Re: NEW processing friction

[Scott Kitterman <debian@kitterman.com>]

On February 8, 2022 2:38:48 AM UTC, Holger Levsen <holger@layer-acht.org> wrote:
>On Mon, Feb 07, 2022 at 09:28:16PM -0500, Theodore Ts'o wrote:
>> The argument why a package which has an upstream-induced shared
>> library version bump, has to go through the entire NEW gauntlet [...]
>
>I hear your frustration but don't you think that language like "gauntlet"
>makes it, uhm, very hard for the "gauntlet team" to reply, and even more
>importantly, reason with you?
>
>IOW: how can we get to 'no NEW (or a much lighter one) for new binary packages'
>or how can we communicate this if we already have this, maybe also?
>
>'cause I think the latter could very well also be true, or very close
>to it.

He either didn't read the rest of the thread or didn't really care about what was said.  It doesn't really leave an impression that communication is the goal.

To restate: when there's a new binary package, there's several reasons to go through New again, many, if not all except the licensing check, can be automated.

Speaking only for myself, I think that if the tools were up to it (they aren't now), we could probably skip New for packages that aren't empty and don't steal binaries from other sources (when we aren't in freeze - we often catch things from going to Unstable that should really be in Experimental because they aren't targeted at the next Stable release), but with the current state of our tooling that's not a policy that could be implemented (even if there was a consensus among the FTP Masters to do so).

Scott K