"up to date at all costs" is a failed approach

[nia <nia%NetBSD.org@localhost>]

it directly conflicts with pkgsrc's goal of being portable,
and directly conflicts with supporting NetBSD well.

it results in frequent feature regressions.

everyone who spends time reading bulk build reports is burned
out - there's no point working on improving things because the next
major update to some third-party component will steamroll all
the progress made.

people who want stuff up to date don't want to test on ~8
different platforms and it's understandable. everything decays.
pkgsrc's few niches suffer as a result and the whole platform
becomes less attractive to use.

any plans?