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

Irregular Earth Rotation and Rationale

• Earth’s rotation is expected to slow over time due to tidal friction, but measurements since the 1960s show periods of speeding up and complex variation.
• Recent work suggests melting ice redistributing water toward the equator will slow rotation again.
• Leap seconds exist to keep UTC aligned with Earth’s rotation (UT1) within about a second, but the behavior is not smoothly predictable.

Algorithmic Determination and Predictability

• Leap seconds are not computed by a closed formula; they’re decided empirically.
• Rough process described: at the start of January/July, measure UT1–UTC; if the difference exceeds a threshold, schedule a leap second for June/December and publish a bulletin.
• This depends on observational data, not pure math, which surprises some who assumed a simple algorithm.

Practical Complexity and Software Bugs

• Leap seconds are rare corner cases, so many systems under-test them.
• Multiple anecdotes: GPS hardware behaving inconsistently (including bricking), MySQL crashes, JVM “losing its mind,” mail systems freezing, large sites suffering CPU livelocks in 2012, and admins having to restart services during leap seconds.
• Problems often involve non-monotonic clocks (time going “backwards”) when adjusting for the extra second.
• Negative leap seconds (short minutes) have been tested in some OSes but are expected to be another source of chaos if ever used.

Time Standards and Alternatives (UTC, TAI, Smear, TZ)

• Some argue computers should use TAI (no leap seconds) internally and convert to human time via a TZ-like database.
• Others prefer UTC with “smearing” (gradually stretching seconds around the event) to avoid abrupt jumps.
• There is debate whether leap seconds should live in the same conceptual layer as time zones (display-only) rather than core timekeeping.

Abolish, Keep, or Replace Leap Seconds?

• One camp: leap seconds are an unnecessary burden for almost everyone; astronomers and specialists can apply offsets themselves.
• Others: small, regular maintenance is better than a disruptive large correction (e.g., leap hour) thousands of years later.
• Proposals include abolishing leap seconds entirely, switching to infrequent leap minutes or hours, or handling drift via occasional timezone shifts.

Long-Term Drift and Human Relevance

• Estimates from the thread: current leap second rate ≈ 1 minute per ~100 years; a leap hour might not be needed for several thousand years.
• Many note daily life already tolerates large offsets from solar noon due to time zones and DST, so sub-minute solar alignment may be overkill.
• Others counter that even if effects are long-term, building systems on ad-hoc political decisions about time (versus physical measures) is risky.