In the past year, the auto-creation of the browse.json data stopped working twice. Once it was because the code encountered unexpected data. And recently it was because of a change to the structure of the python virtual-environment.
Our goal as maintainers of the project is to be notified of issues before users experience them. Given that, a script that simply loads the browse.json data and checks that the "meta" -> "date_produced" isn't older than one day -- and if it is, notifies us -- should ensure that if something were to happen again, we'd catch it promptly.
In the past year, the auto-creation of the browse.json data stopped working twice. Once it was because the code encountered unexpected data. And recently it was because of a change to the structure of the python virtual-environment.
Our goal as maintainers of the project is to be notified of issues before users experience them. Given that, a script that simply loads the browse.json data and checks that the "meta" -> "date_produced" isn't older than one day -- and if it is, notifies us -- should ensure that if something were to happen again, we'd catch it promptly.