Commit Graph
6 Commits
Author SHA1 Message Date
Snow LeeandClaude Fable 5 fafb7fe24d feat: fold the web viewer into the CLI as bdrive web
One binary instead of two: cmd/bdrive-web becomes the `web` subcommand
(same flags and positional folder-or-URL argument). Measured cost of
carrying the webapp in the CLI: +1.4 MB on a ~56 MB binary (~2.4%) —
the cloud SDKs dominate either way. Drops the second goreleaser build
and the separate go install path.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01HHEUaYfFHhmDvqLYw74Ehz
2026-07-07 16:06:26 -07:00
Snow LeeandClaude Fable 5 1ac128db2c rename: sfs -> BearDrive everywhere; CLI becomes bdrive
Product and project are BearDrive; the CLI binary is bdrive (bdrive mnt,
bdrive sync, ...), the web viewer is bdrive-web. All conventions follow
the full name: .beardrive settings file, .beardriveignore, ~/.beardrive
home, BEARDRIVE_HOME, .beardrive-conflict-* / .beardrive-tmp-* files.
Plugin/skill/marketplace renamed to beardrive (/beardrive:mount).

Module path is now github.com/runbear-io/beardrive — merge only after
renaming the GitHub repo, or go install breaks.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01HHEUaYfFHhmDvqLYw74Ehz
2026-07-07 15:02:40 -07:00
Snow LeeandClaude Fable 5 7f098b3ff5 rename: brand the product BearDrive, keep sfs as the CLI
BearDrive is the product and repo brand (open source + future managed
cloud); sfs stays the CLI command, and .sfs/.sfsignore/~/.sfs conventions
are unchanged, so nothing breaks. Brew formula becomes
runbear-io/tap/beardrive (still installs the sfs binary).

Not included here (need the GitHub repo rename first): the Go module
path and repo URLs.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01HHEUaYfFHhmDvqLYw74Ehz
2026-07-07 14:52:32 -07:00
Snow LeeandClaude Fable 5 2d29ff8aac feat: web viewer, project-local config, selective sync, Claude Code plugin
- sfs-web (cmd/sfs-web, internal/webapp): read-only Obsidian-style web UI
  serving a local folder (default) or an sfs remote; markdown rendering
  with [[wikilinks]], task lists and tables, file downloads with ETags,
  per-file provenance from the journals; added to goreleaser builds
- .sfs project file (internal/config): per-folder volume/remote/include
  settings that travel with the folder, win over the global registry, and
  never sync; daemon picks up edits live
- .sfsignore + include lists (internal/syncer): gitignore-style selective
  sync with ! re-includes, applied symmetrically in scan and materialize;
  newly ignored files stop syncing without being deleted anywhere
- Claude Code plugin (plugin/, .claude-plugin/): sfs skill, /sfs:mount and
  /sfs:status commands, turn-boundary sync hooks (blocking pull on prompt,
  async push on stop); installable via the repo's marketplace manifest
- CLAUDE.md and .claude project settings for Claude Code development

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01HHEUaYfFHhmDvqLYw74Ehz
2026-07-07 14:05:36 -07:00
snowandClaude Fable 5 27b3adf86f fix: make daemon and state cache per-mount, not per-volume
One volume can be mounted at several folders (e.g. ./shared in multiple
repos defaulting to the same volume name). Previously the daemon pidfile
and materialization cache were keyed per volume, so mounts fought over a
single daemon and shared one folder-state cache. Key both by a mount ID
(hash of the absolute folder path); blobs, journals, and the lamport
clock stay shared per volume. Content now propagates between co-mounted
folders even offline. Also default remote polling 30s -> 10s.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
2026-06-12 08:46:44 -07:00
snowandClaude Fable 5 2caa09702f sfs v0.1: mountable synced file system for AI agents
- sfs mnt/umnt/sync/status/log/remote/whoami CLI (cobra)
- per-device append-only journals + content-addressed blob store
- deterministic lamport-ordered merge, LWW with conflict-copy preservation
- cloud-agnostic backends: S3, GCS, file:// (S3-compatible via AWS_ENDPOINT_URL)
- offline-first: real files on disk, journal locally, push on reconnect
- background sync daemon per mount with change tracking (device/author/time)
- macOS + Linux; Homebrew formula + goreleaser release pipeline

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
2026-06-11 23:40:26 -07:00