Conversation
…radable Selected rows were rendered as one flat badge-text-on-highlight span, so the status bullet (green/yellow/red/muted) lost its color and blended into the row the moment it was selected — the only place that signal mattered least, since it's the one thing that lets you tell installed from upgradable from security-update at a glance down the list. The bullet now keeps its own foreground color (with the row's background still applied to it) while the rest of the row stays uniform, without reintroducing the original problem this design was avoiding: mixing segments that each carry their own ANSI reset with an outer Style.Width() wrap, which only paints background up to the first reset and leaves the rest of the row un-highlighted. Padding to fill the row width is now done on plain text before styling, sidestepping that entirely. Separately: the Upgradable view never showed the held (◆) marker, even for a package that's both upgradable and held, because parseUpgradableOutput built packages straight from "apt list --upgradable" output, which says nothing about hold state. It now cross-references dpkg's own selection state (already fetched for the Installed view) the same way ListInstalled does.
A left-click's row was computed from the click's Y position and handed straight to list.Select() without checking it was actually a valid item index. bubbles/list doesn't validate that itself, so an out-of-range cursor later panicked inside its own render (a slice-bounds crash in list.Model.populatedView) — reproducible with a single click anywhere on an empty list (search with no results, an empty view), or a few clicks below the last row of a short one, matching exactly what was reported live: 'basta cliccare sulle voci in fondo, bastano 3 click, su una pagina vuota anche meno.' Fix is a bounds check before calling Select; added a regression test covering both the empty-list and short-list-click-past-the-end cases.
A plain TERM=xterm (no -256color suffix, and no COLORTERM) gets detected as a 16-color terminal, which downsamples every theme's 256-color codes to the nearest basic ANSI color — several can end up looking similar or low-contrast. Reported live from a minimal Docker/SSH session; not a pkgtui bug (forcing a wider palette on a terminal that actually can't show it would just produce different garbage), but worth documenting since it's an easy trap to fall into outside a real desktop terminal.
pkgtui now installs a .desktop launcher (opening in a terminal, since it's a TUI) and an icon, so it shows up in a desktop environment's app menu after installing the .deb or .snap; a headless/server install just never looks at either file, no detection logic needed on our end. - assets/icon.svg: a small isometric-box mark (256x256, scales cleanly down to launcher sizes), used as-is for the hicolor icon theme. - assets/pkgtui.desktop: the launcher entry, passes desktop-file-validate. - .goreleaser.yaml: installs both into the .deb via nfpm's contents list (/usr/share/applications, /usr/share/icons/hicolor/scalable/apps), plus a postinstall/postremove pair (scripts/deb/) that refreshes the desktop and icon caches so the launcher entry appears without a re-login. - snap/snapcraft.yaml: adds the top-level icon field and the app's desktop field. The go plugin only installs the built binary, so an override-build step copies the .desktop file into the part's install directory — snapcraft's desktop-file generation reads it from there, not from the source tree, and errors out otherwise. Verified for real, not just by reading the config: built both packages locally (goreleaser snapshot + snapcraft pack --destructive-mode), installed each, and confirmed the launcher entry and icon land exactly where the desktop/snapd tooling expects — /usr/share/applications + hicolor for the deb, meta/gui + snapd's own desktop registration for the snap — then removed both test installs and all build artifacts.
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