xping v1.5.1

[1.5.1] - 2026-10-06

Added

• A default check file, ~/.xping/checks.toml: xping check and
  xping monitor use it when no file is given, and say how to create it
  when it does not exist. xping check --init creates it with checks that
  work anywhere (your router, 1.1.1.1, 8.8.8.8, DNS, HTTPS); it never
  overwrites an existing file.
• xping monitor HOST HOST:PORT URL… without a file: hosts are pinged,
  HOST:PORT gets a TCP connect, URLs an HTTP check.
• Homebrew: brew install mehaskari/tap/xping (man page and completion
  for bash, zsh and fish included). The tap follows PyPI releases on its
  own, after building and testing each new version on macOS.
• blocklist checks IPv6 addresses too (nibble format) on the lists
  that publish IPv6 data, Spamhaus ZEN and DroneBL — both for an IPv6
  address given directly and for a domain's IPv6 mail and web servers,
  which were previously listed as "not checked".
• xping report: the saved history as one HTML page. An overview of
  every target (state, uptime, outages, latest value), and per target the
  uptime, outages and downtime, latest and median value, a chart of the
  headline metric over time with failures and outages marked, and the
  list of outages. One self-contained file, no scripts, nothing loaded
  from the network, light and dark theme. -o FILE (or - for stdout),
  --since, --last, --title; narrow to one command or target.
• mtr --tcp / --port PORT: every probe is a TCP SYN to that port
  (443 by default), like trace --tcp, so mtr works where ICMP is
  dropped. Also in check files (tcp = true, port).
• --fail-after N and --recover-after N for watch mode (--watch /
  --until-up on ping, tcp, udp, http, health) and xping monitor
  (also per check: fail_after, recover_after). A target counts as
  DOWN only after N failed checks in a row and as UP again after N
  passing ones, so a single blip neither flips the state nor alerts
  anybody. Uptime still counts every check; outages, the "down at the
  end" verdict and alerts follow the confirmed state, and an outage is
  measured from its first failed check. --until-up waits for N passes.
  ping --watch keeps its 3-lost-pings rule as the default.
• xping monitor marks a check ◐ while its latest runs disagree with
  the confirmed state.