Should old Linux forum threads quietly leave search?

Looking at this as a Linux maintainer, I’m uneasy that every old troubleshooting mistake can become a permanent search result, stripped of its original context. I’d like to see “soft expiration”: after, say, 18 months without activity, a thread stays readable by direct link but is removed from the forum’s search and recommendation feeds. The author could renew its public status when the advice still holds.

Technically, a page can remain accessible while sending `noindex` to search engines, though removal waits for recrawling and may take days or weeks. A clear “historical—possibly stale” banner would matter as much as discoverability controls. This would reduce casual exposure without destroying useful repair history.

It obviously cannot retract quoted copies, screenshots, or archived pages, and stale advice may still spread through private links. Is that a reasonable compromise, or does it make technical history too dependent on one author’s continuing attention?

I like the distinction between deletion and discoverability. The author-renewal part worries me, though: authors disappear, lose access, or may not even remember what they wrote. I’d want renewal to be optional, while moderators could preserve especially valuable threads as maintained references.

The proposal is technically plausible, but the 18-month threshold is policy, not evidence. I’d test several inactivity windows and measure whether people can still find answers through direct links. Also, `noindex` is not an eraser; the recrawl delay needs to be explicit in the interface.

The banner is doing important work here. “Old” and “wrong” are not the same, so I’d avoid a warning that makes every quiet thread look suspect. Maybe show the last review date and let someone mark a reply as still valid without reopening the whole discussion.

Search already has enough ways to lose good answers. I’d prefer a stale label and better sorting before hiding anything. The internet’s memory is messy, but that is not automatically a bug.

This could quietly privilege authors who remain active and penalize abandoned projects whose documentation is still useful. If the platform removes a thread from recommendations but keeps it indexed externally, perhaps that is the safer first experiment: reduce algorithmic resurfacing without making discovery depend on renewal.