do not skip span for newline textnodes otherwise compute_new_comment_positions() will return bad results for pandoc texts.
from django.contrib import admin
from cm.models import *
admin.site.register(Text)
admin.site.register(Role)
admin.site.register(UserProfile)
admin.site.register(UserRole)
admin.site.register(TextVersion)
admin.site.register(Email)
admin.site.register(Activity)
admin.site.register(Notification)