Hi, folks! Just as a recommendation to (regular) contributors: Any patches changing the database layout or the way information is stored in the database should be tested by rebuilding and reloading dummy data into the database in addition to usual testing before submitting them to aur-dev: ---- $ cd support/schema/ $ ./gendummydata.py dummy-data.sql && ./reloadtestdb.sh ---- This way, we can ensure that the patch does neither break the schema file nor the dummy data generation script. Maybe we could also make "./gendummydata.py" a bit more modular and run a slimmed down version to avoid delays here. I came up with this cause I just pushed two patches fixing regressions in "aur-schema.sql" and "gendummydata.py" once again. Please don't get the wrong end of the stick: This is no accusation but a simple proposition to improve our workflow. Affected patches are broken because those changes are hard to spot by just skimming through the source code and because I forgot to do these tests before signing off and pushing.