summaryrefslogtreecommitdiff
path: root/src/detached/wire.zig
Commit message (Collapse)AuthorAge
* views: a client that joins again has its own view back, layout, focus and ↵Gabriel Schneider2026-10-30
| | | | | | the panes opened in it: an Upgrade's clients and one whose bye says it joins again (a divergence's, wire 19's `bye.again`) leave their view kept for them (`views_away`, digested and dumped, so an Upgrade's new build has it; taped `awayView`), the client names it in its next hello (`Hello.view`, handed on by the daemon with the connection) and its join takes it back (taped `rejoinView`) rather than a copy of the session's; the live view is never a kept one, `views.count` counts the others, and a join with every view taken lets one go (every Upgrade made each client's own panes leave its view: the hunt's 80 ms cell lost `+Search` and `+New`) Co-Authored-By: Claude Opus 5.5 <[email protected]>
* replica: an input's time is its client's: it is sent stamped with the ↵Gabriel Schneider2026-10-30
| | | | | | client's guess of the session's clocks (wire 18's `stamp`), the guess its speculation records, and the authority's step reads that, not its own clocks (within a second of them, else its own), so a step's clock read never mispredicts: an edit's second across a turn, a message's millisecond (Ctrl-W's, a jump's) were each a rollback (fixer 1: 4 seen in 2 minutes), and every reapply of one now reads the same time Co-Authored-By: Claude Opus 5.5 <[email protected]>
* remote clients: a pointer's hit carries the place it was made on ↵Gabriel Schneider2026-10-30
| | | | | | (Layer.Place, on the wire, version 17) and a step resolves it from that, never from this process's frame (`surface`, `.local`), which a session never draws for a client: a window's pointer over a tag diverged it at once (the user's `tape byte 119`, 15 rejoins in 25 s) and a click there aborted it; a replay that meets a site its tape lacks, or makes a syscall that is none, is a divergence that goes on harmless (poisonOf), not a panic, and a speculation that makes one waits (Tty's shell lookup, untaped, is a value site now, as Shell's was); a copy over a queue is refused, not asserted; a session's last client, while no one else watches it, leaves it its layout (leaveView, settled as closeView is), which its rejoin copies; a redial waits only after an Upgrade's `refused` (a divergence's redialled a gone session 20 s); a dying session shuts its clients' links before its trace, and an --attach client ends its ssh and socket whatever way it exits; a far join's hashing stops at a blob once its client is let go (its leave waits for it in the step); a confirmed speculation's state is checked against its unit's digest, and the trace gives a key's and a pointer's words; test/daemon.py: 200 far joins level, a window's pointer swept over tags and body, ssh stand-in clients refused and detached, a client of a killed session ending in 2 s; in-process tests for each Co-Authored-By: Claude Opus 5.5 <[email protected]>
* tape: an error crosses by its name, a stat as fs.OsStat, and the tape's ↵Gabriel Schneider2026-10-30
| | | | | | shape is in wire.semantics: a replica of another build read its session's units as its own build's, so a window client of a terminal's session, or of the box's, read every error by a number its build gives another (451 of 505 names numbered otherwise in the Linux terminal's and window's builds, 244 between Linux and FreeBSD: pardes-gui replaying a terminal's tape panicked at `unexpected error code, found error.NonLocalPath`, and so did main's window client of a session on the box at its Look of a fifo) and a FreeBSD stat out of step (nlink 8 bytes there, 4 here); a name this build's set lacks is a divergence (error.NoSuchError), never another error; an anyerror (a host's report) reads back as one of tape.Failure's, a Debug build panicking at one outside it as it is taped; tape.shape (each taped type's fields, tags, enum names and numbers) in semantics refuses, in words, builds whose tapes differ (the stat taped as before: the Linux and FreeBSD builds hash to two), and a site whose result tape.results does not list is a compile error; walk's test of an error by its name Co-Authored-By: Claude Opus 5.5 <[email protected]>
* gui: the window a replica client (pardes-gui --attach=<socket>:<name> and ↵Gabriel Schneider2026-10-30
| | | | | | <host>:<name>, as the tty's, far through ssh as amort's: its blobs by hash and its stream in zstd) and a window session serving replicas (a terminal's client or a window's): the client feeds its replica what its window does, read over its own view (views.within) and applied after, draws its view between steps, takes the stream on its own reader and rejoins as the tty's does (replica.Link, the client's half shared with it, the far path's Have with it); the session takes clients at its step boundary, one person's input a wait, a client's view no window of its own, its font and opacity answers taped; a client says its frontend first (Event.frontend, in wire 16 beside the far word) and the tty and window builds step alike (one desktop semantics: each has the other's words, which a view takes by its own frontend, builtins.registry.here); a check asked 300 ms after the last input and check, not per wake; test/daemon.py --window-clients Co-Authored-By: Claude Opus 5.5 <[email protected]>
* replica: a far client's content cache and zstd: the session lists its join's ↵Gabriel Schneider2026-10-30
| | | | | | blobs by BLAKE3 hash first (wire 16 `blobs`), the client finds each in the core it leaves or on disk ($XDG_CACHE_HOME/pardes/blobs, hashed again on load, a bad file a miss, the least recently used evicted past 1 GiB) and asks for the rest (`want`), and everything after the welcome goes in zstd envelopes (level 6 over 8 MiB for the join, an undo snapshot being a near copy of its text, then level 3), so `ssh -C` leaves the default command; a far client (through ssh, or PARDES_REPLICA_FAR) keeps what it was sent and, at its end, what it made; a hangup ends a remote client the normal way; through the box a client holding the blobs is drawn in 2.4 s where main took 7.0 to 11.9, and a first join of a session with undo in 7.1 s where main took 11.9; libzstd 1.5.7 vendored, its allocations on c_heap, a second copy refused by shared-libs-check; tests for the cache's verify-on-load, its eviction and a far join's blobs sent once; test/daemon.py's far clients Co-Authored-By: Claude Opus 5.5 <[email protected]>
* replica clients join again: one that diverged (a digest the session's unit ↵Gabriel Schneider2026-10-30
| | | | | | | | | | | | did not end on) and one the session let go to upgrade dial it again, the same join as the first, and say why in a line over their frame until a key; a client of another build is told "session upgraded to a newer pardes: upgrade this client too" tty.zig `remote` is a loop of joins over one terminal: the link ended, its reader joined, the replica let go, the session dialled again (an upgrade's exec and load, and its daemon's, retried for 20 s), its dump loaded, its view's size sent again. `Remote.notify` draws the line (`noticeLine`, the client's own, never the session's core). The welcome carries the session's semantics: the daemon judged the hello by the hash it last heard, which an upgrade under way may not have said yet. replica.zig `Peers.refuseAll`: before an Upgrade's dump each client is sent `refused` after what it was sent (up to 200 ms) and let go, its view with it. "A newer pardes is installed" is a Host call (`newer_build`) the core makes in its focus step, so replica clients take the line from the tape; the GUI still says it itself. PARDES_TEST_DIVERGE=<n> puts a client's core off the session's once quiet after its n-th unit. test/daemon.py: the clients test has one rejoin and agree; the upgrade test has a client of the old build told to upgrade and one of the build running join again and agree. Co-Authored-By: Claude Opus 5.5 <[email protected]>
* replica clients over a socket: the session streams each its dump and every ↵Gabriel Schneider2026-10-30
| | | | | | | | | | | | unit after it (replica.zig `Peers`, a writer and a reader thread a client, units at each step boundary, a client's input stepped there in its own view), and `pardes --attach=<socket>:<name>` and `--attach=<host>:<name>` (through `ssh -tt -L`) are a client in the tty host, its keys applied at once by its own replica and drawn by its own process wire.zig version 15: welcome (the client's id and view, the dump's length, the session's UTC offset and digest), dump pieces, units (with the authority's clocks), terminal messages both ways, a check. A unit carries one digest, at its end (`unit_digests`). The client: tty.zig `Shell.remote`, `feed` (a terminal's input to the replica and the session); the session's clock guessed from the last unit's (tape.zig `Guess`); a pump woken only by a client's post reads no new time (`mail_only`), and a unit of another's ending where the last ended leaves the speculation as it is (`neutral`); the rollback copy taken while idle (`Replica.ready`); views a frontend here does not draw left undrawn. A lagging client never holds the session: its queue is cut off past 64 MiB (PARDES_REPLICA_LAG_KB). Detach in a client ends that client; a Restore lets every client go. daemon.zig: a replica the daemon hands over is the session's loop's (a note), not a stub thread's. A window session refuses one in words. test/daemon.py `--clients`: two clients (one through the delay relay) type into a detached session and agree with it, one killed outright, one detached, a third typing into a shell. Co-Authored-By: Claude Opus 5.5 <[email protected]>
* the replica socket: the daemon listens on pardes-replica.sock, a stream `ssh ↵Gabriel Schneider2026-10-29
| | | | | | | | | | | | -L` can forward, refuses another build's hello in words and hands the connection to its session, which welcomes it and answers pings The daemon binds `pardes-replica.sock` beside its own (SOCK_STREAM, so ssh can forward it; the SEQPACKET socket that carries fds stays local) once it holds its lock. A replica's first message is a hello (detached/wire.zig, now version 14): the protocol version at offset 0, the client's semantics hash and the session's name. The daemon reads exactly that much and no further, so whatever the client sends next is the session's, keeps up to 16 hellos still on their way (ponytail: one that never comes holds its slot until it hangs up), and refuses in words another protocol, a session nobody has, or another build: every session now says its own hash in its registration (daemon protocol 2), so a mismatch never reaches it. An accepted connection goes to the session over its link (SCM_RIGHTS) and the daemon keeps no copy; the session welcomes it with its pid and name on a thread of its own and answers each ping with a pong, until bye or a hangup. ponytail: that stub is all a session does with a replica until phase 5 proper; it is what times a round trip through the handoff and `ssh -L`. The semantics hash (wire.semantics) is phase 1's undone item: the protocol, the release and commit, the platform, the build options, every dependency the manifest pins by its content hash (grammars, ghostty, the Unicode tables: build.zig's new `deps` option) and the walker codec's assumptions (pointer width, byte order). Not the OS: a Linux client of a FreeBSD session is the main case. ponytail: an unstamped build has no commit, so two dev trees of one release with the same options are not told apart. `pardes --version` prints it on a second line. The hello lost its grid (a replica's view takes a resize like any other input). test/daemon.py: a hello of this build welcomed by the session itself (its pid, one fd more in the session and none in the daemon, and back once it hangs up), 50 pings, another build's hello and an unknown session's refused, and `--ping <socket> <session> <pardes> [n]`, which times a hello and n pings on any replica socket (a forwarded one, say). Co-Authored-By: Claude Opus 5.5 <[email protected]>
* the frame wire goes: detached/server.zig, client.zig and the cell frames of ↵Gabriel Schneider2026-10-29
| | | | | | | | wire.zig Nothing draws a session's screen from frames any more: a session is always an in-process authority with its terminal or window, and remote clients come back as replicas (docs/next-daemon.md, phase 5). server.zig, the third host and its frontend socket, the smallest-common-grid rule and its 32 frontends; client.zig, the frontend half; wire.zig's frames, cell and layer codecs, server messages and the post chain on the wire. wire.zig keeps the framing, the hello and its version (now 13) and the input events, with their tests, for phase 5. Co-Authored-By: Claude Opus 5.5 <[email protected]>
* anim 11: a PDF's strip moves in the text's motion system: the integrator is ↵Gabriel Schneider2026-10-25
| | | | animation.Follow (the Glide's, moved out), here the strip's lag behind document_scroll_y in pixels; whatever moves the core's position (a notch, a spin, PageDown, Ctrl-d/f, ge/gg, a Look at a page, Back, the rail) kicks it, a touchpad's fraction and a relayout move it directly; the pages a glide crosses render in its input's frame, a far jump's only where it comes to rest; a slow stream eases onto its mark (it ran into it at speed and passed it); pictures neither scroll nor pan
* drag 2/3: a pane's or column's grip let go past the window's edge opens it ↵Gabriel Schneider2026-10-24
| | | | in a new window (Handover.zig)
* Search converge 4/6: the disk engine, Grep and Find's listing on a worker ↵Gabriel Schneider2026-10-24
| | | | (rg when there, else a sorted walk), generation-numbered batches, a debounce, a 2000-line cap and More, and the declarations parse off the loop
* Search converge 2/6: a pane's places, every match or listed place tinted and ↵Gabriel Schneider2026-10-24
| | | | ticked on the rail, n/N stepping them silently, shifted by an edit rather than searched again
* Audit: the hosts' wire on std.Io Reader/Writer, Lsp and selection_pipe ↵Gabriel Schneider2026-10-23
| | | | | | | | | | deliver through a sink, std.Io clock and sleep wire.zig reads and writes through std.Io.Reader/Writer instead of its own framing helpers; host_io.Lsp.start and selection_pipe.start take a callback sink, so each shell keeps no copy of the worker plumbing. Co-Authored-By: Claude Opus 5.5 <[email protected]>
* zig fmt for Zig 0.17Gabriel Schneider2026-10-22
|
* Build with Zig 0.17Gabriel Schneider2026-10-22
| | | | | | | | | | | ghostty, libvaxis and uucode move to ghostty's 0.17 pins, cloud9 to its 0.17 main, SDL to castholm's release-3.4.x (3.4.18), zstbi and tree-sitter to forks with 0.17 build files (see build.zig.zon). A bare zig build copies the binaries into ~/.local/bin (-Dhome-install=false leaves it out): 0.17 keeps the prefix from build scripts. A terminal's edited copy goes when the program clears the screen. tinywin.snap is gone; perf baselines are re-recorded. Co-Authored-By: Claude Opus 5.5 <[email protected]>
* A program that asked for focus reports (1004) hears CSI I when its pane ↵Gabriel Schneider2026-10-01
| | | | | | | | | | | takes the keyboard and CSI O when it gives it up, to another pane, a tag or another window The core has a window-focus event now, from SDL's window focus and from the tty's own focus reports (vaxis turns 1004 on outside); the detached wire carries it: version 10. The test shell's ESC]777;mouse;focus;<1|0> fakes the window's. Co-Authored-By: Claude Opus 5.5 <[email protected]>
* A mouse event carries Shift and Alt as well as Ctrl, from SDL and from the ↵Gabriel Schneider2026-10-01
| | | | | | | | | | | tty's input; a program's wheel report carries them as xterm counts them, and Shift keeps the wheel pardes's own scroll The reports are ghostty's encoder now (shift+4, meta+8, ctrl+16), with the mode and format ghostty tracks. The detached wire carries the two new bits: version 9. The test shell's ESC]777;mouse takes an optional fifth field of modifiers. Co-Authored-By: Claude Opus 5.5 <[email protected]>
* shader_build.zig is ShaderBuild.zig: a file that is a struct (its fields are ↵Gabriel Schneider2026-10-01
| | | | | | | | the post chain's builds) takes a type's name The naming the split agreed on: a file with fields is a type, TitleCase, as Messages.zig and Layer.zig are. Its importers bind it as ShaderBuild; comments and docs/render-pipeline.md name the new file. No behaviour changes. Co-Authored-By: Claude Opus 5.5 <[email protected]>
* The theme's chrome is worked out once, not per grapheme; still post passes ↵Gabriel Schneider2026-10-01
| | | | | | | | | | | | | | | | | | | | | | | | | | | | let the GUI rest; a perf gate somtrmsz's contrast floors made ChromeTheme.fromTheme run its focus-tint and separator searches (pow calls each), and Output's RowDecoration.styleAt called it for every grapheme of every highlighted row: the 50k-line file's render went from 0.11 ms to 13 ms. Pardes.bodyChrome now keeps the theme's chrome, worked out again only when the theme differs; recolorSyntax asks for it once a decorated row, and a plain row (nearly every row of a file) never asks. ReleaseFast, medium fixture, median us, before -> after (main): render 13150 -> 76 (111), key-down 14658 -> 79 (110), wheel 13589 -> 76 (109), open 14891 -> 2095 (2055), edit-char 16703 -> 3006 (1251; the rest of that gap is editing and tree-sitter, not this). Post.animating asked for frames whenever the window had focus and any pass was ready, so a still pass kept the GUI drawing at the display's rate (Bloom: 49% of a core idle, in a hidden test window). A pass now says whether it moves on its own: the CRT (its hum and dither) and a Shadertoy file whose source reads iTime, iFrame or iDate (shader_build.readsTime; the flag rides the wire's post message); Bloom, Vignette and Grain are still. Bloom idle: 49% -> 1.2%, as with no pass. zig build perf-gate: the 50k-line file's gestures, each's fastest sample within 3x of the recorded baseline's (test/perf-baseline-<platform>- <optimize>.json, recorded from this build), run with every unit-test. On the regressed code it fails at 180x for render.
* Shader files compile again on save; an attached GUI runs the session's post ↵Gabriel Schneider2026-10-01
| | | | | | | | | | | | | | | | | | | chain shader_build.zig takes the Shadertoy compile out of the GUI's Post: the process that holds the core compiles each chain file behind the prefix on a thread, watches its directory (file_watch's new shader slots), and on a change reads and hashes each file, compiling only one whose bytes moved. A failed compile keeps the last good SPIR-V and is said once; the same bytes are never compiled or reported twice. A file is read with std, not fs.zig's readFile, whose turn hand-off a worker thread does not hold. A detached session compiles for its frontends and sends them the chain (wire post: each pass's scene and level or the file's SPIR-V, and ShaderAnimation) on attach and on every change, so an attached GUI runs the same passes, levels and animation as a local one while still reading no disk and running no program. The attached GUI describes its frames to the chain from the session's chrome and redraws an animating chain on its own.
* G13: Parallax, lapis's dots drift with the text at a quarter of its speedGabriel Schneider2026-10-01
| | | | | | | | | | | | | | | Parallax on|off (off, a pixel shell's; Chrome.fx bit 3). A body region carries its first line (Region.line, a file's scroll or a terminal's grid offset; the wire's region record grows a u32). The GUI offsets each body's page dots by a quarter of its line in rows, and of G6's glide offset while one runs, so the pattern drifts with the text and settles with it; the cell shader reads it from the instance's serial slot, whose per-cell dissolve noise G7 left unused (the dead noise code goes). Tests: a scrolled body tells its line; the shift handed to the shader is a quarter of it within the dots' period. Measured in a capture: four lines move the dots 13 px (140 px / 4 mod 22). Gates as before, goldens byte-identical. Feel review: .scratch/render/parallax/.
* G12: Occlusion, pane bodies darkening faintly toward their edgesGabriel Schneider2026-10-01
| | | | | | | | | | | Occlusion on|off (off, a pixel shell's; Chrome.fx bit 2 — the glow bits field is now `fx`): four soft strips cast from just outside each pane body inward, clipped to the body and round a cursor, black pre-warped to linear light, 2% at most (§9.1) and under the lift's contrast ceiling, in the pane's own group. Tests: the shade stays inside the body, off the cursor, at most 2%. Gates as G11. Feel review: .scratch/render/occlusion/ (with an x40 diff).
* G11: HoverGlow, a soft underline of light under a look-hover's wordGabriel Schneider2026-10-01
| | | | | | | | | | | | | | HoverGlow on|off (off, a pixel shell's; Chrome.glows bit 1, SelectionGlow bit 0 now that there are two): under the word a look-hover would open (its cells in the affordance's tint) a strip three logical pixels tall along its foot, blurred by one, in the theme's accent, at most 22% and under the text's contrast floor, fading in over 80 ms; kept to the bodies and split round a cursor. glowBlocks serves G9 and G11. The strip is cast from a row away so addSoft's caster cut-out misses it. Tests: the glow comes up by 90 ms and never falls on the cursor's cell. Gates as G10 (a loaded machine shifted the GUI goldens a scene once; a rerun matched 01-08). Feel review: .scratch/render/hover/.
* G9: SelectionGlow, a soft halo round the body's selectionGabriel Schneider2026-10-01
| | | | | | | | | | | | | | | SelectionGlow on|off (off; a pixel shell's, Chrome.selection_glow): each block of selected rows in a pane body casts Lift's soft kind in the selection's own colour, the block its caster (never touched) and a cursor beside it joined to the caster; clipped to the bodies (never a tag or a grip); at most 22% (§8.2), less where the text round it would drop below min(its contrast, 4.5); fading in over 100 ms with an ease-out, on the core's continuous path (shell_continuous); none while panes move. Blocks come from the cells in the theme's selection colour (documented). Tests: the halo's colour, its clip to the body, nothing at its first frame and up to its cap after 100 ms, fading done. Gates as G8. Feel review: .scratch/render/selection/.
* G5: lapis, 0x4200.cafe's manuscript, as a theme with its ornamentGabriel Schneider2026-10-01
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | A native theme `lapis` (src/themes/lapis.zig) from web/theme/lapis.css: a lapis-deep page, lapis tags with gold names, vellum text, vermilion rubrics, gold rules and a vermilion tag rule, gold selection in lapis ink. Its ornament is the theme's new `decor` (colors.Decor), drawn by a pixel shell inside the chrome it already has (decision 1), all of it off every focus indicator: - page_dots: the site's gold dot grid (#e8c46a22, 22 px), drawn by the cell shader on page-coloured cells alone (flag 0x10000000), so it lies under the text and never under a selection, a tag or the cursor; fixed to the window, a page the panes lie on. - rail_checker: the dither frame's gold and lapis checker in the scroll track (a decor kind, flag 0x08000000), anchored to its rail; the thumb stays solid gold. - tag plaques (tag_border, tag_shadow, tag_stripe): every tag a raised plaque inside its own band, the site's trail: a 2px gold frame and a hard 5px vermilion shadow down and right, both within the band's rows and columns (down, what the slack under the text leaves), so no shadow reaches a body, a grip or the next pane. The focused pane's plaque is the code window's title bar, gold and lapis stripes of 2px, its words on plates of the focused ground; unfocused ones are plain; column and workspace tags take the object-label language (1px frame, 2px shadow toward the page). Drawn as decor under the cells (a group's under range), with blank tag cells see-through and the cells blended premultiplied (pipeline_over) on such frames; opaque windows only. - title_shadow: the titles' vermilion offset shadow under a file name's letters, sampled in the cell shader inside the letters' own cells (flag 0x08000000), none on a cursor. Chrome carries the decor (decor_*, name_ink, active_name_ink); a theme file folds a nested decor literal field by field. Left out, with why, in docs/effects.md: the fonts, the ornaments, a decor crossfade, and in the tty anything but the colours (a tag row has no spare cell for a shadow). An empty column is the page. The effect id is masked to 27 bits (ui.vert/ui.frag) for the fifth flag. Tests: the cell flags land only on page ground and file-name cells, never a cursor, nothing without decor; lapis's plaques stay inside their tag's band, clear of every grip, only the focused one striped; its checker is in the track; no other theme draws any; the separator and focus tests cover lapis. Goldens: theme and themesel (the ring gains lapis after acme). Gates as G4. Feel review: .scratch/render/lapis/ (lapis.png, lapis-2x.png, tags-focused-vs-unfocused-2x.png, a PanelSlide at 60 fps and a quarter).
* G3: the focused cursor glides a jump on the Motion flavour's springs, and blinksGabriel Schneider2026-10-01
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The core owns the motion, as Lift: animation.CursorGlide keeps the focused cursor's four corners on springs (leading corners at the flavour's pace x2.4, trailing ones x follow, so a jump smears and collapses; a long jump bows by the flavour's arc), sampled at each frame's own time. A step of a cell or less, insert mode, a focus change and any scroll (a file's view, a terminal's output) land at once; a snap lands a glide under way too. The quad is clamped to the focused pane when drawn (never its springs). Chrome carries it on the wire (cursor_glide: bit 1 a cursor, bit 0 gliding; cursor_alpha; cursor_quad, grid cells; cursor_idle_ms): the GUI draws the quad as an overlay in the cursor's colour, at 85% (70% on the long flavours) rising to solid as it lands, and hides the cells' cursors while it does: one or the other every frame. Only the focused cursor glides. Blink (CursorBlink, on by default in pixel shells) is the shell's: from cursor_idle_ms and when the frame arrived, solid while typing and 500 ms after, 530 ms halves with 80 ms eased edges, solid after 10 s idle, and never while the window is unfocused or minimized or on a virtual clock. An edge redraws the last frame with no core render (level B, blink_redraws); an attached GUI blinks from the same field without a frame per edge. Tests: snaps a cell, glides a jump with smear and bow, arrives under a pixel by 150 ms on smooth and lands exactly, shifts with a scroll; blink holds, eases, stops; the core glides a jump (ge) frame by frame (gliding exactly while it moves, alpha rising, exact at rest), snaps a step (j) and a switch to insert mid-glide; a terminal's scrolling output lands its cursor; wire round-trip of the cursor record. Shared files: src/animation.zig, src/pardes.zig, src/config.zig (toggle CursorBlink), src/surface.zig, src/detached/wire.zig.
* acme looks like acme; every theme gets acme's button, scroll column and 2px ↵Gabriel Schneider2026-10-01
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | rules acme takes plan9port's own colours (acme.c iconinit, draw.h): selections #eeee9e in the body and #9eeeee in the tag, black ink; exec and look sweeps #aa0000 and #006600, white ink; a #99994c scrollbar under a #ffffea thumb; black rules; a #8888cc tag/body rule and column button. Every theme: the grip is acme's button, its tag row's full height, in one scroll column with the scrollbar (rail_px, 12 at a 17px tagline, scaled, times GripWidth, a percent, default 150; same x0/x1 to the pixel): focused clean = solid box; focused dirty = box_dirty filled to a box ring (acme's modbutton); unfocused clean = a box_border ring round the tag's ground; unfocused dirty = box_dirty a pixel inside that ring. The grid, which has no ring, marks dirty with a bold * on box_dirty in the grip's second cell and fills an unfocused grip halfway to its ring. Regions carry a dirty flag. Column buttons are the same shape, solid. 2px rules between columns and stacked panes (inside the tag band's slack), none at the window edge; a 1px tag/body rule halfway between tag and page. Theme data: tag_sel_bg, sweep_bg/fg, tag_rule, rule_px, rail_px, box_border, box_dirty. Chrome: rule_px, tag_rule, rail_px, grip_border, grip_focus_ring; a comptime guard on Chrome's layout forces a wire bump once v8 ships. Structural rules are exempt from the focus-indicator rule (docs/effects.md). Tests: acme widths; stacked rule in the slack at two tagline sizes; every grip-column pixel is band, button, ring or rule, and the button shares the scrollbar's x0/x1 (4 metrics x top/bottom x focus x dirty); the four grip states distinct in every native theme with a non-colour dirty cue in the grid; wire round-trip of the new fields. Shared files: src/colors.zig, src/surface.zig, src/body_layer.zig, src/tagline.zig, src/detached/wire.zig, docs/themes.md, docs/config.md, docs/effects.md, 21 snapshot goldens (the grip cells only), test/gui-goldens.txt. Motion flavours differ by design, and drive notice drops and pane moves
* Motion flavours, and Lift is shadow, rim or auto: auto on a dark page ↵Gabriel Schneider2026-10-01
| | | | | | | | | | | | | | | | | | | | recedes the other panes (InactiveDim) Motion off|crisp|smooth|bouncy|playful is one parameter set (animation.Motion) every fx animation reads. One focus spring per pane drives both the lift and the dim, at the flavour's pace; a grid snaps. Lift drops glow and surface. A shadow falls on pane bodies and rails only, capped so text and the selection keep min(contrast, 4.5); rim is a hairline along the focused tag's top. auto is a shadow on a light page; on a dark page the core recedes the unfocused panes' text (InactiveDim, ported from 2e4d97f, default 30 under auto) with the same floor. Tests: focused text contrast is never below unfocused in any style, dim or theme; the focused pane's cells and tag are untouched by a dim. Shared files: src/config.zig, src/builtins.zig, src/ninep/ctl.zig, src/surface.zig, src/detached/wire.zig, docs/config.md, test/snapshots/builtins.snap (one more wheel for the longer help).
* Lift takes a style: shadow, surface, rim, glow or autoGabriel Schneider2026-10-01
| | | | | | | | | | | | | | | | | | | The user wants Lift to read on dark themes, where a shadow has nothing to darken. `Lift shadow|surface|rim|glow|auto` (`on` is auto, bare flips auto and off): surface lightens the lifted part's ground about 5.5% (Material-dark), rim is a thin soft light off its top and left edges, glow a faint halo in the theme's accent (Chrome.accent, the focused box colour), auto a shadow on a light page and surface with half a rim on a dark one (by the page's luminance). Rim and glow are the soft kernel in a light colour (no pre-warp: coloured light is laid on as is, within §8.2's budget); each style is capped per theme so text keeps min(its contrast, 4.5) (tests: darkening and lightening over the native themes). The style travels in the palette (Chrome.lift_style, one byte on the wire). Config reports and ctl reads `Lift <style>`. Shared files touched: config.zig, builtins.zig, ninep/ctl.zig, detached/wire.zig (palette byte), gui.zig. Not touched: pardes.zig, Messages.zig, mouse.zig.
* G1: Lift, soft elevation shadows (off by default)Gabriel Schneider2026-10-01
| | | | | | | | | | | | | | | | | | | | | | `Lift` (a GUI toggle, capability lift) lifts the focused pane on a critically damped spring (animation.Spring: closed form, about 240 ms to settle, velocity kept across retargets, settled = no frames) and floats notices. The core puts the lift on the regions (Region.lift, on the wire); the GUI casts from it: a soft shadow is decor, an erf rectangle on one quad grown by 3 sigma (decor.frag, flag 0x10000000), in the quad's own space so a track's transform carries it; darken-only, the caster's rect left alone, alpha pre-warped for linear light and IGN-dithered. Cast last in its group, so a neighbour's rails and rules darken with its text; a notice's before its rule. Strength is capped per theme so text over the darkest shadow keeps min(its contrast, 4.5) (test over the native themes). Opt-in until the focus lift/dim default is decided. Feel review material: docs/effects.md. Test-mode PARDES_TEST_SERIES keeps every captured frame. Shared files touched: pardes.zig (two fields, one nextWake line), detached/wire.zig (lift on the region), config.zig, builtins.zig (capability), gui.zig. Not touched: Messages.zig, mouse.zig, tty, host_io.
* Trial merge: 9P + helix + renderGabriel Schneider2026-10-01
|\
| * The GUI draws in tiers from the regions; pane chrome moves with its paneGabriel Schneider2026-10-01
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Stage 8 of docs/render-pipeline.md. The frame is drawn in groups: tier 0, one per track in paint order (tiers 2 and 3), tier 4 (notices, then guides and the debug box), tier 5 (bar cursors), each cells, images, decor. Decor (rules, rails, thumbs, grip marks, spines, the workspace and column rules, the bottom band, notice rules, bar cursors) is whole-pixel rects through the new decor.frag over ui.vert, so it carries its track's transition and clip. A closing pane's chrome comes from Surface.previous_regions. Deleted: taglineBaseRgb, topbarPaneBorderHeight, bottomTaglinePresent, frameChromeBg, cellBackgroundIs and the rail inference, paneGripCell's scan, transient_on, PaintPlan. One cover map (coverFrame) is marked from the layers and regions once a frame. New regions column, guide and debug; Surface.chrome is the palette, carried in wire v8 (not shipped yet), so an attached GUI draws the same chrome (test). A per-instance clip replaces the vertical transition's scissor. Goldens: 01-12 byte-identical. 13-17 differ only past the grid: the image pass left its scissor at the grid's size, cutting rule ends, rail feet and the bottom band in the leftover pixels whenever a picture was on screen. 16-debug now shows the debug box, which a context-row layer had hidden. 18-mid-transition is new: virtual clock (PARDES_TEST_CLOCK in the GUI), PanelSlide Newcol with the picture, frame 6 of 12. Also: the GUI sleeps when idle instead of polling every 16 ms, and a minimized or occluded window sleeps through animation. Tracy 'gui frame build' at 200x60: 992/1015 us median before, 989/987 us after. Shared files touched: pardes.zig (one export), detached/client.zig, detached/server.zig, detached/wire.zig (all additive), gui.zig. Not touched: Messages.zig, mouse.zig, tagline.zig, colors.zig, tty.zig, host_io.zig, dump.zig, exec.zig, panes.zig.
| * One Layer for tags, notices, headers and bodies; a taller tag is one layer ↵Gabriel Schneider2026-10-01
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | of rows; wire v8 src/Layer.zig merges TagLayer and BodyLayer. `rows` (0 = no layer) and a cursor at {x, y}. A tag of N rows is ONE layer of N grid rows: tagHit answers the row as `line`, bodyHit keeps its meaning, and the per-line layer bases (TAG_LINE_LAYER_BASE, HEADER_LINE_LAYER_BASE) are gone, not aliased. Wire v8, the one bump: tag layers carry rows and cursor y, and the frame carries the placed region list. v7 and v9 peers are refused in both directions (server test over both, a new frontend test over both). web: tag_layer_value 11 = rows, 12 = cursor y; app.mjs lays every row. macOS: the Zig side compiles against Layer; pardes.h still sees one row per tag layer (accepted, the macOS shell is ignored for now). No visual change: snapshot goldens and the 17 GUI goldens byte-identical. Shared files touched: pardes.zig, Messages.zig, gui.zig, macos.zig, detached/client.zig, detached/server.zig, detached/wire.zig (plus web.zig, app.mjs, edit.zig, look.zig). Not touched: mouse.zig, tagline.zig, colors.zig, tty.zig, dump.zig, exec.zig, host_io.zig, panes.zig.
* | A command pane's command is over when its process exits, not when its pty closesGabriel Schneider2026-10-01
|/ | | | | | | | | | | | | | | | A job left in the background (sleep 100 &) held the pty open, so the pane stayed running and the child a zombie until the job ended; a command that closed its terminal and ran on got its end of file at once, the host waited 100 ms for an exit, reported exit ? and hung it up. Now each command's child is watched on a thread (waitid with WNOWAIT, so its pid stays its own until the host reaps it), and the host tells the core the exit from that: after the pty's end of file, so the output before the exit is in, or 50 ms after the exit without one, a job holding the pty. The pty stays open until both, so a command that let go of its terminal is never hung up by it. All four front ends; a host that cannot start the watcher reads the exit at end of file as before. Tests: host_io's for both cases, and cmdexit.snap end to end. Co-Authored-By: Claude Opus 5.5 <[email protected]>
* Point docs and comments at the files code moved toGabriel Schneider2026-10-01
| | | | | | | | | | | | | Documentation only, no code change: README's reading order and layout, docs/design.typ's paragraph on where pane kinds and editor parts live, docs/helix-keys.md's code map (normalInput and the executors now in normal.zig and edit.zig, insertTab in edit.zig), docs/open-questions.md (execute and ttyForDir in exec.zig), and the comments that named pardes.zig for fold, Cell, takesCommandLine and runBuiltin (tools/gen_themes.zig, themes/helix.zig, detached/wire.zig, host_io.zig, lsp_zls.zig). docs/design.pdf is a retained fixture and is not regenerated. Co-Authored-By: Claude Opus 5.5 <[email protected]>
* Answer 9P on the connection's task, so a session can open its own treeGabriel Schneider2026-10-01
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The editor's loop was the only thing that could answer a 9P request, which made the editor's own syscalls through a mount of its own tree -- a Look at /mnt/9p/pardes/<me>/anything under a `9ns --mntgen` view, a Save into it -- requests only the blocked loop could serve. The name-based refusal that followed (ownMountSuffix) and the in-process routing of a mount of oneself (Client.sameSession) were patches over that, and both are gone, with the mailbox that shipped every request to the editor's thread. One rule replaces them, `pardes.turn`: the core is single-threaded, the editor's thread has the turn by default and gives it up in two kinds of gap -- while it waits for input and while a step of it is out in a host syscall -- and a cloud9 connection task takes it in those gaps to answer. `out` counts the steps that are out, from any thread: while one is, the core reads consistently but that step still holds pointers into it, so a request that would change a pane (a write, a truncation, an rmdir) is parked in the engine and retried when the turn is next given up with nothing out, and the editor's own wake waits for the count to reach zero. It is never a write of its own that a step waits on out there -- writes come from a shell performing a save between steps -- so a parked request is never the syscall's own, and making a pane or rendering a screen need not park: every yield sits before its step's mutation, so the layout and the surface are whole under it. A changing request that queued effects is answered once the editor has performed them (`echo Save > exec` returns with the file written, as acme's `put` does), and it settles the way a step does, because without that a /log reader waited for the user's next keystroke. Every host syscall on a user path has to give the turn up, not fs.zig's alone: the first end-to-end run hung in `inotify_add_watch` performing the new pane's watch effect. PDFs and images are read whole at open, so no draw goes out into the host. The core's allocator takes its fixed buffer through the lock-free interface, since a connection task allocates while the editor's thread is out in a syscall that allocates too. A Restore puts the replacement in first and releases every task waiting on the old core. cloud9 (pinned at eb1a104) parks an open, a truncating wstat, a clunk and a remove on `again`, not only reads and writes, and answers a parked job whose fid was clunked without asking the backend. Verified: test/selfmount.py runs the editor under `9ns --mntgen` and Looks at, reads and Saves its own tree through the mount; a unit test pins that a change parks while the editor is out mid-step and lands when it rests, while a read is answered in the window. 9P over the Unix socket against a tty session, same machine, Debug builds: a read of /index 278us -> 61us, a truncating body write 1184us -> 609us, exec Save 718us -> 583us; the gesture benchmark is unchanged (geometric mean 0.997 over 53 cells). Also from the reviews: a notice chip over an image or PDF pane was painted out by the picture drawn after the cells, so pictures give up the rows; in the GUI a tree-sitter context band painted over the chip, so body layers are emitted first; a message is one row of printable text, its 256-byte cut never leaves half a glyph, and one wider than its pane keeps its tail (the file name, the reason) rather than its head. Co-Authored-By: Claude Fable 5.1 <[email protected]>
* Notices become tagline bands at the top of the bodyGabriel Schneider2026-10-01
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | A message, a leader chord and a prompt used to share one row of body text at the bottom of a pane, wearing the tagline font and nothing else about a tagline. Now each one is a TagLayer of its own, emitted through the same renderHeaderLayer the pane and column tags go through, so it gets the tagline height, the small-font metrics, the band offset and the border for free -- none of which a body-grid row can have by changing its font role. The text is right aligned. The prompt stays on the canonical grid because it owns a cursor, and a cursor has to sit on a real cell. The body starts BELOW the bands rather than under them, the way tree-sitter context rows already worked. Pane.body_offset is how many rows they took and Pardes.bodyTop(pane, rect) is the one place that answers "where does the body begin" -- replacing fifteen copies of `if (tag_bottom) r.y else r.y + BOX_H` spread across the paint, hit-test, scroll, PDF and image paths, which is what let the bands and the text under them come adrift. Every notice is painted on the grid as well, because the grid is what a terminal client draws and a band it cannot see is a message it never gets; the GUI skips grid cells a tag layer covers, so nothing is drawn twice. Three bugs the bands exposed, fixed here: - a prompt band flush with the right edge put its caret one column past the pane, which the detached wire refuses -- so every frame was dropped for as long as the prompt was open. The band now reserves that column. - a click on a band mapped to Sel row 0, which is the TAG row: clicking chrome expanded a word out of the tagline and ran it as a builtin. - a watched file reloading under the editor changed the core without going through update, so needs_frame was never set and the reload was never drawn. Pardes.invalidate() is the name for that, and the file and theme reloads call it. A session can now drive its own 9P namespace instead of being refused one: ownMountSuffix answers what a path names inside this editors own tree and resolve, readLimit and write serve it from memory rather than making the syscall that never returns. The match is anchored to whole components under the registrys 9p/pardes/<name>, because a bare /pardes/<name> anywhere in a string would claim ~/src/pardes/<name>/README -- and, before write learned the same trick, write the trees bytes over the real file. readFileLimit and writeFile refuse instead, having no core to answer from. A toggle setting SETS when given `on` or `off` and only flips when it is bare, so the report LocationsConfig prints can be fed back as configuration and mean what it says. Snapshots: 97/98, from 0/98. The goldens were several commits stale and 17 scripts had stopped running; `config <line>` is a new script command that appends to the per-script startup config, so a script that clicks body coordinates pins `Verbose off` instead of counting the rows an announcement moves. nested-optout is left failing on purpose: two levels of nesting prepend vaxis F3 codepoints to typed lines, which is a real bug and is written down in docs/divergences.md with a repro. Co-Authored-By: Claude Opus 5 (1M context) <[email protected]>
* Separate tag text geometry from physical pane gripsGabriel Schneider2026-09-15
|
* Separate discontinuous pinned source context with thin bordersGabriel Schneider2026-09-15
|
* Add optional compact tagline styling for source contextGabriel Schneider2026-09-15
|
* Follow embedded PDF links through LookGabriel Schneider2026-09-15
|
* Route mouse thumb buttons to jump historyGabriel Schneider2026-09-15
|
* Refactor panes and filesystem; replace FUSE with 9PGabriel Schneider2026-09-07
| | | | | | Consolidate pane, layout, memory and host code. Serve 9P by default over Unix sockets, with runtime mounts and optional TCP/QUIC transports. Remove FUSE and obsolete proof-of-concept examples. Fix highlighting and terminal-history performance, expand differential and stress-test infrastructure, sort navigation results while preserving the next occurrence, add syntax-colored Braille minimaps, remove SPC-k, and document 9P interaction as a repository skill.
* detached: a big screen can attach, and the frame it asks for fits the wireGabriel Schneider2026-09-03
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Two defects, and either fix alone makes the other one worse. No frontend ever clamped its window to the protocol's grid ceiling — the hello carried it raw — so a 4K display at a small font, already past max_rows 128, had its geometry refused by the session's decoder as BadValue. That path answers with close(.protocol) and no refuse behind it, so the frontend was told only that the session "hung up on the connect": at a session with all 32 slots free. client.zig now asks for the largest grid the wire carries, which is what its own GEOMETRY note already promises a frontend gets — the session is drawn at its own size in the corner of a bigger window, exactly as when another frontend is the smaller one. A ZERO geometry is dropped rather than clamped, because the session grid is the smallest common one and a frontend reporting 1 would collapse everybody else; TIOCGWINSZ answers 0x0 during a teardown and the tty shell forwarded it, which was the same mute hangup by another route. That clamp alone would have replaced one bug with a worse one. max_cols * max_rows is 65536 and a run's length prefix is a u16, so the single grid legal at both bounds is the one grid whose full frame — and an attach always produces a full frame — cannot be described by one run. encodeFrame's @intCast panicked in a safe build and was illegal behaviour in a fast one. The encoder splits the run instead, bounding the CURSOR rather than the run because the gap lookahead runs ahead of it, and frameBound had already paid for the extra header. wire.version 1 -> 2 for the same reason: the geometry a v2 frontend now asks for is one a v1 daemon panics encoding, and `zig build` replacing the binary under a running session is exactly what that field exists for. A v1 daemon answers Refusal.version instead of dying with every pane shell it owns. Co-Authored-By: Claude Opus 5 (1M context) <[email protected]> Claude-Session: https://claude.ai/code/session_016Q4RATpafkwahrovHQLKRf
* acmefs: a pane's terminal gets pty/data, pty/ctl and pty/statusGabriel Schneider2026-08-27
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Step 3 of the 9P chain (docs/9p.typ 12.3, docs/registry.typ 9P-8). Nothing here is about 9P: it lands in the FUSE-served tree and any later transport inherits it. A script could write into a terminal that already existed and read its rendered scrollback. It could not START one, RESIZE one or SIGNAL one. Two of those were already effects the core emits, so `exec` and `winsize` are existing capabilities acquiring a name; only `sig` is new, and it brings the one new host method, `push_pty_signal`. pty/ctl winsize <cols> <rows> | sig INT|TERM|HUP|QUIT|KILL | exec one verb per line, validate-all then apply-all, EINVAL applies nothing -- `writeCtl`'s shape and `writeCtl`'s reason pty/status cols, rows, tty-taken as three %11d fields pty/data write is input to the process; read is the RAW output stream, gated on a reader count so a pane nobody reads costs one branch A pane that is not a terminal has no pty/ at all: the lookup is ENOENT and readdir does not list it. `PaneFile` is an enum(u4) and this takes it from 11 values to 15. ONE REMAINS. That is also why pty/ is a DIRECTORY and not three more flat names -- a subdirectory costs one value and buys its own namespace, so `ctl` and `data` did not have to be renamed. Two things the core does not know, and which are therefore not invented: a child's EXIT STATUS (a shell's death is `Event.eof`, which removes the pane, so there is no directory left to read it in) and RAW/COOKED (the core never sets a termios; the mode belongs to the program on the far side). Verified live against a daemon: pty/ appears only on the terminal pane; a `winsize 0 24` and a `sig SIGINT` are refused; a bad verb beside a good one applies neither; `echo pty-works` written to pty/data runs in the shell and its output reaches the body; and a blocking read of pty/data returns the raw stream, OSC 133 marks and all. fs-bench unchanged and still zero allocations.
* An edited row keeps its colours, four copies of forkShell become one, and ↵Gabriel Schneider2026-08-27
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Esc stops recentring ## A terminal row's ANSI colours survive being edited The loudest colour bug this editor had: one keystroke anywhere in a coloured shell row turned EVERY column of it grey. `EditAnchors` anchored a buffer line only when it was BYTE-IDENTICAL to the shell row it stood over, so a single differing byte dropped the whole row's colour projection. Worst shape is invisible: append past the pane's right edge, where the text is clipped, and the row looks the same and only its colour goes. Anchoring is byte-level now. An edit leaves the row's own bytes at both ends, and being the same bytes they keep the same colours; only what was typed has no cell under it, so only that takes none. Live, on real `fastfetch`: a 32-column blue run split into 6 + 26 around one typed character. Three defects underneath it, all found by machinery rather than by reading: * A JOIN removes a buffer line while the buffer's covered span grows, so `lines == covered` and both aligned guesses — Nth line over the Nth covered row, and the same counted from the bottom — resolved to the SAME wrong row. Every untouched row below a join went plain. Anchoring is now a streaming monotone matching: one shell-row cursor that only ever moves forward, advanced once per buffer line, linear in the buffer where the version before it was quadratic. * An EMPTY line is not evidence. Splitting a row makes one, it equals every blank row in the span, and left free to look ahead it claimed the blank row below the last output and took every coloured row in between out of reach of the lines that owned them. * Reflow under a scrolled viewport. `PageList.getTopLeft(.viewport)` returns the viewport pin verbatim, x and all, while `PageList.pin` forces x to 0 — so after a reflow remapped a tracked pin into the middle of a row, the text pass dumped row 0 from that column while the colour pass paired the fragment with the row's FIRST cells. Row 0 wore its left half's colours until the pane snapped back to live output. `bodyText` dumps from column zero now, which is also what ghostty's own renderer draws. Also here: DECSCNM (reverse video) was silently dropped whenever `tty_filter` was off, because the raw path resolved a `.none` colour by role and never consulted the mode. The test that found the first two is the one worth keeping: random editing against an ABSOLUTE oracle — every row's own text names the colour it must have — because the differential oracle it replaced was blind by construction. It skipped the edited row, which is the row the user is complaining about. ## Esc returns to a pane without moving its view Esc in body normal mode runs `Last`, "the pane you were in before this one", and that went through `focusPaneLine`, which recentred a file on the target line unconditionally. So returning to a buffer repainted the whole screen to show a line that was already on it. `focusPaneLine` takes a landing now: `.center` for the three callers going somewhere you have not been (a look target, a path a pane already holds, `@pN:LINE:COL`), `.keep` for Esc. `.keep` leaves the view alone and lets `ensureCursorVisible` — which already existed and already scrolls by the minimum into the `scroll_off` band — be the only thing that may move anything. Not `line = 0`, which `focusPaneLine` already understands as "focus and touch nothing": a background pane's view can move while you are away, because the wheel scrolls the pane under the POINTER and a resize reveals no cursor, so the recorded cursor plus a minimal nudge is what actually gets you back. Ctrl-o and Ctrl-i keep centring, and the asymmetry is structural rather than arbitrary: `Last` only ever CROSSES panes, so the pane it lands on already holds the view you left it with, while `jumpBy` can land in the SAME pane, where a long in-file jump would arrive on the very top or bottom row with `scroll_off` lines of context on one side. Helix splits the same pair the same way — its jumplist centres, its buffer switch does not. One deliberate consequence: under `.keep` a PDF's page is not restored AT ALL, because a page reveal IS that pane's view and a reveal of the page you are already on still snaps `document_scroll_y` to that page's start, discarding where you had read to. When something moved the pane while you were away — the wheel again — Esc leaves it where the wheel left it, and Ctrl-o is how you reach the recorded page. ## host_io.zig: the machine-local half of a host, once `host.zig` is the seam. The part of the answer that is identical on every host with an operating system under it — fork a pane's shell, put bytes on a disk — was written FOUR times: in tty.zig, gui.zig, macos.zig and detached/server.zig. What those copies had in common says what they were for: all four were missing FD_CLOEXEC on the pty master, so in every shell pardes has shipped, a program in one pane could read another pane's terminal. One copy now, and the wire got smaller for it: `ServerMsg.spawn` is gone. A frontend never asked the server to fork anything — the server has an operating system under it and forks through `host_io` like every other host — and `decodeClient` lost the scratch buffer that message needed.
* One core behind N frontends, the board's own runner moved in, and every ↵Gabriel Schneider2026-08-27
board cap on one screen ## The wire is the effect stream, not a new protocol `pardes --detach` leaves a core running with no terminal; `pardes --attach` is a frontend that owns a terminal and a socket and nothing else. N frontends on one core all look at the same screen — `screen -x`, not N sessions. The codec (`src/detached/wire.zig`) carries exactly one `Event` or one `Host.VTable` call per message. That is not a coincidence and it is why there is no third vocabulary to keep in step: the core's IO seam was already a struct of function pointers with plain-data arguments, so a socket is a legal implementation of it. `nested.zig`'s socket could not be reused — it carries a builtin command line, and a command line cannot carry a frame. ARCHITECTURE-NEUTRAL on purpose, not as decoration. The frontend on the far end may be riscv32-freestanding on the ESP32-P4 while the core is x86_64 Linux, so every field is an explicit little-endian fixed width and no message is a blit of a native struct. A protocol that only works between two builds of the same compiler would have thrown away the one frontend that motivated it. ## The board comes in; its toolchain stays out `src/p4.zig` becomes `src/esp32p4.zig`, and the pardes half of `../05-zig-p4` — the vaxis-over- serial runner, the UART editor terminal, the keystroke rescue ring, the on-die test suite — moves into `src/esp32p4/`. `build.zig.zon` gains `.zig_p4 = .{ .path = "../05-zig-p4" }`, so `zig build -Dplatform=esp32p4 -Desp32p4-firmware` builds, flashes, monitors and self-tests the board from this repo's `build.zig`. The DIVISION is the point. What moved is what only pardes wants: the runner that drives a pardes core over a serial line. What stayed is everything a second project would also want — the HAL, the register/radio/oracle layers, the linker script, `_start`. `zig_p4` declares no dependencies of its own and its `build()` early-returns when it is not the root package, so this costs the package graph exactly zero packages and the editor's own builds nothing at all. ## limits.zig: nine forgettable places become one budget Nine `platform == .esp32p4` capacity tests lived in nine files. They were never nine decisions — they are ONE decision, how much memory this build may spend, taken nine times where no reader could see the total. `src/limits.zig` puts the whole budget on one screen with every cap named against what it is measured against, derived from two booleans. The payoff is testability on a machine that is not the board: the caps are ordinary comptime values, so a host build can be compiled against the board's numbers and the parking, eviction and clamping paths a 240 KiB core takes get exercised by the normal test suite instead of only over a UART. ## A bare `zig build` `zig build` with no arguments now builds the tty and GUI binaries and installs them into `~/.local/bin`, and says so once on stdout with the flag that overrides it. The old default built one binary into `zig-out` — a path nothing on a `PATH` ever looks at, which made "build it" and "use it" two different commands for no reason.