Skip to content

Present: put the slide show on the other display - #433

Open
NikitkaRa wants to merge 1 commit into
storytold:mainfrom
NikitkaRa:fix/presenter-displays
Open

NikitkaRa wants to merge 1 commit into
storytold:mainfrom
NikitkaRa:fix/presenter-displays

Conversation

@NikitkaRa

Copy link
Copy Markdown
Contributor

Summary

  • When two displays are connected, the slide show goes borderless-fullscreen on the first non-primary display and presenter view opens maximized on the primary.
  • One display (and the browser, and transition previews) keeps the current behaviour: fullscreen on the current screen, presenter view a 1100×700 window.
  • The choice is a pure function of the display list, so it is tested without a second monitor. Swap and a manual monitor choice are still missing. Wayland ignores window positions, so the presenter window may not land on the laptop screen there.

Test plan

  • cargo test -p deckcraft-ui-egui --lib show::tests
  • On a Mac with an external display, start the show (F5): the slides fill the external display and Presenter View fills the laptop screen
  • End the show: the editor window returns to where it was
  • With one display, the show still goes fullscreen on that display and presenter view stays a floating window
  • Transition Preview does not move the window to another display

Made with Cursor

anishguduru pushed a commit to anishguduru/deckcraft that referenced this pull request Oct 16, 2026
…es (storytold#630)

Release builds said "Text recognition isn't installed: its models are
missing" because no packaging script or release job fetched the OCR
models: only `cargo xtask models` does, into the source tree.

- packaging/env.sh gains stage_models DEST, which runs
  `cargo xtask models DEST` (every ATTRIBUTION.toml kind = "model" file,
  SHA-256 verified, with its licence text). No model file is named, so
  the models PRs storytold#433/storytold#572 add ship without packaging changes.
- `cargo xtask models` also writes ATTRIBUTION.txt crediting each model
  (CC-BY-SA-4.0 requires attribution when redistributed).
- macOS: PdfCraft.app/Contents/Resources/models, before signing.
- Windows: models\ beside pdfcraft.exe in the MSI (WiX v5 <Files>
  harvest) and the portable zip; test-msi.ps1 checks the folder.
- Linux deb/rpm/tar.gz/AppImage, Flatpak bundle and FreeBSD tarball:
  share/pdfcraft/models beside bin/. Models::search_dirs now also looks
  in <exe dir>/../share/pdfcraft/models (Models::dirs_beside_exe); a
  test builds every package layout on disk and finds the models in each.
  release.yml and freebsd.yml check the packages contain them.
- The missing-models messages no longer tell end users to run cargo,
  and the 11 catalogs that keyed an old PRINTCRAFT_MODELS wording (or
  the previous one) now key the current English text; ATTRIBUTION
  hashes updated. A test checks the translations match.

The web (wasm) build, which the issue reports, still has no models; this
is the desktop half only.

Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
When two displays are connected the show goes borderless-fullscreen on the first non-primary and presenter view fills the primary. One display stays full screen on the current one.

Co-authored-by: Cursor <cursoragent@cursor.com>
@NikitkaRa
NikitkaRa force-pushed the fix/presenter-displays branch from 603119e to d0081ca Compare October 16, 2026 18:41
@pick2510

Copy link
Copy Markdown
Contributor

Tested on a Mac with three displays. Before testing, I rebased this branch locally onto current main (128cb0b). The two conflicts were trivial: in show.rs, the #844 tests from main and the display tests here go side by side, and in ROADMAP.md both sides' rows stay. After the rebase, cargo test -p deckcraft-ui-egui --lib show::tests passes (19 tests), as do cargo clippy -D warnings for deckcraft-ui-egui and deckcraft, cargo xtask layers and cargo xtask wasm. Happy to push the rebased branch to my fork if that saves you the rebase.

Setup: MacBook (built-in Retina, 1512×982 pt at 2×) on the left; a Dell U2417H set as the macOS main display (menu bar) in the middle; a second U2417H on the right.

Test plan item Result
Two or more displays: slides on one, presenter view on another Presenter view filled the primary display (the middle Dell, below the menu bar), and the show went borderless full screen on the built-in screen, the first non-primary display in winit's order. Both draw correctly, menu bar hidden on the show.
Ending the show: the editor returns In a hands-on run (F5, presenter-view buttons and arrow keys, End Show), the editor came back where it was, on the middle Dell. ⚠️ Twice, when I started the show through the control channel (show.fromStart) and ended it about 4 s later with Escape, the editor window stayed listed at its old position but drew nothing and couldn't be captured. A resize brought it back, one display width to the right (x 240 → 2160). The same steps on main restore the window normally. This may be restore_window sending Fullscreen(false), OuterPosition and InnerSize in one frame while the window is still on the 2× display; applying the position and size a frame after leaving full screen might help.
One display The show goes full screen on the current screen, and presenter view is a floating window on it. ✅
Transition Preview The editor stays on its display. ✅

The main point: which display is which. The PR treats the primary display as the presenter's screen. On a Mac, the primary display is just the one with the menu bar, and with an external monitor that's often the external one. Here, that put presenter view on the big monitor and the slides on the laptop, the opposite of what a presenter wants: the show on the main or projector display, and presenter view on the laptop's own screen.

Suggestion: prefer the built-in display for presenter view when there is one, and show the slides on the primary display if it's external, otherwise on the first external one. Fall back to the current rule (presenter on primary) when no display is built in. On macOS, CGDisplayIsBuiltin tells you, and winit's monitor name for the built-in screen is "Color LCD" or "Built-in …". With three displays, a manual choice (already listed as missing) matters too.

@Harrison-Booth

Harrison-Booth commented Oct 27, 2026 •

Copy link
Copy Markdown
Contributor

Comparison of the four open PRs that change where the slide show and Presenter View go on one or two displays: this one, #1156, #1184 and #1225 (#709 is the closed earlier version of #1156). Nothing here picks one.

#433 (this PR) #1156 #1184 #1225
Question it answers which display gets the show and which gets Presenter View whether Presenter View opens at all both, plus startup backend both, plus a saved choice and Swap Displays
Two displays show on the first non-primary display, Presenter View maximized on the primary unchanged show on another screen (SetMonitor), Presenter View fullscreen on the presenter's screen show on the display chosen in Set Up Slide Show (default: the editor's), Presenter View on top, remembers its window
One display show fullscreen, Presenter View a 1100×700 window show only, no Presenter View show only show fullscreen, Presenter View on top of it
How displays are found winit monitor list wl_output / X11 RandR in a new displays.rs (new wayland-client, x11rb deps); unknown on web, macOS, Windows winit available_monitors, duplicates counted once winit, reported each frame
Extra scope Wayland becomes the default when WAYLAND_DISPLAY is set (DECKCRAFT_X11=1 opts out) Set Up Slide Show ▸ Multiple monitors, Swap Displays, notes under the slides, A-/A+ colours
Size +357/-123, 11 files +95/-1, 5 files +276/-180, 6 files +533/-31, 15 files
New tests 5 2 2 8
Merge state conflicts with main (595 behind): lib.rs, show.rs, ROADMAP.md, 4 docs conflicts: lib.rs conflicts: lib.rs, show.rs clean, 0 behind
CI old run (22 Oct): pass only CodeRabbit gate pass gate pass

Every pair conflicts with every other, mostly in ui-egui/src/lib.rs and show.rs (#433 with #1156: show.rs and ROADMAP.md; #1156 with #1225: lib.rs only; #433 with #1225 and #1156 with #1184 also in apps/deckcraft/src/main.rs). The author of #1156 says in the thread that it will be rebased on #433's display list.

PowerPoint check (real PowerPoint 16.0.20527 through COM, one temporary two-slide presentation, ShowPresenterView = True, SlideShowSettings.Run()): on this machine, which has one display, the only visible PowerPoint window during the show was "PowerPoint Slide Show - Presentation1"; no Presenter View window opened. That matches #1156 and #1184 (one display: slides only) and differs from #433 and #1225, which keep Presenter View on one display. I only tried F5-style Run(), not Alt+F5.

Not checked: PowerPoint with two displays (this machine has one, so Slide Show Monitor, Swap Displays and where PowerPoint places each window were not tested); any of the four PRs built or run; behaviour on Wayland or macOS (the author of a comment here tested #433 on a Mac with three displays after a local rebase). The table comes from the diffs and PR texts at the heads fetched on 27 Oct.

@FredApps FredApps left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

ReviewerA — source-read COMMENT (not an approval)

I read the head of the diff: Present puts the slide show on the other display. The ROADMAP table is rewritten by hand (feature breadth 80% weighted over 191 features with P0 92%, P1 75%, P2 36%; menu bar coverage 59%; component scores for transitions, animations, slide show and review), which are numbers that cargo xtask parity and the project's tooling produce and are better left to it. The feature (choose a screen for the show) is PowerPoint's Set Up Slide Show › Show on. It overlaps deckcraft#1156 and #1184 (counting displays, Presenter View), all three adding display logic, so the maintainers must order them. I did not read the display code. I have not run it, and no exact-head CI result exists for this head. Nothing was run.

I ran nothing and have no exact-head CI result for this PR, so this is not an approval. Where a diff was long I read its code hunks rather than every generated or documentation line.

Status: waiting_author.

This branch has not been deployed

No deployments
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

missing feature The reference application has it and we do not, or only part of it needs decision Needs a product or scope decision by a maintainer

Projects

None yet

Development

Successfully merging this pull request may close these issues.

5 participants