<feed xmlns='http://www.w3.org/2005/Atom'>
<title>pardes.git/test/monkey9p.zig, branch main</title>
<subtitle>Pardes</subtitle>
<id>https://git.0x4200.cafe/pardes.git/atom?h=main</id>
<link rel='self' href='https://git.0x4200.cafe/pardes.git/atom?h=main'/>
<link rel='alternate' type='text/html' href='https://git.0x4200.cafe/pardes.git/'/>
<updated>2026-11-09T04:28:28Z</updated>
<entry>
<title>tests: monkey and replay are modes of the T1 runner and snapshot.zig goes: behaviour-test runs test/monkey9p.zig bare (its smoke), `-- --monkey &lt;seeds&gt;` drives an in-process session with test/monkey.zig's input (keys, leader bursts, the mouse, tiny resizes, over /kbdin, /mousein and Session.resize) hunting panics and hangs, `-- --replay` records each seed's sessions as a tape in memory and replays it (tape.replayBytes, replayMain's loop: three replicas and the dump), and `process-test -- --replay` records and replays the tty cases' real sessions as `zig build replay` did (with the addons it lacked, so images passes); the snap, snap-driver-test, monkey, monkey-test (it no longer compiled), replay and monkey-session steps, pardes-snap and build/snap.zig (now build/process.zig) go; checked: a counter outside the core bumped at the 40th update diverges every T1 seed and 89 of 103 old tapes, one bumped on terminal output diverges no T1 seed but 69 tty cases and 70 old tapes (why the tty replay stays), a panic on a window under 5x5 is found by the T1 monkey (seeds 1 8 17 20) and the old one (seed 13)</title>
<updated>2026-11-09T04:28:28Z</updated>
<author>
<name>Gabriel Schneider</name>
<email>gbrls@0x4200.cafe</email>
</author>
<published>2026-11-09T02:49:04Z</published>
<link rel='alternate' type='text/html' href='https://git.0x4200.cafe/pardes.git/commit/?id=174e031985db910b4907ba324d6fd12c67898a5a'/>
<id>urn:sha1:174e031985db910b4907ba324d6fd12c67898a5a</id>
<content type='text'>
Co-Authored-By: Claude Fable 5.1 &lt;noreply@anthropic.com&gt;
Claude-Session: https://claude.ai/code/session_016Y9MNAtZenFv4weoMD3cge
</content>
</entry>
<entry>
<title>settings #2, five presentation variants go, each now the style it had by default: Colors (text panes and terminals coloured, with SPC t c), TreeContextTagStyle (TreeContext rows in the tag's style), FocusTint (the focused tags tinted), SyntaxBold (keywords not bold) and Ligatures (the window shapes them; gui.zig's syncLigatures and the words' ligatures flag go, and Capabilities.ligatures); the reads deleted are in File.zig, body_layer.zig, Tables.zig, exec.zig's applySettingBuiltin, Core.zig (SessionView.layered left at its default), draw.zig, surface.zig's comment and gui.zig; deleted tests: "settings: Colors off draws a source plain, and on again colours its keywords", "Ligatures is a setting only in the shell that shapes text" and "Ligatures off draws a ligature as its cells' own glyphs, and on shapes it again", and the toggled half of TreeContext's Core test; rewritten through what stays: SyntaxBold's every-row test as keywords coloured in every row, the tag-click tests on JumpTrail, the tag-name tint tests with the focused tag's ink plain, Core's tab test by the keyword's ink; golden changes, each the cut's: syntax, theme, mini and ninep lose their Colors toggles and frames (the message row after them names the builtin before, Mini), and leader, ttyhelp and builtins' Help listing is four columns narrower, TreeContextTagStyle having been its longest word; perf.zig's harness id moves again; the window scenes from 09 on take new gui-goldens lines, the scene no longer writing FocusTint on and its word gone from the message row (the base window build running the same scenes gives the same frames); gates on tui: core-test 965/965, behaviour-test 242/242, unit-test 1482/1483, pane-test 171/172, snap all 105 scripts ok, process-test 0 of 37 failed; gui unit-test 1526/1527, gui-golden 0 of 1 failed</title>
<updated>2026-11-06T02:58:31Z</updated>
<author>
<name>Gabriel Schneider</name>
<email>gbrls@0x4200.cafe</email>
</author>
<published>2026-11-06T01:40:49Z</published>
<link rel='alternate' type='text/html' href='https://git.0x4200.cafe/pardes.git/commit/?id=dfcc040ac979b695dc2e68b9d866c2769ba6d928'/>
<id>urn:sha1:dfcc040ac979b695dc2e68b9d866c2769ba6d928</id>
<content type='text'>
Co-Authored-By: Claude Fable 5.1 &lt;noreply@anthropic.com&gt;
Claude-Session: https://claude.ai/code/session_016Y9MNAtZenFv4weoMD3cge
</content>
</entry>
<entry>
<title>settings #1, Tagbottom goes, a pane's tag at its head the one value anyone used: layout.bodyTop and tagTop, mouse's seam and clampBorderRow, applyRowSplit, body_layer's copy and the tests' row arithmetic lose their other arm (reads deleted in Core.zig, layout.zig, mouse.zig, draw.zig, gui.zig, macos.zig, whose ABI slot 11 says 0 now, and ninep/events.zig, ninep/tree.zig, look.zig, Messages.zig), with the registry row, SPC t b, Help, /commands, the builtins JSON and the reference row; deleted: the snapshot scripts tagbottom, tagbottomshort and tagbottomimage with their goldens, the behaviour test "placement: under Tagbottom a look from a pane's text goes under its text too", the window scene 10-tagbottom with its gui-goldens line and the macOS parity check tags-bottom; tests that ran both ends of the tag run the head; a dump that kept the field loads without it; perf.zig's harness id moves in the two debug baselines (its reads of the field went); the window scenes after 10-tagbottom (12 to 15, 17) take new gui-goldens lines, the two Tagbottom writes gone from the scene and with them their words on the message row (the base window build running the same scenes gives the same frames); gates on tui: core-test 966/966, behaviour-test 243/243, unit-test 1484/1485, pane-test 171/172, snap all 105 scripts ok, process-test 0 of 37 failed; gui unit-test 1529/1530, gui-golden 0 of 1 failed</title>
<updated>2026-11-06T02:58:31Z</updated>
<author>
<name>Gabriel Schneider</name>
<email>gbrls@0x4200.cafe</email>
</author>
<published>2026-11-06T01:27:44Z</published>
<link rel='alternate' type='text/html' href='https://git.0x4200.cafe/pardes.git/commit/?id=f930d53c58187381b2f0df6892546ac1c217c9db'/>
<id>urn:sha1:f930d53c58187381b2f0df6892546ac1c217c9db</id>
<content type='text'>
Co-Authored-By: Claude Fable 5.1 &lt;noreply@anthropic.com&gt;
Claude-Session: https://claude.ai/code/session_016Y9MNAtZenFv4weoMD3cge
</content>
</entry>
<entry>
<title>dumps: a Dump is a labelled entry of the state file (src/Db.zig, architecture §14.3), one fsynced transaction: `dumps/&lt;label&gt;` holds the session as a database of its own, which the walk writes by field name into xitdb's maps (walk.Put: a struct a map keyed by its fields' names, a union a map of its one field, records a list, plain numbers one bytes value tagged with their kind and width, an enum its name, a number a uint or an int, an optional an absent key, a type with dumpWrite its bytes, a .blob text of 4 KiB and up its BLAKE3 hash) and another build reads by name (walk.Get: defaults, notes, a pane it cannot read left out, as the byte loader did), with the hashes of the texts it keeps once under `blobs/`, so an unchanged text is written once across dumps; labels are `pardes-&lt;date&gt;-&lt;time&gt;` (`-2` and on in one second), `last` names the newest, the newest 32 are kept and the texts no kept one names are dropped in the same transaction, and at a session's end the file is compacted into `pardes.xitdb.new` and renamed over once it is past twice what it holds plus 64 MiB (another session notices the new file by its inode); a dump travels as an image, a database in memory holding it alone, deterministic (std.Io.failing's zero ids, so the replay's fixpoint holds): what Dump hands its host, what `Dump export &lt;path&gt;` and $PARDES_DUMP write whole (the snapshot goldens' `Restore &lt;path&gt;` unchanged), what a Restore of a label is read into, what a handover carries (`handover-&lt;seq&gt;`), and what `-l` and `Restore &lt;path&gt;` read; Restore takes a label or a path (bare: this session's last dump, else the newest), a label the file keeps none of refused in words (`Restore: no dump labelled X`, asked of the authority as the closed panes are); `/dumps` lists them, `&lt;label&gt; &lt;date&gt; &lt;size&gt; &lt;panes&gt;`, the newest first; Upgrade commits `upgrade-&lt;pid&gt;` before the exec, its record naming the state file and the label (a record without one, an older build's, names its `.dump` file, read as before), and the label stays; the byte dump stays for the join, the closed panes and the old `.dump` files (dump.byteDump), and dump.outPath, uniqueIn and upgradePath go for dump.label; measured (Debug, a session with a 3.2 MB file and three scratch panes, 20 dumps): the image on the loop 64-66 ms (59 ms of it BLAKE3 of the file, 0.65 ms in ReleaseFast) where the byte dump was 0.7 ms, the commit 9-16 ms with its fsyncs where the file was written in 2, a Restore 1.8 ms to read and 11.4 to load, and the file 3.55 MB after the first dump and 379 KB a dump after it with the big text unchanged (3.58 MB a dump with it edited each time, as twenty `.dump` files would be); tests: dump.zig's 'two dumps in one second get names of their own' and 'a dump makes every directory missing on the way to its DumpDir' become Db.zig's 'two dumps with one label get labels of their own' and 'a Dump makes every directory missing on the way to its DumpDir, and the file there', 'a session read by name, field by field as another build reads it, is the same session' becomes 'a session's byte dump read by name...' beside the new 'a session's image, read by name as every load reads it, is the same session, and dumps to the same bytes again', two replica tests compare the join's stream with dump.byteDump, the Handover tests read the carried image, upgrade's record test carries a label, the root listing tests and the monkey list `dumps`, the behaviour suite gains five root-ctl tests (a Dump and Restore round trip through 9P with `/dumps` and the tag's word, a label kept none of refused in words, the 33rd dump dropping the first, an unchanged text written once across two dumps by the file's size, a killed Dump leaving the one before it whole), the test host performs a Dump's write and a Restore as a host does, the monkey lets go of its log follower at a Restore, and daemon.py's upgrade case finds the kept dump in `/dumps`</title>
<updated>2026-11-05T22:16:37Z</updated>
<author>
<name>Gabriel Schneider</name>
<email>gbrls@0x4200.cafe</email>
</author>
<published>2026-11-05T17:42:45Z</published>
<link rel='alternate' type='text/html' href='https://git.0x4200.cafe/pardes.git/commit/?id=15a1b94eaa73747afaa299262a2d6e4fd6af42fd'/>
<id>urn:sha1:15a1b94eaa73747afaa299262a2d6e4fd6af42fd</id>
<content type='text'>
Co-Authored-By: Claude Opus 5.5 &lt;noreply@anthropic.com&gt;
Claude-Session: https://claude.ai/code/session_016Y9MNAtZenFv4weoMD3cge
</content>
</entry>
<entry>
<title>the 9P monkey sends InactiveDim again: the word left its list with Lift (uqrqpqqo) and came back to the core with the window's dim (rsmzxvpt), the same setting byte, parsed as before, that both frontends take, so the monkey sends it as it did before the programme; zig fmt lines the list's columns up again (SyntaxBold, put in where Motion and InactiveDim were, had left them out of line); gates on tui: monkey-session-smoke, the behaviour gate's fixed seed, passes; monkey-session over seeds 1 to 5: 1, 3, 4 and 5 are clean, and seed 2, its draws moved by the longer list, finds a ctl bug the effects never touched: a /ctl write refused for a control character leaves the start of its line, held from the write before, to run at the clunk, two err records for one failed write (tree.zig's control-character check refuses the write but keeps the fid's held text), shrunk to two ops: zig build monkey-session -Dplatform=tui -- --seed 2 --steps 763 --only 761,762</title>
<updated>2026-11-05T18:28:16Z</updated>
<author>
<name>Gabriel Schneider</name>
<email>gbrls@0x4200.cafe</email>
</author>
<published>2026-11-05T17:45:07Z</published>
<link rel='alternate' type='text/html' href='https://git.0x4200.cafe/pardes.git/commit/?id=3205f21d9b5f222f7228aecda36d9d2f3a7742e4'/>
<id>urn:sha1:3205f21d9b5f222f7228aecda36d9d2f3a7742e4</id>
<content type='text'>
Co-Authored-By: Claude Opus 5.5 &lt;noreply@anthropic.com&gt;
Claude-Session: https://claude.ai/code/session_016Y9MNAtZenFv4weoMD3cge
</content>
</entry>
<entry>
<title>effects out, the core's motion: the view's glide is the frontends' alone, and the core keeps no motion and no clock but its deadlines; glide.zig is a frontend's picture of a scroll, measured on the frame it is about to draw against the last it drew (a body's first line and the text it shows in its region, Region.doc new; the last frame's cells kept by the glide itself), on the frontend's own clock, the tty drawing it at whole rows and the window at the pixel; each frontend draws its own motions' frames from the last frame (the tty's poll presents it again, the window's loop renders it again, as its blink did) and paces its own waits, so the core's continuous path goes (Pardes.continuous and shell_continuous, the pump's one-millisecond wait, the view's drawn fields, the tape's watch of it); the Motion word and its five flavours go (the glide and the caret move at the default's pace, and no switch for reduced motion is left), and animation.zig with them (Motion and Spring gone, Follow now glide.zig's with its test, CursorGlide gone with the core's caret, Blink the window's with its test, Box and layout.panelBox gone, the frame's length now Pardes.frame_ns); a PDF strip no longer glides, its scroll landing at once (its follow, stripAt's motion and Pdf.moving go); the window's glide tests are fed frame by frame as the window feeds them; the window monkey runs by seed, not by flavour, and the 9P monkey loses Motion; the tty wakes a frame at a time while its glide or trail moves, as the core's continuous path woke it; all 109 snapshot goldens and the 16 GUI goldens unchanged; tests gone: a carried spring keeps its speed; a spring settles without overshoot; each flavour is its own motion; a cursor eases a cell at a time (animation.zig's); a PDF strip glides a notch and a page to their marks; a PDF glide renders the pages it crosses</title>
<updated>2026-11-05T18:26:36Z</updated>
<author>
<name>Gabriel Schneider</name>
<email>gbrls@0x4200.cafe</email>
</author>
<published>2026-11-05T16:56:13Z</published>
<link rel='alternate' type='text/html' href='https://git.0x4200.cafe/pardes.git/commit/?id=3488111273764c17a42870484fabc36abb0634a1'/>
<id>urn:sha1:3488111273764c17a42870484fabc36abb0634a1</id>
<content type='text'>
Co-Authored-By: Claude Opus 5.5 &lt;noreply@anthropic.com&gt;
Claude-Session: https://claude.ai/code/session_016Y9MNAtZenFv4weoMD3cge
</content>
</entry>
<entry>
<title>effects out, Lift and InactiveDim: the core's per-pane lift springs (Pardes.lifts, lift_serials, Region.lift, Chrome.lift_style) and the unfocused panes' recede (draw.zig's dim, its contrast memo and linear blend) go with the words, their settings and tests; the window loses its soft shadows (addSoft, the lift and rim casting, the shadow a sliding pane cast, the contrast ceilings, decor.frag's Gaussian path) and Chrome.sel_fg and sel_bg, which only the ceilings read; a pane closed mid-lift left its spring unsettled and continuous() true for ever, so a window could never go idle (monkey-gui failed `never settled` in the Lift combinations, and passes now); GUI golden 18-mid-transition changes, its sliding pane casting no shadow (the scene leaves with the transitions); root_ctl's out-of-range case is SearchRows', settings' GUI-only case GripWidth's, the window monkey loses its Lift axis</title>
<updated>2026-11-05T18:15:07Z</updated>
<author>
<name>Gabriel Schneider</name>
<email>gbrls@0x4200.cafe</email>
</author>
<published>2026-11-05T14:34:44Z</published>
<link rel='alternate' type='text/html' href='https://git.0x4200.cafe/pardes.git/commit/?id=2bd4066a932ec8fc401e4b85e9f8441b17df3cfb'/>
<id>urn:sha1:2bd4066a932ec8fc401e4b85e9f8441b17df3cfb</id>
<content type='text'>
Co-Authored-By: Claude Opus 5.5 &lt;noreply@anthropic.com&gt;
Claude-Session: https://claude.ai/code/session_016Y9MNAtZenFv4weoMD3cge
</content>
</entry>
<entry>
<title>config, exec, look: Placement, BootShell, LookWord, DirLook and DiffLook go, with what each kept of pardes before it took acme's ways (exec's pardesPlace, atFoot, belowParent, pardesDoc and evictLonePristineTty with mouse.zig's call, layout's splitParent and columnFitsHalves that only they called, look's ls in a terminal for a directory, the word and patch looks' other branches, the SPC c p key, their Help and builtins-JSON rows, their lines in docs/typ and a changelog line), and the tests of them: Dir.zig's "DirLook terminal: a look at a directory opens a terminal there, no pane listing it", pardes.zig's "first document fallback splits like Term from the originating pane", "first document keeps its originating shell and only opens a column when it fits", "BootShell keep leaves the untouched boot shell a dragged document joins; replace closes it" and "a Look on a directory does not type ls into an occupied terminal", keys.zig's "keys: SPC j n and SPC j p walk on from the place a Look's +Search listing chose, its keyboard left in the file", look_and_exec.zig's "look-and-exec: a second LookWord list look in a directory refills its +Search rather than stacking another", placement.zig's "placement: Placement pardes puts a new pane right under the asking one, and bare it steps back to acme", output.zig's "LookNext/LookPrev keeps the result output a Look chose through a no-match answer, and a stale serial once it is deleted", and the snapshot scripts that turned them on (cmdword, doccol, look-dir, lookpanes, rsearch, splitparent, ttylook); the Del question's test makes its middle panes the acme way, a New at the column's foot, and the others that turned one on for their layout run without it</title>
<updated>2026-11-05T03:32:56Z</updated>
<author>
<name>Gabriel Schneider</name>
<email>gbrls@0x4200.cafe</email>
</author>
<published>2026-11-05T03:08:16Z</published>
<link rel='alternate' type='text/html' href='https://git.0x4200.cafe/pardes.git/commit/?id=2367ed34beccf3f2f1b8bbf6b8a171f50933d7a1'/>
<id>urn:sha1:2367ed34beccf3f2f1b8bbf6b8a171f50933d7a1</id>
<content type='text'>
Co-Authored-By: Claude Opus 5.5 &lt;noreply@anthropic.com&gt;
Claude-Session: https://claude.ai/code/session_016Y9MNAtZenFv4weoMD3cge
</content>
</entry>
<entry>
<title>rename: the terminal frontend is the tui (-Dplatform=tui, src/tui/tui.zig, --tui with --tty kept as its old spelling, the process tests' backend .tui and --backend=tui, docs/typ/builtins-tui.json, the Settings report's platform), and the builtin Tty is Term (its registry entry, SPC n t, every default tag, the docs, tests and snap scripts), Tty still run as Term through exec.acme_words, Tty+fish too on a tag and a ctl; tty as a pane's raw mode and as the OS's terminal unchanged</title>
<updated>2026-11-05T00:12:33Z</updated>
<author>
<name>Gabriel Schneider</name>
<email>gbrls@0x4200.cafe</email>
</author>
<published>2026-11-04T23:54:21Z</published>
<link rel='alternate' type='text/html' href='https://git.0x4200.cafe/pardes.git/commit/?id=0ff1fc99eb870f8a1e714a71c1a853066737c822'/>
<id>urn:sha1:0ff1fc99eb870f8a1e714a71c1a853066737c822</id>
<content type='text'>
Co-Authored-By: Claude Opus 5.5 &lt;noreply@anthropic.com&gt;
Claude-Session: https://claude.ai/code/session_016Y9MNAtZenFv4weoMD3cge
</content>
</entry>
<entry>
<title>test: review of the mky port: the monkey's child exits 3 (a harness failure the parent reports) when its run errors out, as when a fresh session's /index will not parse or its log refuses `follow new`, where it exited 1 with no failure line and the parent took it for a clean run (both breaks passed the smoke before); /index's blank line unparsable again, as Python's splitlines saw it (tokenizing skipped it); served_lines' fourth self-check back (a piece that fills its Twrite goes on), its room a parameter again; the port checked against nine breaks of the reviewer's own, each failing: /index ending in a blank line (harness, now), log's `follow new` refused (harness, now), /pane's readdir listing new twice (chunked equals whole), a repeat counted (x+2) (one failure rule's upper bound, at 3x3000 steps, not the smoke), a held name's fault no longer said by its write (close runs only line files, at 3x3000 steps), a panic on writing 1 to scroll (a crash, shrunk 188 -&gt; 1 op), /screen answering .again (a hang), pty/data swallowing ^U (all three terminal cases), a terminal's body from the active screen alone (history, raw key); behaviour, process, fs-unit and core tests and agent_session_test.py pass</title>
<updated>2026-11-04T20:20:56Z</updated>
<author>
<name>Gabriel Schneider</name>
<email>gbrls@0x4200.cafe</email>
</author>
<published>2026-11-04T20:13:31Z</published>
<link rel='alternate' type='text/html' href='https://git.0x4200.cafe/pardes.git/commit/?id=4b4752d66cfdb45d9d5bcaa3c7bb132c35b591f4'/>
<id>urn:sha1:4b4752d66cfdb45d9d5bcaa3c7bb132c35b591f4</id>
<content type='text'>
Co-Authored-By: Claude Opus 5.5 &lt;noreply@anthropic.com&gt;
Claude-Session: https://claude.ai/code/session_016Y9MNAtZenFv4weoMD3cge
</content>
</entry>
</feed>
