alias-server was a custom implementation of an OP_RETURN spec. It was intended to support prod implementation of the alias spec, and Cashtab was modified to work with this spec.
The pace of infra development on eCash has left alias-server obsolete. It would have been a difficult, bug-prone, and maintenance-heavy approach in the best case scenario. This was originally thought justifiable for the value of the feature and the then-seemingly-low likelihood of better options.
Now though we have numerous better options to choose from. With the success of Agora as a chronik plugin, it is clear that attempting to implement any sort of spec with a custom indexer is the wrong approach. Either modify chronik or create a plugin in chronik.
Agora also presents a ready avenue for selling aliases, provided we work out the spec properly. Before this was always a pipe dream "to figure out later."
Removing this repo and confirming it is the wrong approach is a tough call but it needs to be seen as a sunk cost. The cost of maintaining this repo has already been high, despite no prod use. Aliases are an important feature but the cost of getting it wrong to have them now is much higher than the cost of waiting to get it right.