Sean Whitton dijo [Tue, Aug 29, 2017 at 02:08:37PM -0700]: > Packaging mistakes can be rectified, but rectifying them sucks volunteer > time. The reason why we expect highly trusted contributors to have a > longer track record of packaging updates is to reduce the chance they'll > make mistakes which cause work for other people. > > Becoming someone who doesn't make common packaging mistakes just takes > time and lots of uploads. Not having that ability doesn't reflect on > someone's general technical ability. Umh... /methinks that this volunteer time sucking will happen regardless of whether the person in question is a DM or a DD. The number of botched uploads a DD can make is _usually_ n+1 the botched uploads the same person would do being a DM (that is, a package gets reviewed and access is granted to do unsupervised uploads). Of course, a DD will be able to NMU. However, how often will a newbie DD NMU something they are not familiar with? Or, uploading to NEW... NEW gets reviewed no-matter-what, so that's not _so_ different. People tend to get confident over time. I guess that's one of the reasons I have made some of my mistakes: Because of not triple-checking some stuff I would be checking otherwise. Say, I have uploaded to backports something that was lacking quite a bit of dependencies. Silly me. Or, as keyring-maint, I have often made formatting errors in the git log (which we consume for some automated tasks) requiring me to rewrite history. Both are things that Should Not Happen™, but happen nevertheless. I guess that if I were a new DD, I would be more careful. I understand your point and won't argue more about it — But what I am stating is... There is no clear point as to where a person has "done enough" to be trusted to be careful not to botch too much. I can assert the level of care I have observed in this person's interactions is high enough that I trust he won't be a serial upload botcher ;-)
Attachment:
signature.asc
Description: PGP signature