Rebuild the README GIFs around normal usage - #46
Merged
Conversation
Two changes, both aimed at what the recordings show. Event source ids are now the domain's own — an ISBN keys a book, a handle keys a member — instead of generated GUIDs. Chronicle takes any string, and it is the difference between a read model table of readable rows and two columns of wrapped hex. The serve and serve-failing modes keep the client connected. A seeder that has exited leaves every observer Disconnected, which looks like a broken system when it is only a stopped process. serve-failing arms the fault before connecting rather than after: Chronicle retries a pending failed partition the moment a client reconnects, so a late arm lets that first retry succeed and clears the failure the clip is about. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
The set was built entirely around a broken store. The hero opened on red crosses and a four-deep stack trace, three observers showed Disconnected because the seeder had exited, and the palette clip led with a failure row. It read as a tool for when everything is on fire rather than a tool for looking at an event store. Three of the four clips now run against a healthy connected system: - demo is it healthy, what happened, what state did that produce - workbench filtering a view, then searching every artifact kind at once - completions tab completion asking the server, using a read-only command - triage the one failure clip: one stopped partition in a healthy store completions no longer types `observers replay` to demonstrate completion. It was never run, but a stray keystroke in a tape stood between a demo and a replay, and completing a read model name shows the same thing without that. record.sh renders the set and asserts the store is in the state a clip needs before rolling. A clip recorded against the wrong state still produces a perfectly good-looking GIF of the wrong thing, which is how two broken clips nearly shipped. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Rendering against a live client exposed two things the README had wrong, both artifacts of a fixture whose seeder had exited rather than facts about the tool. Next# does not sit behind the log tail for observers that skip event types — with a connected client every observer reaches tail+1. It is LastHandled# that differs, because it records the last event each one actually cared about. The note explaining the columns said the opposite. Output-format byte counts are re-measured; shorter identifiers moved them. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
The README GIFs showed a broken system throughout; they now show the CLI in normal use, with one clip about a failure.
Changed
LastHandled#that lags, notNext#.