@jpgilldev / steid

steid/plans/ROADMAP.md
6.3 KBRaw
1# Steid
2
3The source of truth for everything in a developer's or organization's portfolio.
4
5## Vision
6
7Steid is a developer/org identity and portfolio platform. It's where you host your
8code, your writing, your projects — everything you do, in one place, under your
9control.
10
11Git repos are a core feature, not the whole product. Blogs, documentation, project
12showcases, and anything else a dev or org wants to present lives here too.
13
14Personal-first, but organizations can collaborate on it.
15
16**What makes it different from Gitea/Forgejo:** those are GitHub clones scaled down.
17Steid is portfolio-first — the profile page is the product, and repos are one kind of
18thing that appears on it. That framing should win any tie-break.
19
20## Core Principles
21
22- **Fast** — Rust, minimal overhead
23- **Reliable** — correct first, then optimize
24- **Personal-first** — great for a single dev, scales to orgs
25- **Identity-centric** — everything is built around who you are (user/org); auth is
26 designed in from day one, not bolted on
27
28## Stack
29
30Starting intent, not settled decisions. Each one gets a record in
31[decisions/](decisions/) when it's actually made and confirmed in this build.
32
33| Concern | Choice | Decision |
34|---|---|---|
35| Web framework | Topcoat (tokio-rs) | — |
36| Git transport | smart HTTP via `git http-backend` | [0001](decisions/0001-git-over-http-not-ssh.md) |
37| Git auth | personal access tokens over HTTP Basic | [0001](decisions/0001-git-over-http-not-ssh.md) |
38| Git operations | the `git` binary, behind narrow ports | [0006](decisions/0006-git-binary-behind-narrow-ports.md) |
39| Crate layout | single crate | — |
40| Ownership | personal org owns repos | — |
41| URLs | root handles, grouped routes | [0004](decisions/0004-root-handles-grouped-routes.md) |
42| Database | SQLite via sqlx | — |
43
44Architecture and conventions: [architecture.md](architecture.md).
45
46## Milestone Ladder
47
48A from-scratch rebuild on Topcoat. Much of this was built and verified end-to-end in a
49previous attempt — see [progress.md](progress.md). Treat that code as reference, not as
50a baseline.
51
52| # | Milestone | Status |
53|---|---|---|
54| 0 | **Skeleton** — Topcoat app boots, config, one page, SQLite wired | done |
55| 1 | **Identity, thin** — claim on first run, login, session | done |
56| 2 | **Profile page** — `/{handle}` as the real profile | done |
57| 3 | **Repo model** — records + bare repos on disk | done |
58| 4a | **Clone over HTTP** — `git http-backend`, public repos, no auth | done |
59| 4b | **Push and tokens** — PATs over HTTP Basic, push, private clone | done |
60| 5 | **Repo browsing** — tree, blob, commit log | done |
61| 5b | **Deployable by anyone** — binary release, installer, TLS proxy | active |
62| 6 | **Writing** — posts, markdown | not started |
63| 7 | **Identity, full** — multi-user, orgs, invites, registration policy | not started |
64| 8+ | Projects/showcases · issues & PRs · CI/CD · SSH transport · federation | not started |
65
66**CI/CD is a milestone, eventually, and it is 8+.** Push-to-release — a repository that
67builds and publishes itself on push — is the obvious end state for a forge that already
68receives pushes, and it is what "source code integration" means for anyone arriving from
69GitHub. It needs receive-pack hooks, a job model, isolation to run untrusted code, and
70somewhere to keep artefacts, so it is a milestone in its own right rather than a feature.
71Recorded now because it was asked for; deliberately not scheduled.
72
73`/api` is not a milestone. It's a standing practice: every milestone that adds a use
74case exposes it under `/api` where that makes sense. An API isn't a feature, it's a
75second surface on work already being done — and having a third consumer keeps the use
76case layer honest.
77
78### Why this order
79
80The vision calls Steid portfolio-first, but every previous ladder put portfolio behind
81four milestones of git plumbing. Across two attempts and ~6,000 lines, Steid has
82shipped identity, repo records, clone, push, and key auth — and not one portfolio
83feature. Attempt #2 ended with all the git infrastructure working and nothing that
84distinguishes it from Gitea.
85
86The differentiator therefore shipped early: the profile page is milestone 2, not
87milestone 6. Full multi-user identity stays deferred to 7 — personal-first means one
88user, and there is nothing to collaborate on until there is something worth
89collaborating on.
90
91**Git moved ahead of writing after milestone 2 shipped**, deliberately reversing the
92original order. Two reasons. The profile has a Repositories section that a visitor
93expects to be full, and a gitforge whose repo list is permanently empty is its own kind
94of wrong. And git-over-HTTP is the least-understood work left — everything else is CRUD
95over SQLite, while that is pack data streamed through a subprocess — so learning what
96Topcoat makes awkward is worth more now than after three more features are stacked on
97top.
98
99The risk being run knowingly: disappearing into protocol work is what ended both
100previous attempts. Milestone 3 is deliberately small and ships something visible on the
101profile; if milestone 4 starts to sprawl, that is the signal to bank it and go back to
102writing.
103
104**Milestone 5b was inserted rather than renumbered.** Deployment became urgent the
105moment Steid could host its own source: a gitforge nobody can install is a demo. It is
106lettered rather than numbered for the same reason 4a/4b were — decisions
107[0003](decisions/0003-scoped-urls.md) and [0004](decisions/0004-root-handles-grouped-routes.md)
108both refer forward to "Milestone 7" meaning multi-user identity, and renumbering would
109quietly falsify accepted records to save a table row.
110
111**Milestone 4 was split before it started**, acting on that warning rather than waiting
112to be surprised by it. It bundled two subsystems — personal access tokens and the smart
113HTTP protocol — and neither shipped anything until both were done. Split, 4a is
114anonymous clone of a public repository with no authentication code at all, which is a
115working `git clone` in one small milestone; 4b adds tokens, push, and private clone on
116top. If 4b sprawls, 4a has already banked the capability.
117
118**One numbering scheme.** The previous attempt kept two (a roadmap ladder and a
119separate build log with conflicting numbers) and they drifted apart within three
120milestones. If the plan changes, edit this table — don't start a second list.
121
122The active milestone is broken into steps in [current.md](current.md).