From ddafc51fc1fd12e3b975710eb68a783550905ded Mon Sep 17 00:00:00 2001 From: Jeffrey Aven Date: Fri, 20 Nov 2026 16:32:50 +1100 Subject: [PATCH] docs: record the dialog and the read-only install runs What was run after the check for a missing WebView2 runtime went in: - The optimised build from CI, with the loader pointed at an empty folder, showed the app's own dialog and no terminal, and exited with code 1 when the dialog was closed. A click on the link brought Microsoft's page up in the browser. - The app and an engine in a folder the account could not write to, which is what Program Files is to an ordinary account. The engine beside the app was started, its data folder was under the app data directory, and nothing was added to the folder of the app. Two questions for the maintainer came out of the second: whether the provider cache of an installed app belongs in the roaming profile, and whether it should be the one the stackql command shell uses. Co-Authored-By: Claude Fable 5.2 --- CLAUDE.md | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/CLAUDE.md b/CLAUDE.md index 46b9637..dba5314 100644 --- a/CLAUDE.md +++ b/CLAUDE.md @@ -41,7 +41,8 @@ What has and has not been run: - Which engine an installed app starts was checked on Windows with a build of the app and a copy of `stackql.exe` in one folder, the way the MSI lays them out. With nothing set it started the engine beside it. With `STACKQL_BIN` naming another engine, and with it pointing nowhere, it still did, and said the variable was not used. With a path in settings it started that engine, and with a path in settings that pointed nowhere it stopped with the reason. The folder was one an ordinary account can write to, so the fallback of the data folder under `Program Files` was not part of this. The macOS rule has run only as unit tests. - What the Windows executable asks the system for was read from its import tables before and after the C runtime was linked in. Before, it asked for seven libraries of the Universal C runtime. After, for 19 libraries that are all Windows itself. - The page renderer was checked on Windows 11 with a build of the app in a folder of its own. With nothing beside it, the renderer process ran from the system's WebView2. With a copy of that runtime in a `WebView2Runtime` folder beside it, the renderer process ran from that folder and the app came up with its engine. With a `WebView2Runtime` folder that held no runtime, the system's was used. The copy was of the runtime Windows had installed, version 157, not of Microsoft's Fixed Version package, which has the same layout and was not downloaded. -- No machine without a WebView2 runtime was tried, and no freshly installed Windows 11. A machine without one was imitated on Windows 11 by pointing the loader's variable at an empty folder. The optimised build showed Tauri's own dialog, "Could not find the WebView2 Runtime", with Microsoft's download address, and exited when the dialog was closed. A debug build shows no dialog and only panics. Neither writes the app's log, which starts after the window is made. +- No machine without a WebView2 runtime was tried, and no freshly installed Windows 11. A machine without one was imitated on Windows 11 by pointing the loader's variable at an empty folder. The optimised build from CI showed the app's own dialog, the heading and the one sentence each on one line, and no terminal: the process had no console host, where a debug build has one. It exited with code 1 when the dialog was closed. A click on the link, sent to the dialog of a debug build, brought Microsoft's page up in the browser and left the dialog open. Nothing is written to the app's log, which starts after the window is made. Before the check was added, the optimised build showed Tauri's own dialog and then ended in a panic. +- The data folder of an installed app was checked on Windows with the app and a copy of `stackql.exe` in a folder the account could read and run from but not write to, which is what `Program Files` is to an ordinary account. The app was started with that folder as its working folder. It started the engine beside it, the engine's data folder was `/engine`, and nothing was added to the folder of the app. The real `Program Files` was not used, since that needs an administrator, and the note in the settings pane was not looked at in that run. - The token counts were seen in the built app with a scripted model, and for real through the Copilot CLI: one prompt to a probe to see what the CLI sends, and one turn through the app, which showed the count on the steps line and in the log. The counts of Gemini, Anthropic and an OpenAI-compatible server are read as their documentation describes and tested against stubs, not against the real services. - The supervisor's tests see a child end by waiting on a handle to it on Windows, and with `kill -0` on Unix. They used to ask `tasklist`, which on a busy machine could use up the time a test allowed. They passed 30 runs in a row with every core busy. With twice as many busy loops as cores a different test, `an_unexpected_exit_is_reported_with_its_code`, missed its 10 second limit twice in 30 runs. @@ -453,6 +454,8 @@ Raise these with the maintainer when the work reaches them. Do not decide them a - How Workbench versions are pinned to engine releases, and the minimum engine version check at startup. - How a build that ships is released from this repository: a tag and a release whose files the StackQL release fetches, in place of workflow artifacts that expire. +- Where an installed app keeps the engine's provider cache. The MSI installs for the whole machine under `Program Files`, which an ordinary account cannot write to, so the cache goes to `/engine`. On Windows that is in the roaming profile, and a cache of two providers measured 95 MB. The local app data directory is the usual place for a cache. +- Whether an installed app shares the provider cache of the `stackql` command shell the MSI installs, which is `.stackql` in the home folder. Today each has its own, so a provider installed in one is not seen in the other. - Whether `srv` should later co-host the MCP server so the IDE and the agent share one engine session. - Whether loopback pg wire needs TLS with a per-launch client certificate.