I think with the migration that has happened at Galileo (from Oracle to PostgreSQL) and maybe one or two other institutions following this road, we would have a proven path for dropping Oracle Support altogether.
Though this might sound painful for the current users of Oracle, dropping support does not mean we are not supporting it in terms of you can't run Oracle even if you want to. It is more a "if you run Oracle, you are on your own (with all the other Oracle users)". Well, this is the current situation anyway, so...
I'd suggest to tag the packages once they run oracle with an "oracle-works" or something like this tag. This way we could get a state of a package where Oracle support has been working and if PostgreSQL goes further, then organizations running Oracle will at one stage be able to update the script to PostgreSQL and be happy with it.
