Hi Antonio, > I'll just point out that this number is completely meaningless, for > the simple reason that it counts *pkgnames* and not *pkgbases*. > So, for instance, firefox-developer-edition contributes almost 100 > packages to the total count. Would an accurate count of how many pkgbases are flagged as out of date, charted as how long they've been flagged, be interesting enough to spend the effort creating it? Could the data then be recorded over time so the change can be observed. This might then assuage concerns that ‘Arch is dying!’ or help highlight times when there's a dip in maintenance and decide what should be tried to help. Perhaps the historical data is already present, e.g. flag changes are recorded by date rather than just current state, and so that can be trawled to give us that over-time view today. -- Cheers, Ralph.
