml:ids on words isn't being used on wg.
Invalid xml:ids are appearing on elements in lowfat.
For example: MAT 1:1