No leap second will be introduced at the end of December 2026

Status of leap seconds and negative leap seconds

• Announcement: no leap second will be added at end of 2026; decision windows occur twice a year with ~6‑month notice.
• Earth’s rotation has been slightly fast recently, so accumulated offset doesn’t justify a leap second.
• A negative leap second has not yet occurred; some commenters push back on claims it was “skipped out of cowardice.”
• There is broad expectation that leap seconds will be phased out around 2035, with a pause of ~100 years and potential replacement by a “leap hour” much later.

Earth rotation and unpredictability

• Rotation rate is affected by many factors: geology, core flows, atmosphere, ocean currents, melting ice, reservoirs, earthquakes, and human water redistribution.
• These variations are irregular and only predictable to limited horizons, hence leap seconds can’t be scheduled far in advance.
• Plots from IERS and Wikipedia are referenced to illustrate recent trends and the lack of need for a negative leap second so far.

Technical and systems impact

• Leap seconds are called a “huge problem” for distributed and high‑reliability systems.
• Common workaround: “leap smearing,” stretching or compressing seconds over ~24 hours so clocks remain monotonic but seconds aren’t uniform.
• This complicates semantics of clocks like CLOCK_MONOTONIC and CLOCK_TAI, especially when smearing is undocumented.
• Many systems (Unix timestamps, most datetime APIs, Temporal) effectively ignore leap seconds and treat all days as 86,400 seconds.
• NTP, GPS, and PTP distribute leap information differently; support for true TAI time is patchy.

Time standards: UTC, TAI, GPS

• UTC is tied to atomic seconds but adjusted with leap seconds to keep mean solar noon within ~1 s of 12:00 at the prime meridian.
• TAI is a continuous atomic timescale; UTC‑TAI = −37 s since 2017‑01‑01.
• GPS time differs from both UTC and TAI but maintains fixed offsets; current UTC‑GPS offset remains −18 s.
• Some argue the need for “solar alignment” is weak given large existing offsets from time zones and the equation of time.

Design and abstraction debates

• Several argue leap seconds belong at a higher abstraction layer (like time zones/DST), not in the global timescale.
• Counter‑arguments: rotation irregularities are global, so the global standard should record them; pushing them into time zones would create even more chaos and inconsistency.
• Leap days vs leap seconds:
  • Leap days are predictable and encode orbital mechanics.
  • Leap seconds handle rotational irregularities and are irregular, making implementation error‑prone.

Calendar and time reform ideas

• Proposals include:
  • 12×30‑day months plus a timeless festival period.
  • 12 months of 5×6‑day weeks, with solstice intercalary days.
  • 13×28‑day months plus 1 “extra” day.
• Historical and alternative systems (French Republican calendar, Julian period, intercalation concept) are referenced.

Cultural, legal, and political tangents

• Thread contains extensive humorous riffs: Time Lords, Douglas Adams/Pratchett references, Superman/jets to change Earth’s rotation, XKCD/SMBC links.
• British legal hacks (e.g., resignation via accepting an “office of profit,” long leases) are compared to the “leap hour” workaround for international time treaties.
• A long tangent discusses UK by‑elections, joke candidates, and parliamentary rituals as another example of elaborate legal fictions.

Attitudes toward importance and future

• Some see leap seconds as vital for “accurate” civil time; others say only niche scientific/space applications truly care and can maintain their own timescales.
• Several argue distributed systems’ fragility around leap seconds is a strong reason to abandon them; others retort robust systems must already handle many forms of time jumps and skew.
• Consensus in the thread leans toward: leap seconds solved a relatively minor human‑facing problem while creating substantial technical complexity, which motivates their planned phase‑out.