Version tested
v0.2.0
Install method
Official release binary
What did you try?
Fresh local install (obsidianlog init, local backend)
Platform
macOS (Apple Silicon / aarch64)
How far did you get?
Completed the init → ingest → query → verify loop
What was confusing, slow, or friction-y?
The first xattr -d com.apple.quarantine run returned "No such xattr" errors for both binaries, which is harmless but looks alarming if you're not familiar with macOS extended attributes. A small note in the docs explaining that this error is safe to ignore (it just means quarantine was already cleared or never set) would reduce confusion for new users. Other than that, the flow was smooth.
Approximate init time (optional)
Near-instant, the interactive prompts completed in under 5 seconds total.
Approximate first-ingest-to-query time (optional)
Under 2 seconds from the curl POST returning to obsidianlog query showing the result.
Would you use this for real? Why or why not?
Yes, for lightweight tamper-evident logging on projects where I don't want the overhead of a full log aggregation stack. The local backend with chain verification is a compelling combo for audit trails on small services. The main thing I'd want before using it in production is clarity on what the verify chain actually covers (just content hashing, or sequence ordering too?) and whether the local storage format is stable across versions.
Anything else?
The verify output (OK tutorial: 1 chunk(s) verified, chain intact) is clean and confidence-inspiring. Would love a --json flag on verify output for piping into CI scripts.
Version tested
v0.2.0
Install method
Official release binary
What did you try?
Fresh local install (obsidianlog init, local backend)
Platform
macOS (Apple Silicon / aarch64)
How far did you get?
Completed the init → ingest → query → verify loop
What was confusing, slow, or friction-y?
The first
xattr -d com.apple.quarantinerun returned "No such xattr" errors for both binaries, which is harmless but looks alarming if you're not familiar with macOS extended attributes. A small note in the docs explaining that this error is safe to ignore (it just means quarantine was already cleared or never set) would reduce confusion for new users. Other than that, the flow was smooth.Approximate init time (optional)
Near-instant, the interactive prompts completed in under 5 seconds total.
Approximate first-ingest-to-query time (optional)
Under 2 seconds from the curl POST returning to obsidianlog query showing the result.
Would you use this for real? Why or why not?
Yes, for lightweight tamper-evident logging on projects where I don't want the overhead of a full log aggregation stack. The local backend with chain verification is a compelling combo for audit trails on small services. The main thing I'd want before using it in production is clarity on what the verify chain actually covers (just content hashing, or sequence ordering too?) and whether the local storage format is stable across versions.
Anything else?
The verify output (OK tutorial: 1 chunk(s) verified, chain intact) is clean and confidence-inspiring. Would love a --json flag on verify output for piping into CI scripts.