Comment 2 for bug 1788256

The debdiff for this upload includes changes e.g. to fonts-noto-color-emoji-0~20180810/emoji_aliases.txt that change existing behavior:

-1f46a;1f468_200d_1f469_200d_1f466 # FAMILY -> family: man, woman, boy

The Makefile has started to pass a -S "small metrics" flag to emoji_builder.py that was not previously in use; and previously the script defaulted to "big metrics".

How do we ascertain that these changes do not regress consumers of these fonts that may be relying on the existing behavior? I don't think the regression potential has been adequately addressed with the current SRU template.