The page you meant to cite was live Tuesday. By Thursday it was a 404 and the Internet Archive had not crawled it yet. You do not need a research lab to know that waiting on someone else's crawler schedule is a bad preservation plan.
Wayback from wabarc (~2.3k GitHub stars, GPLv3, Go) is not the Internet Archive's Wayback Machine. It is a self-hostable archiver that pushes URLs to several backends at once: Internet Archive, archive.today, Ghostarchive, IPFS (local node or pinning services), Telegraph, and optional on-disk copies. You can drive it from a shell, a web entry point, or chat bots that feel like IM: Telegram, Discord, Matrix, Mastodon, Slack, IRC, and XMPP.
What it actually does
- Batch CLI —
wayback https://example.comhits multiple targets in one go; flags like--ia,--is,--ip,--ga,--phpick backends. - Daemon modes — run as
-d telegram(or web, discord, matrix, etc.) so a team drops links in chat and gets archive links back. - Sharing — optionally post results to a Telegram channel, Mastodon, or GitHub Issues.
- Privacy paths — Tor snapshots (
--tor), Tor hidden service hosting, local web UI instead of a public bot. - Media — streaming capture with FFmpeg when you need video preserved, not just HTML.
- Ops — Prometheus metrics, config via
wayback.confor env vars documented underdocs/environment.md.
Upstream ships install scripts, apt/rpm repos, Homebrew, Snap, and docker pull wabarc/wayback. Docs: docs.wabarc.eu.org.
Why self-host?
Sensitive URLs stay on your network. Public Telegram bots are convenient; they are also someone else's computer seeing every link your team archives. Your own daemon keeps tokens and logs under your policy.
Pick your backends. Pin to IPFS through Pinata or a local Kubo node, keep Ghostarchive on, turn Telegraph off if you do not trust it for your content class.
Always-on capture. A small Canadian VPS running the web or Matrix daemon catches links from night-shift researchers without laptop sleep killing the job.
Compliance realism. Archiving third-party sites can touch copyright and terms of use. Wayback is a tool; your lawyer still decides what you may store and replay. Respect robots.txt and do not use this to evade paywalls you do not have rights to bypass.
Quick start
curl -fsSL https://get.wabarc.eu.org | sh
wayback https://www.example.org
# Docker + Telegram bot
docker pull wabarc/wayback
docker run -d wabarc/wayback wayback -d telegram -t YOUR-BOT-TOKENFor IPFS pinning, set WAYBACK_SLOT, API key, and secret per their Pinata docs. Run --migrate when the project ships SQL migrations for local playback/search features (check current docs for your release).
Who it's for (and who should skip it)
Good fit: journalists and OSINT folks who want redundant snapshots; communities already living in Matrix or Discord; anyone building a personal "save this before it disappears" workflow.
Skip it: if you only need occasional IA saves (bookmarklet or manual save at web.archive.org); if you cannot maintain bot tokens and container updates; if you expected a full visual replay engine like archive.org's viewer (this tool fans out to external archives).
Hosting in Canada
Wayback is light compared to a media vault: a modest VPS in Beauharnois or Toronto runs the daemon, stores local copies on attached disk, and exposes a TLS-terminated web front on your subdomain. Lock down bot admin interfaces and rotate Telegram/Discord tokens like production secrets.
Talk with Swift Host if you want a Canadian box for always-on archiver bots plus disk for local snapshots. We are in St. Albert and actually answer the phone.
