summaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAge
...
* Grep and Find say when a cap hid what they might have found: `cut at 512 ↵Gabriel Schneider17 hours
| | | | | | | | | | | | | | | | hits`, `N files read only in part (first 256 KiB)`, `walk cut at N entries`; one that finds nothing leaves the +Search as it was Grep reads 256 KiB of a file and stops at 512 hits, and its walk at 20000 files or 100000 entries; Find stops at 512 names. A cap hit was silent, so a match past one read as `pattern not found`. The walk now notes each cap it hit (fs.WalkNotes) and the +Search says it in a line under the hits, as it says directories skipped; a search that found nothing but hit a cap opens the +Search with those lines, not a miss. A Find or Grep that finds nothing and passed nothing over leaves the directory's +Search with the last search's rows, its failure saying so, where it used to empty it. Co-Authored-By: Claude Opus 5.5 <[email protected]>
* A language server that failed a moment ago is said to have failed, with when ↵Gabriel Schneider17 hours
| | | | | | | | | | it is tried again, not `language server for .zig not found` During the backoff after a failed spawn or handshake, ensure() answered NoServer, which the client turned into "not found" for a server that is installed. It is now Backoff, said `zls failed recently; retry in Ns`. Co-Authored-By: Claude Opus 5.5 <[email protected]>
* A Rename the server answers for this file alone, while other open files of ↵Gabriel Schneider17 hours
| | | | | | | | | | | | | | | | | its language say the name, is previewed, never applied; Lspwhy shows the files synced first zls renaming at a declaration answers with that file's edits only, cold or warm, whatever it was told of (checked against zls directly: from a call site in main.zig it renames both files, from the declaration in util.zig it never does), so the first Rename applied in util.zig and left main.zig broken. The client now looks for the old name in the other open files of the language; when one says it and the answer left it alone, the edits are a preview with a row naming each such file, and the message row says the server renamed in this file only. Lspwhy carries the open files too, so its trace shows `synced X first` for the query it explains. Co-Authored-By: Claude Opus 5.5 <[email protected]>
* Restore reads a clean file from disk, and an unsaved one whose file changed ↵Gabriel Schneider17 hours
| | | | | | | | | | | | | | | | | | | | | since the Dump is said changed, its first Save asking before overwriting it A Restore put back the dump's copy of every file pane and armed the watch with the disk as it now was, so a file changed on disk after the Dump came back as its old text, marked clean, and the next Save wrote the old text over the new without a word. A clean file pane's text is now not in the dump: Restore reads its file (an older dump's copy is read past too, and used only when the file is gone). An unsaved pane keeps its text once, base64, the plain copy beside it dropped, with the hash of the file it was read against (`disk_hash`); a Restore that finds another there says the file changed on disk, and Save asks once before overwriting it, as for a change seen while it was open. file_watch.zig's own tests were never collected (no test root imported it): main.zig's test block does now, and the restored watch test says what a clean restore now reads. Co-Authored-By: Claude Opus 5.5 <[email protected]>
* Release Pardes 0.22Gabriel Schneider19 hours
| | | | Co-Authored-By: Claude Opus 5.5 <[email protected]>
* The docs take round 35: cloud9's errno table, Restore's parser words, insert ↵Gabriel Schneider19 hours
| | | | | | after a match, LookWord list rows, the last-run awk without exit, command pane directories, rename previews, listings that keep the keyboard, set-value toggles, pty/run one line a write, /pager resolving, PDF :N:H, skipped directories and Lspinfo Co-Authored-By: Claude Opus 5.5 <[email protected]>
* Petscii, Palette, Ascii, PdfFit and PdfTint take the state their tag word ↵Gabriel Schneider19 hours
| | | | | | | | | | | | | names, set rather than flipped: `Petscii on`, `Palette terminal`, `PdfFit width` Each could only flip, so a script had to read the tag to know which way a write would go. Each now takes an optional argument, the value its tag word shows (on|off for Petscii and Ascii, commodore|terminal for Palette, width|height for PdfFit, disabled|filtered|full for PdfTint), and sets it; bare, it flips as before. Anything else is refused with what it takes. Co-Authored-By: Claude Opus 5.5 <[email protected]>
* A file name with a newline is refused naming it once: `pardes: ↵Gabriel Schneider19 hours
| | | | | | | | | | two\nlines.txt: a file name is one line...`, not `a file name: a file name ...` The refusal's subject was the words "a file name", and its reason began with them again. It now names the name given, its newline shown `\n`, as /log and /index show one. Co-Authored-By: Claude Opus 5.5 <[email protected]>
* A second LookWord list look in a directory refills its +Search, as a ↵Gabriel Schneider19 hours
| | | | | | | | | | terminal's word looks already did, rather than stacking another Each word looked at with LookWord list opened a +Search of its own in the same directory. Word looks now share the directory's one, the next word refilling it; a `/` search still keeps a +Search each. Co-Authored-By: Claude Opus 5.5 <[email protected]>
* `@pN:` looks at pane N, its trailing colon dropped as any look's is, and a ↵Gabriel Schneider19 hours
| | | | | | | | | | | line past the end is said one way: `<file>:N has no line N`, for `:N` as for `file:N` lookAt dropped a trailing `:` from what it read but resolved the target from the text as written, so `@p33:` was `no match for "@p33:"`. And `:99999` said `address out of range` while `file:99999` said `has no line 99999`: a bare line number past the end now says the latter too. Co-Authored-By: Claude Opus 5.5 <[email protected]>
* A terminal's petscii:on survives Dump and Restore: the dump says it, and the ↵Gabriel Schneider19 hours
| | | | | | | | | | restored terminal draws its graphics as glyph art again An image pane's Petscii choice was dumped; a terminal's was not, so a Restore brought it back drawing pixels. dump.Terminal carries petscii, and a restored terminal is turned to glyph art when it was. Co-Authored-By: Claude Opus 5.5 <[email protected]>
* A name write to a terminal, image or PDF and Edit on a pane with no text are ↵Gabriel Schneider19 hours
| | | | | | | | | | | refused EINVAL, in words a mount maps so: `invalid rename: ...` and the data write's own refusal `rename not allowed: ...` went out EPERM, and `Edit ,d` on a PDF ran and failed late, EIO, while a data write there was EINVAL. The rename refusal now says `invalid rename` with EINVAL, and Edit on a pane with no text is refused before anything runs with the data write's words. Co-Authored-By: Claude Opus 5.5 <[email protected]>
* `:N:H` looked at on a PDF pane turns to page N, as `book.pdf:N:H` does from ↵Gabriel Schneider19 hours
| | | | | | | | | | anywhere, not a refusal that a PDF has no text A PDF's own pane took only `:N`; `:5:3` was refused as no text to address while `book.pdf:5:3` turned to page 5 (its search hit 3). lookPdfPage now reads the page before a colon and the hit after it. Co-Authored-By: Claude Opus 5.5 <[email protected]>
* /pager resolves its directory, `~` and `..` alike, and refuses one that may ↵Gabriel Schneider19 hours
| | | | | | | | | | | | | not be written, permission denied; `pardes -` from such a directory pages into the session's `/tmp/../etc` named a +Pager `/tmp/../etc/+Pager`, a second one beside `/etc/+Pager`, and `~/x` was refused as relative. The directory is now home-expanded and resolved before it is checked, and one that may not be written is refused as a file's name there is (fs.deniedAbove). `pardes -` run in such a directory (`git log` under /usr/src) asks for the session's +Pager instead, so its text is still paged. Co-Authored-By: Claude Opus 5.5 <[email protected]>
* A refused /pager write leaves its open no answer, not the +Pager it answered ↵Gabriel Schneider19 hours
| | | | | | | | | | before writePager emptied the answer only once the directory had passed, so a read after a refused write on the same open still gave the last write's serial. It is emptied first, as a refused look or exec leaves none. Co-Authored-By: Claude Opus 5.5 <[email protected]>
* Lspinfo before any query says a server is not started yet, its first query ↵Gabriel Schneider19 hours
| | | | | | | | | | starting it, not `off` while it is on PATH A server is spawned on its first query; until then its row said `off`, which reads as disabled. It now says `not started yet (its first query starts it)`; a server disabled by its environment still says so. Co-Authored-By: Claude Opus 5.5 <[email protected]>
* A Rename, references and the like tell the server of every open file of its ↵Gabriel Schneider19 hours
| | | | | | | | | | | | | | | | language first, so a rename reaches files it had not seen and the preview lists them zls renames only in the documents in its store, and syncDoc told it of a file only when a query was asked from it, so renaming in progress.zig left main.zig unchanged and broken, with no preview. For a query whose answer can reach other files (rename, references, implementation, workspace symbols, calls and type hierarchy), the host snapshots every other open file and the client syncs those its server answers for first. The mock server answers a rename in each document it was told of; lsp-rename-files opens main.rs by a look, renames from lib.rs, and its preview lists both. Co-Authored-By: Claude Opus 5.5 <[email protected]>
* A Rename that previews other files answers its preview pane to the 9P write ↵Gabriel Schneider19 hours
| | | | | | | | | | | | | that asked it, and says on the message row that it previewed and applied nothing A multi-file Rename filled a +Search preview but the waiting exec write answered the pane it was asked from, and nothing was said. lspResponse now notes the pane it filled (fs.lsp_result); the write that waited for the answer reads that pane back, as a look or an exec answers the pane it went to, and the message row says `Rename: N edit(s) across files, previewed, not applied`. Co-Authored-By: Claude Opus 5.5 <[email protected]>
* Grep and Find pass over a directory they may not open and keep every other ↵Gabriel Schneider19 hours
| | | | | | | | | | | | hit, saying in one line how many they skipped One `chmod 000` directory anywhere under a pane made every Grep say `pattern not found` and every Find `access denied`: the walk's `try w.enter`/`try w.next` ended it at the first refusal. Each is now counted and passed over, and the +Search ends `N directories skipped: permission denied` under the hits. Co-Authored-By: Claude Opus 5.5 <[email protected]>
* Every EINVAL or ENOENT refusal a unit test meets must say its errno in words ↵Gabriel Schneider19 hours
| | | | | | | | | | | | | | a 9ns mount maps back: pardes re-pins cloud9, whose fs.enameErrno knows acme's address and argument refusals The tests' request helper asks cloud9's fs.enameErrno, the rules 9ns turns a refusal's words back into an errno with, for every EINVAL and ENOENT a request is refused with, and fails the test when a mount would hear another (EIO, mostly): "this pane has no text to address" was one. The re-pinned cloud9 maps acme's own refusals, which pardes keeps in acme's words: no match for regexp, address out of range, addresses out of order, range past end of body and the rest. Co-Authored-By: Claude Opus 5.5 <[email protected]>
* Release Pardes 0.21Gabriel Schneider20 hours
| | | | Co-Authored-By: Claude Opus 5.5 <[email protected]>
* addr, dot and limit on a pane with no text are refused `invalid address: ↵Gabriel Schneider20 hours
| | | | | | | | | this pane has no text to address`, which a mount reads as EINVAL, not EIO 9ns maps a refusal back to its errno by its words; "this pane has no text to address" held none it knows. The reference quotes the new words. Co-Authored-By: Claude Opus 5.5 <[email protected]>
* /pager's refusals say invalid, so a mount's shell hears EINVAL: `pager: ↵Gabriel Schneider20 hours
| | | | | | | | | | | invalid directory "relative": must be absolute` 9ns maps a refusal back to an errno by its words, and "the directory must be absolute" and "one directory a write" held none it knows, so a shell through the mount got EIO for both. They now say `invalid`, as every EINVAL here does, and the reference quotes the new words. Co-Authored-By: Claude Opus 5.5 <[email protected]>
* Recipes work through a pane of their own, its look and exec, and keep a ↵Gabriel Schneider20 hours
| | | | | | command pane while their open lasts; the docs give cmd panes, the err grep, alt-screen prompts, /pager's refusals, Tty+ on a ctl, --, and theme failures in words Co-Authored-By: Claude Opus 5.5 <[email protected]>
* A Restore whose ThemeFile fails to load changes no theme: the session goes ↵Gabriel Schneider21 hours
| | | | | | | | | | | | | on showing the one it did, its file still the one named A Restore makes a fresh core, which started at the default theme and asked for the dump's ThemeFile; when that file had broken since, the failure reverted to what the fresh core had loaded, nothing, and ink became orchard. A Restore that asks for a ThemeFile now carries over the theme the old session showed and the file it loaded (and its Theme, unless the dump names one), so a failure there changes nothing. Co-Authored-By: Claude Opus 5.5 <[email protected]>
* A msg record carries what was said, not the message row's clock: `msg 21 ↵Gabriel Schneider21 hours
| | | | | | | | | | dumped themes ...`, no `04:39:41 ` before it Messages a host stamps (saved, reloaded, dumped themes) went into the log with their clock and its double blank. The log has no times; noteMessage records the words after the stamp (Message.body). Co-Authored-By: Claude Opus 5.5 <[email protected]>
* A pane ctl takes Tty+bash as exec and the tag do: a word that takes its ↵Gabriel Schneider21 hours
| | | | | | | | | | argument after a + is no unknown control message there checkBuiltin split a line at its first blank only, so `Tty+bash` was one unknown word on a ctl while exec ran it. It now splits at the + for the words whose registry entry says they take one so. Co-Authored-By: Claude Opus 5.5 <[email protected]>
* /pager refuses a directory that is not there, ENOENT, and a relative one, ↵Gabriel Schneider21 hours
| | | | | | | | | | EINVAL: no +Pager named into a missing directory, none quietly the session's `/(/+Pager` was made for a missing absolute directory, and a relative one fell back to the session's without a word. An empty line still asks for the session's directory. Co-Authored-By: Claude Opus 5.5 <[email protected]>
* A body write to an image or a PDF says its body takes no writes, not that ↵Gabriel Schneider21 hours
| | | | | | | | | | the pane has no text "(a terminal, ...)" A terminal's body takes writes (its keys), so the old words named the one kind that does. data, sel and dirty keep "this pane has no text", which is true of all three. Co-Authored-By: Claude Opus 5.5 <[email protected]>
* Restore and ThemeFile failures say why in words, a ZON one with its line: ↵Gabriel Schneider21 hours
| | | | | | | | | | | | | | never FileTooLarge, ParseZon or a bare `parse zon` Restore read failures named the Zig error (FileTooLarge, NotAFile, ReadFailed), a file that is no dump said `not a pardes dump (ParseZon)`, and a ThemeFile that did not parse said `parse zon`. Messages.errorWords says an error in words (the few that read badly spelled out plainly, the rest spaced and lower case, as reportError did), and zonWhy gives the parser's first error with its line: `not a pardes dump: line 1: expected ...`, `theme file x.zon: line 3: ...`. Co-Authored-By: Claude Opus 5.5 <[email protected]>
* A name under a directory that may not be searched or written is refused, ↵Gabriel Schneider21 hours
| | | | | | | | | | | | | | permission denied, and a Save there says so, not no such directory `pardes noperm/a/b/c` and `pardes /proc/1/root/x` took the missing directory for one Save would make, opened a pane and exited 0; its Save then said "no such directory". fs.deniedAbove finds the nearest directory there and asks whether it may be searched and written: forwarding refuses such a name, exit 1, "permission denied", and a failed Save says "permission denied", which the writer's 9P error carries as EPERM (9ns: EACCES). Co-Authored-By: Claude Opus 5.5 <[email protected]>
* `--` ends pardes's options: `pardes -- -dash.txt` opens the file named -dash.txtGabriel Schneider21 hours
| | | | | | | It said `no such option: --`. What follows `--` is a name, one at most, as for any other positional. Co-Authored-By: Claude Opus 5.5 <[email protected]>
* A shell's prompt drawn on the alternate screen is a prompt: a terminal whose ↵Gabriel Schneider21 hours
| | | | | | | | | | | | | | | | program was killed there is no longer busy for ever htop killed with `sig KILL` leaves the emulator on its alternate screen; bash comes back and draws its prompt there, marks and all, but promptInputReady refused any prompt on the alternate screen, so pty/status said busy 1 and pty/run answered a bare `busy` for good. A prompt mark there now counts. A run sent at such a prompt leaves the alternate screen in the emulator first, as a reused command pane does, so its output lands on the primary screen and is read whole, not `exit 0 cut` with nothing. Busy there with no prompt drawn, and no program name to give, a run is told `busy alternate screen`. Co-Authored-By: Claude Opus 5.5 <[email protected]>
* /index names a command's pane kind cmd, not term: a script picking a ↵Gabriel Schneider21 hours
| | | | | | | | | | terminal by kind no longer picks a pane with no shell A command pane runs one line and takes no typing; a script choosing `$2=="term"` to type into got one. Its kind is now cmd; a shell's stays term. Co-Authored-By: Claude Opus 5.5 <[email protected]>
* A command pane an exec open was answered is leased to it while that open ↵Gabriel Schneider21 hours
| | | | | | | | | | | | | | stays open: another client's command in the same directory gets a pane of its own A finished command pane was the next command's for its directory, from any client, so with four clients scripting one session, one read another's output as its own last run. An open of a look or exec file whose write was answered a pane's serial now holds that pane: reuse passes over it and a new command pane runs the line. The open's own next command may run there again, and once it is closed the directory's one command pane is reused as before. Co-Authored-By: Claude Opus 5.5 <[email protected]>
* Release Pardes 0.20Gabriel Schneider22 hours
| | | | Co-Authored-By: Claude Opus 5.5 <[email protected]>
* Recipes read look and exec on the open they wrote, the reference says a ↵Gabriel Schneider22 hours
| | | | | | fresh open reads anyone's last answer, and /pager is documented Co-Authored-By: Claude Opus 5.5 <[email protected]>
* The docs say Grep is a literal search, that a write's own status comes ↵Gabriel Schneider22 hours
| | | | | | before the log's last err, and that Delcol opens no +Unsaved Co-Authored-By: Claude Opus 5.5 <[email protected]>
* --wait waits on the pane its look answered, read on the look's own open, not ↵Gabriel Schneider22 hours
| | | | | | | | | | | | | one found again by name in /index The forwarded look was a write and nothing more, and --wait then looked its file up in /index. /index shows names escaped (`\\` for a backslash, `\xNN`), so a file named with one matched no pane and --wait exited 0 at once, as if the pane were already closed. The look is now asked on one open (Client.ask) and its answer is the pane waited on; /index by name stays only for a look that answers none. Co-Authored-By: Claude Opus 5.5 <[email protected]>
* pardes - asks /pager for its +Pager and reads the serial on the same open: ↵Gabriel Schneider22 hours
| | | | | | | | | | | | | | another client's exec meanwhile can never take the text into its pane pardes - wrote `pager <dir>` to /ctl, then read /exec on a fresh open, which answers the session's last answer: a client that exec'd in between had its pane overwritten by the paged text. /pager takes the directory and answers the +Pager's serial on the open that wrote it, as /pane/new answers its own open; the ctl verb is gone. Client.ask writes and reads on one open. The README's look recipe reads its answer on its own open too. Co-Authored-By: Claude Opus 5.5 <[email protected]>
* Release Pardes 0.19Gabriel Schneider22 hours
| | | | Co-Authored-By: Claude Opus 5.5 <[email protected]>
* The docs take round 32: petscii words in terminal and image tags, open ↵Gabriel Schneider22 hours
| | | | | | records and EMFILE, ThemeFile failures and replays, trailing-colon looks, pty/run after clear, no-text addr refusals, held log reads at Restore, pardes - in one open, and a detached pty in pixels Co-Authored-By: Claude Opus 5.5 <[email protected]>
* Scripting shows a bash event helper that owns two tag words, writes other ↵Gabriel Schneider22 hours
| | | | | | clicks back and lets event go, and warns that cat event | while read swallows the next click Co-Authored-By: Claude Opus 5.5 <[email protected]>
* A detached session's pty is sized in pixels too, at the last frontend's cell ↵Gabriel Schneider23 hours
| | | | | | | | | | | | size or a nominal 8x16 before one attached forkShell sized a new pty in pixels (ptyWinsize), but the detached server's resize set xpixel and ypixel to 0, and a pane is resized as it is placed, so a program in a detached session with no frontend (yazi) saw a pty with no pixels and drew no images. CSI 14/16/18 t were already answered at the nominal 8x16 cell. Co-Authored-By: Claude Opus 5.5 <[email protected]>
* A log follower hears every record queued before a Restore: the read it holds ↵Gabriel Schneider23 hours
| | | | | | | | | | | | takes all that fit, before the hangup cuts it A follower reads one record per read, and a Restore hangs every client up, so records queued just before it -- the ones a script most wants, what led up to the Restore -- were lost: the read after the first was cut. Before the cut, each held read of /log is answered with every whole record queued that fits it. Co-Authored-By: Claude Opus 5.5 <[email protected]>
* An open past the session's 64 open records reaches a mount's caller as ↵Gabriel Schneider23 hours
| | | | | | | | | | | EMFILE, "Too many open files": pardes re-pins cloud9, whose 9ns maps the words The refusal's words said too many open files, but 9ns matched none of its needles and handed the shell Input/output error. The re-pinned cloud9's 9ns maps them to EMFILE. selfmount opens the root exec for writing eighty times through the mount and hears EMFILE at the 65th. Co-Authored-By: Claude Opus 5.5 <[email protected]>
* addr, dot and limit on a pane with no text, a terminal, an image or a PDF, ↵Gabriel Schneider23 hours
| | | | | | | | | | | are refused alike: "this pane has no text to address" A terminal took an addr write and evaluated it over nothing, saying "bad address syntax" or "address out of range" by the expression, while dot refused it as no text at all. A range file of a pane with no text now refuses every write the same way, before the expression is read. Co-Authored-By: Claude Opus 5.5 <[email protected]>
* A size refusal quotes the bad value, "5 2", not the word sizeGabriel Schneider23 hours
| | | | | | | | size quoted its own word in every refusal, as bad values were quoted before round 31's change for the settings. The value is the offending word here too, and is now what each size refusal quotes. Co-Authored-By: Claude Opus 5.5 <[email protected]>
* pty/run after a clear answers what the command printed from the clear on, ↵Gabriel Schneider23 hours
| | | | | | | | | | | | whole; cut only when its start really scrolled away A clear (ED 2, ED 3, a reset) under a running command marked its output lost, so `clear; ls` answered only "exit 0 cut". The output's start is now pinned again at the screen's top after the erase: what the command prints from there is its answer, whole, and a cut is said only when that start leaves the scrollback. Co-Authored-By: Claude Opus 5.5 <[email protected]>
* A look line's trailing colon is dropped, as a click leaves it off, and a ↵Gabriel Schneider23 hours
| | | | | | | | | | | | look whose address fails never leaves the file it opened `a.txt:2:3:`, as a compiler prints a place, failed as bad address syntax, and still opened a.txt: the file was opened to read the address in, and stayed when the address failed. The trailing colon is now no part of the place, and a pane opened only to evaluate an address that fails goes again, the miss said from the pane looked from. Co-Authored-By: Claude Opus 5.5 <[email protected]>
generated by cgit v1.3 (git 2.53.0) at 2026-10-02 04:48:31 +0000