docs: terminal screenshots, plural fix, and a container image - #31
Merged
Conversation
The documentation described the output without ever showing it. Four terminal frames now do, rendered as SVG rather than screenshotted: the assets stay text, so they diff and scale, and regenerating after an output change produces a visibly different file instead of a stale PNG nobody notices. Three are real output captured from a live workspace, with that workspace's identifiers replaced before the transcripts were stored. chat is the exception and is reconstructed from the literal strings in ChatCommand.cs, because the REPL refuses to start without a terminal and cannot be piped to a file. The generator and the transcripts are both committed, so the images are reproducible rather than binaries nobody can rebuild. Capturing real output found a real defect: auth check reported "1 Agents visible". Wording.Count now agrees in number, applied to the four call sites that had the same bug -- including "1 rows" on export and chat, and pack's "question(s)", which is the form you write when you have given up. Added a Dockerfile for the case that actually justifies one: running a Question Pack on a schedule where installing .NET is inconvenient. Non-root, self-contained on runtime-deps, globalization left on so non-ASCII cell values do not silently become question marks. No image is published to a registry -- publishing one is a promise to rebuild it for every base-image CVE, and docs/containers.md says so rather than implying an image exists.
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Screenshots
The docs described the output without ever showing it. Four terminal frames now do — rendered as SVG, not screenshotted, so the assets stay text (they diff, they scale, no binary blobs) and regenerating after an output change produces a visibly different file rather than a stale PNG nobody notices.
Three are real output captured against a live workspace, with that workspace's identifiers replaced before the transcripts were stored.
chatis reconstructed from the literal strings inChatCommand.cs, because the REPL refuses to start without a terminal and cannot be piped to a file.docs/assets/generate.pyand the transcripts are both committed, so the images are reproducible rather than binaries nobody can rebuild.A bug the screenshots found
Capturing real output surfaced
auth checkreporting "1 Agents visible".Wording.Countnow agrees in number, applied to all four call sites with the same bug — including "1 rows" inexport/chatand pack'squestion(s), which is the form you write when you have given up. Four tests cover it.Container image
A
Dockerfilefor the case that actually justifies one: running a Question Pack on a schedule where installing .NET is inconvenient. Non-root (uid 10001), self-contained onruntime-deps, and globalization left enabled so non-ASCII cell values don't silently become question marks.No image is published to a registry. Publishing one is a promise to rebuild for every base-image CVE, and
docs/containers.mdsays so rather than implying an image exists. It also says plainly thatchatcannot work in a container, and that the OAuth broker is absent by design since a container is an unattended context.Gates
183 tests pass, 8 skip, format clean, all links and image paths resolve, no sensitive strings in the tree.