Drops the install/distribution machinery in favor of two simpler paths:
- Local: download a tool .html from website/releases/ and open it.
- Server: run zddc-server, which embeds current-stable HTMLs at compile
time via //go:embed (delivered in the next commit).
Removed:
- bootstrap/install.sh, bootstrap/README.md, bootstrap/level{1,2}.html.tmpl
- website/bootstrap/{level1,track-stable,track-beta,track-alpha}/
- website/install.sh symlink
- The bootstrap-stub generation in top-level build.sh
Rewritten:
- website/index.html "Install on your server" section: two cards
(server / local) replacing the old four script-based snippets.
- website/reference.html, website/css/style.css: updated to match.
- website/releases/index.html: regenerated by build.sh from filesystem.
Top-level build.sh:
- All five tool HTMLs now copied into zddc/dist/web/ and into the
new zddc/internal/apps/embedded/ for //go:embed.
- Cross-compile is always containerized via podman or docker against
docker.io/golang:1.24-alpine (the same image the helm prod chart
uses), with named volumes for module + build cache. ZDDC_GO_BUILD_IMAGE
overrides for air-gapped or pinned-mirror deployments.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
bootstrap/install.sh is a POSIX-sh script that handles all three ZDDC
deployment patterns plus both target shapes in a single command. The
homepage's "Install on your server" section now prints `sh -c
"$(curl -fsSL https://zddc.varasys.io/install.sh)"` invocations instead
of inlining four separate cURL loops.
Modes / channels / target:
--mode copy (default) fetch actual HTML files locally; site is
self-contained after install.
--mode track install level-2 stubs that fetch the channel
from upstream on every page load. Only valid
with channel names (stable/beta/alpha); for
pinned versions, use --mode copy.
--channel stable (default)
--channel beta | alpha channel mirrors (auto-update upstream)
--channel 0.0.2 exact version
--channel 0.0 latest patch within 0.0.x (symlink-resolved)
--channel 0 latest within 0.x (symlink-resolved)
Optional leading 'v' is accepted.
--target root 5 tool HTMLs (incl. landing as index.html)
plus _template/ directory of level-1 stubs
--target project 4 level-1 stubs that fetch ../<tool>.html
--target auto (default) detect from CWD: 'project' if parent has a
ZDDC-looking index.html, 'root' otherwise
--source URL Override the upstream base URL (default:
https://zddc.varasys.io). Use this to point
at a private dev server during alpha work;
install.sh sed-rewrites the embedded source
inside fetched track-channel stubs so they
use your URL.
Also:
- website/index.html — replaces the four inlined install-card snippets
with four use-case-oriented cards that all invoke the same script
(self-contained / track-channel / pin-to-version / project-subdir).
- website/index.html — fixes the stale "pre-built image at codeberg.org/
varasys/zddc-server (channel-tagged :stable, :beta, :alpha)" reference;
zddc-server is now distributed as Codeberg release-asset binaries with
helm chart examples in the repo.
- website/install.sh — symlink to ../bootstrap/install.sh so the upstream
serves the script at https://zddc.varasys.io/install.sh.
Verified end-to-end against a file:// source: copy:root, copy:project,
track:root with custom --source (rewrite of base URL inside fetched
stub), and auto-detection of project vs root from CWD's parent
index.html all behave correctly.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
The "Install on your server" section of the home page now prints four
short shell snippets — copy-paste into a terminal, files land in CWD.
Each uses curl to fetch the relevant bootstrap files; nothing else to
install:
1. Self-contained: fetches the 5 current-stable tool HTMLs into CWD
plus a _template/ directory of level-1 stubs.
~1.8 MB on disk; no runtime dependency on the
site after install.
2. Track stable: fetches 5 tiny level-2 stubs (~10 KB total)
that fetch zddc.varasys.io's stable channel
on every page load.
3. Track beta: same, for beta.
4. Track alpha: same, for alpha.
Each snippet card explains when/why to use that option directly inline.
Implementation:
- build.sh now produces website/bootstrap/level1/<tool>.html and
website/bootstrap/track-{alpha,beta,stable}/<tool>.html as
standalone files (rather than packaging them into zips).
- install.zip and track-{alpha,beta,stable}.zip are removed; the
snippets curl the per-channel stubs directly.
- Docs updated: README, ARCHITECTURE, CLAUDE, AGENTS, bootstrap/README,
zddc/README, landing/build.sh comment.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Three improvements bundled because they all ship as zddc-server v0.0.2:
* /.admin/ debug dashboard with /whoami, /config, /logs sub-routes.
Authorization via a top-level `admins:` glob list in <ZDDC_ROOT>/.zddc
(root-only — subdir entries deliberately ignored to prevent privilege
escalation via subtree write access). Non-admin requests get 404 so the
page is invisible. Recent logs surface via a 500-entry slog ring buffer
teed off the existing TextHandler. Lets operators debug without
kubectl exec.
* Default ZDDC_EMAIL_HEADER changes from `X-Email` to
`X-Auth-Request-Email` — the oauth2-proxy / nginx auth-request
convention that the TND helm chart already sets explicitly.
Operators who set the env var explicitly are unaffected; deployments
relying on the previous default need to set ZDDC_EMAIL_HEADER=X-Email
or update their proxy.
* dispatch() rejects any URL whose segments contain a dot prefix other
than the recognized virtual prefixes (.admin, cfg.IndexPath /
.archive). Matches the existing listing-pipeline filter so hidden
subtrees on the served PVC (e.g. /srv/.devshell — used by the
in-cluster dev-shell for persistent home-dir state) become
unreachable via direct HTTP fetch, not just hidden in listings.
Refreshes the X-Email reference in website/index.html accordingly.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Three things on the public website:
1) Cut alpha and beta channel builds for all five tools, so each tool
now has stable + beta + alpha actually published — previously
beta and alpha were vapor for archive (which had been freshened
earlier) and missing entirely for the others. The intro page's
tool cards now point at real artifacts on every channel.
2) New website/releases/index.html — a generated index of every
version + channel of every tool, with stable/beta/alpha pill
links per tool and a "Pin to version" row of every concrete
v0.0.X build. Regenerated by build.sh's new build_releases_index
function (reads the filesystem so it is always consistent with
what is actually under releases/). Linked from the intro page nav
(Releases), from the bottom of the Try the tools section
("Browse all versions"), and from the Learn more list.
reference.html's nav gets the same Releases link.
3) Folded website/zddc-server.html into website/index.html as a new
inline section ("zddc-server (optional)") below the tool cards.
The earlier separate page is removed; the broken Server nav link
that pointed at it is gone too. The new section leads with the
dual-mode insight (the tools work locally on a folder OR via any
web server, including the optional zddc-server) and frames
zddc-server as a small Go binary that adds things a generic web
server cannot: ACL via .zddc files, virtual .archive URL space,
per-request access logging, mundane glue. The What is it?
paragraph also mentions the dual-mode story up front so users
reading top-to-bottom get the framing before they hit the cards.
Also caught two stale _latest.html refs missed by the earlier
rename sweep: 8 tool links in reference.html and a comment line in
CLAUDE.md. Verified with a full link audit — every relative href in
index.html, reference.html, and releases/index.html now resolves to
an existing file under website/.
ARCHITECTURE.md doc-ownership table updated: zddc-server.html row
removed; new row added for the regenerated releases/index.html.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
The 'latest' label for the current-stable channel was inconsistent
with the channel set we use elsewhere (alpha / beta / stable). Rename
to 'stable' so URLs, file names, zip names, and image tags all line
up with the channel terminology used in the bootstrap, AGENTS.md
discipline rules, and chart consumers.
File / artifact renames
- website/releases/<tool>_latest.html → <tool>_stable.html (5 files)
- website/track-latest.zip → track-stable.zip
- shared/build-lib.sh: promote_release writes/refreshes _stable.html
- bootstrap/level{1,2}.html.tmpl: channels map drops 'latest', keeps
'stable' as the canonical name. ?v=stable is now the explicit way
to switch to current-stable for one request (alongside ?v=alpha,
?v=beta, and ?v=X.Y.Z).
- build.sh: install.zip sources from <tool>_stable.html; emits
track-stable.zip instead of track-latest.zip.
Container image (.woodpecker.yml rewritten)
- Tag publishing now cascades:
zddc-server-vX.Y.Z → :X.Y.Z, :stable, :beta, :alpha, :latest
zddc-server-vX.Y.Z-beta.N → :X.Y.Z-beta.N, :beta, :alpha
zddc-server-vX.Y.Z-alpha.N → :X.Y.Z-alpha.N, :alpha
- :stable, :beta, :alpha are now first-class channel pointers; chart
consumers (e.g. tnd-zddc-chart) can FROM :beta for dev and FROM
:stable for prod.
- :latest kept as an alias for :stable per Docker convention.
Documentation sweep
- AGENTS.md, ARCHITECTURE.md, CLAUDE.md, README.md
- bootstrap/README.md, zddc/README.md
- website/index.html, website/zddc-server.html
- transmittal/template.html, transmittal/README.md
all updated to reference _stable.html / track-stable.zip / the
'stable' channel name. ARCHITECTURE.md's manual freshen example
points at ./freshen-channel instead of the old git-checkout snippet.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
The intro page's "zddc-server" link previously pointed at a Codeberg
blob URL (which uses /src/branch/main/, not GitHub's /blob/main/, so
the link 404'd anyway). Replace with a hand-edited concept page on
the website itself.
The page is structured around two access modes:
- Local directory mode — open a tool, point it at a folder, work
via the File System Access API. No upload, no server.
- Online mode — take that same local directory and put it behind
any web server (nginx, Caddy, Apache, even python -m http.server).
The Archive Browser tool works against the server's directory
listings the same way it works against a local folder.
zddc-server is then introduced as a Go binary that gives you online
mode out of the box, plus four conveniences a generic web server
can't: ACL via .zddc YAML files (gated on email-header trust),
virtual /.archive/ URL space, per-request access logging, and the
mundane glue (TLS, ETags, conditional GET, CORS).
Closing section: the on-disk layout is the same in both modes — the
server doesn't transform the archive, it serves it. Stop the server
and the directory is still a valid ZDDC archive. The "Zero Day"
promise: server is convenience, not lock-in.
Also:
- Add Server nav link to website/index.html and reference.html.
- Fix the bootstrap/README.md link that used GitHub's /blob/main/
pattern (Codeberg uses /src/branch/main/).
- Update ARCHITECTURE.md doc-ownership table: new row for the concept
page, clarify that zddc/README.md is the operations reference.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
ZDDC — Zero Day Document Control. A file-naming convention plus five
single-file HTML tools (archive, transmittal, classifier, mdedit,
landing) and an optional Go HTTP server (zddc-server) with ACL and a
virtual archive index. Self-contained, offline-capable, dependency-free.
See README.md for an overview, AGENTS.md and ARCHITECTURE.md for the
build/release/architecture detail, bootstrap/README.md for the
two-level deployment install pattern, and zddc/README.md for the
HTTP server.
