On 17885 March 1977, Guillem Jover wrote:
I'm not seeing how it is feasible for the Archive team to properly handle the overrides, when they no longer have custody of the NEW processing, which is where overrides were applied. (I guess though, no new/unknown sections and priority values should be let through regardless of them coming as-is from the NEW processing, so perhaps either packages getting rejected or the invalid/unknown values should be getting coerced to known values, say unknown/optional?)
But if the DFSG team would be in charge of the overrides, would they be taking care of updates outside of the NEW processing? What about their definition?
If the intention is for the DFSG team to be in charge of the overrides during NEW, and the Archive team to be in charge of the overrides in other contexts including definition of values for example Priority and Section (and for Sections, their description/meaning), how are discrepancies going to be agreed on and coordinated?
If the intention is really for the Archive team to handle all the overrides, then any such fix will need to happen after the fact, and will require filing a request once the packages have been accepted through NEW, which looks like unnecessary overhead (because fixing the metadata in the source packages is not enough to trigger a change).
I don't think this is a problem worth spending much time on. First, DFSG team. The way NEW processing works, each accept submits the overrides for the package in question. So they can adjust them to whatever they think is best. (Sometimes a maintainer may make a mistake there). Then, changes. They aren't that many, so no need to adjust ways of dealing with them. And last, definition of what exists and can be set - also keep it where it was (ie. archive team). It's not like this is changing all that often anyways. -- bye, Joerg
