JasonTibbitts/ReviewTriageProposal

Random ideas on triaging package review tickets

 * All packages could use a build test (generally on rawhide if it's not broken) and rpmlint run at some point. The submitter is encouraged to provide this but some are not able or do not wish to install mock, and those who have not been sponsored do not have access to scratch builds.
 * Could we get koji access for those willing to triage?
 * Do we want folks doing a lot of scratch builds? Can koji handle it?  Is it sufficiently secure?
 * Review tickets which have not had comments in a while should be pinged.
 * Most old tickets could use a fresh build test under rawhide anyway as there's plenty of potential for old packages to fail to build in rawhide, especially after things like gcc updates.
 * We have existing policy for closing out tickets where the submitter does not respond.
 * The submitters need to be checked against the account database to make sure that they don't need to block the NEEDSPONSOR ticket.


 * NEEDSPONSOR tickets could use some additional handling.
 * We ask new contributors to provide various bits of information which is of use to the sponsors. Nobody does this, so they need to be prompted.
 * Perhaps some use of the keywords field would be useful here to identify packages submitted by the upstream developers (which we really should try to address quickly) or those with other special circumstances.
 * Some periodic report about submitters needing sponsors would be useful.

-