| Commit message (Collapse) | Author |
|---|
|
OSC 5577 only works for programs that know slosh exists. The program
status protocol is the same idea written down for everybody, by ghostty's
author and implemented by ghostty and Rex, so a program that already
speaks it needs to know nothing about us:
printf '\033]7501;state=working:app=cargo:msg=YnVpbGRpbmc=\033\\'
Implemented *as well as* 5577, not instead of it. 5577 is a two-way
control channel -- buttons answer on the program's stdin, shaders are
refused with a reason -- and 7501 deliberately is not, so neither absorbs
the other. What 7501 buys is the programs we do not control.
The scanner becomes one pass for four things: 5577, 7501, OSC 133 ; A and
RIS. The last two are here because 7501 ties record lifetime to them -- a
new prompt or a full reset is how a stale `working` dies without a
heartbeat -- and because a pane should pay for that scan once. osc5577_fn
gives way to osc_scan_cbs_t.
Reports are all-or-nothing, which is the spec's rule: no state, an unknown
state, a broken limit, base64 that will not decode, a control character in
decoded text or a malformed id discards the whole report rather than
writing a partial one. A malformed *pair* is skipped and the rest parsed.
A pane keeps up to PANE_PS_MAX records with LRU eviction, and composes the
one line it has from them by precedence -- blocked, error, done, working,
idle -- resolving `app` from the nearest ancestor that has one, the way
the hierarchy is allowed to mean something. It lands in the slot 5577's
`status` already had, because every reader of a pane wants "what is this
doing", not "which protocol said so"; `panes` reports program_state,
program_kind, program_app and progress beside it, with -1 for a progress
nobody reported, since a bar drawn at 0% says something false.
The feature-detection reply is rate-limited here. Note for the next
re-vendor: upstream lib-vt implements 7501 too, newer than our pin, and
answers the support query itself through write_pty without rate-limiting
it. See "What upstream did" under D1 in DESIGN.md before swapping this
half of the file for GHOSTTY_TERMINAL_OPT_PROGRAM_STATUS.
|
|
`tab_bar_width` was a reservation: the configured columns came off the
screen whatever the screen was, and the only fallback was all the way back
to the top row. On a narrow terminal that is chrome displacing the thing it
is chrome for, and the jump to the top moves every target you were aiming
at.
It is a ceiling now. app_tab_bar_cols takes the app, because the answer
depends on the terminal and not only on the config: held to the columns
left after a pane's collapse width, and to a third of the screen, and
narrowed rather than dropped. TAB_BAR_FIT (14) is the narrowest width this
code will invent on its own -- a config that asks for less is taken at its
word -- and 0 means there is no room for a sidebar at all, which is the one
case that still falls back on top.
Height counts too, which it did not before: a sidebar spends a row per tab,
so a wide but short terminal runs out of list before it runs out of screen,
and a tab with no row is not clipped, it is unreachable. tab_bar_min_rows
is that floor. Not when the top strip could not seat them either -- moving
them would gain nothing, so the side you configured stands.
All of it is re-derived per frame, like the layout, so a resize moves the
bar and nothing remembers that it did.
test_tab_bar also stops depending on where the suite is run from. An
unnamed tab borrows the basename of its focused pane's cwd, so eight checks
read `tests` under `make test` (which cds there) and `slosh` from the
repository root, where they failed. They are about padding, alignment and
the index, not about that derivation, so they name the tab first -- by id,
since `set-name` takes 0 for the focused *pane* only. 147 checks, green
from either directory.
|
|
A tab's focus outlived the tree it pointed into. close_leaf's last-pane
branch freed the leaf and removed the tab but never cleared `t->focus`,
and tab_remove only shifts the slots *after* the one it removes -- so with
the last tab gone, `a->tabs[0]` kept the stale tab_t while `ntabs` was 0
and `cur` was 0. `cur(a)` is `&a->tabs[a->cur]` with no bounds check, and
`a->quit` is a request rather than a barrier: the loop keeps serving
commands, and the next one to ask for the focused pane was handed freed
memory.
From outside it looked like a random crash, because what the freed chunk
held depended on what had been allocated since. Two from one afternoon: a
pane_t recycled into a `panes` JSON reply, whose `term` pointer was then
the ASCII of `k":true,` and faulted as a general protection fault; and the
same node read back zeroed, which left pane_clear_bell testing a flag at
NULL+0x5d. Typing `exit` in the last pane is the whole recipe.
close_leaf drops the focus with the tree
tab_remove blanks the slot the shift vacated
app_compose tests ->pane as well as ->focus, the way the search
beside it already did
app_apply_layout clears restore_focus up front -- it is a node pointer,
and an apply that fails before consuming it leaves it
naming a tree the next apply frees
Two more the same reading turned up, both in close_leaf: an emptied split
set the focus to the node just freed, and a leaf whose parent disagreed
with that parent's child list underflowed `p->nkids - at - 1` into a
memmove of nearly SIZE_MAX.
FOCUS_CHECK -- behind -DSLOSH_FOCUS_CHECK, compiled out otherwise -- is
the invariant all of this turns on: every tab's focus is a leaf of that
tab's own tree. It is what found this one.
Committed with --no-verify: the files are deliberately partly staged, and
the staged blobs were checked against clang-format directly.
|
|
Applying a layout landed on the first tab it made, and moving a pane into a
tab took focus there. For a program building a pane in the background that
is two failures at once: it yanks the view off the human mid-sentence, and
it leaves the *next* caller to read focus as "where I am" and put its work
in a stranger's tab.
`apply-layout focus:false` builds the tabs and keeps looking at whatever was
on screen; a layout naming an `active` tab still wins, because that is a
statement about where to be. `move-pane` gains `beside` (which pane in the
destination to split, rather than whatever it has focused) and `focus:false`
-- the two things a caller placing a pane next to *itself* needs.
That uncovered a sizing bug that was unreachable before: a pane's size comes
from a layout pass, and a tab nobody has looked at never had one, so a pane
built in the background stayed 1x1 and its program drew into a single cell.
The jump is what used to size it. layout_tab() gives a background tab a real
pass, borrowing the tab cursor for its length -- nothing composes, reads
input or draws a frame in between.
|
|
`raw` wrote into the focused pane and `snapshot` composited the active tab,
so driving a pane anywhere else meant focusing it or selecting its tab
first. Both of those move the view of whoever is watching, and both lose to
anybody doing the same thing at the same time.
`raw` now takes an `id` (0 is the focused pane, as before), and `capture`
answers one pane's viewport as text -- from the pane's own terminal, so it
works for a pane in a background tab, changes no focus, and leaves the
human's selection alone. It builds its selection as a snapshot rather than
setting one on the terminal, because a read with a visible side effect is
not a read.
`capture` tells "no such pane" apart from "that pane is empty": a caller
polling a dead id would otherwise wait forever on a pane that is quiet
because it is gone.
|
|
A program in a pane could name every pane but its own. `panes` says which
one is *focused*, and focus belongs to the human, so anything that inferred
"where do I live" from it did its work in whichever tab somebody else
happened to be looking at -- and lost a race with anybody else asking the
same question.
So export the id into the child, the way tmux has $TMUX_PANE and zellij
$ZELLIJ_PANE_ID. The id is taken before the pane is made, because pane_new
spawns the program and assigning it afterwards is one instruction too late;
it lives on the pane rather than the layout, so a rerun keeps it -- an
identity that changed under a rerun would be no identity at all. Unset, not
zero, when there is none: a stale value inherited from an outer session
would read as this one's.
`panes` also reports each pane's pid now, which is the same question
answered for anything that cannot read its own environment: walk your
parents up until one of them is on that list.
|
|
It was drawn in tab_status_busy, the colour of the words it sits beside, which
made one decision out of two. The words have to stay readable at a glance and
a mark does not: a theme may well want the glyph loud and the text calm -- or
the text in the accent and the mark a quiet dot -- and with one entry neither
is available without moving both.
Unset it follows tab_status_busy, derived after it and from whatever that
resolved to, named or derived, so nothing changes for anyone who does not ask
and no theme in contrib needs a line. `--dump-theme` writes both, like every
other colour the compositor draws.
|
|
A status is a line of text, and a line of text cannot say whether it is still
true. "building 3/7" reads exactly the same while it builds and ten minutes
after it stopped, which is the one thing somebody reading it from another tab
wants to know -- and the only workaround available to a program was to write
the answer into the words and keep rewriting them.
So `busy`:
printf '\033]5577;1;status;make test\033\\'
printf '\033]5577;1;busy;1\033\\'
...
printf '\033]5577;1;busy;0\033\\'
A flag of its own rather than a field of `status`, because a status is the
*whole* payload after the verb -- `status;a;b;c` has always been the text
"a;b;c" -- so anything added to that line would change what every existing
sender means, including the five pi extensions D1 promises byte compatibility
to. Orthogonal, so a program can leave one line up and toggle the other;
anything but 0/false/no/off is on; `clear` takes both; and a pane whose program
has gone is never busy, whatever it last claimed, because a spinner turning
under a dead tab is a lie you keep looking at.
In a sidebar the row turns `tab_status_busy` -- the theme's accent, derived
like the other status colours, so no theme needs a line -- and takes a spinner
in the column `tab_bar_padding` was already holding blank, so nothing reflows
when a program starts or stops working. `busy_mark` is the frames, separated by
spaces: one frame is a static mark and costs nothing, several make it turn, and
only then does the session keep a frame clock -- via the same `animating` flag
an animated shader uses, so the cost lands only while such a row is actually
composed, and `""` drops the mark entirely and leaves the colour to say it.
The frame is a pure function of the clock, like every other pulse here, so two
clients attached to one session show the same one.
`panes` now reports `status` and `busy`, which is the scriptable half of the
same fact: what a pane says about itself, without reading its screen.
|
|
`padding` is a pane's: the ring between a frame and the text it holds, in rows
times gap_aspect so a `1` looks square around a block of prose. Handing the
same knob to the sidebar looked tempting and is wrong twice over. The sidebar
is chrome, not a pane, and its width is fixed by tab_bar_width -- so `padding
1` would quietly eat four of a sixteen-column strip (two cells a side, aspect
included) and shorten every tab name, because somebody wanted their terminal
text to breathe. And its two axes are not one idea: the columns are an indent
for labels, the rows are air above a list.
So tab_bar_padding, its own knob in the family the rest of the sidebar's air
already lives in (tab_bar_pad, tab_gap, newtab_pad), taking the same 1/2/4
CSS-order values as `padding` -- one reader for both, so they cannot drift
into meaning different things -- but counted in cells.
The default `0 1` is what the sidebar already drew: that one column is where
every label's leading space has been hard-coded since the sidebar existed, and
moving it into a setting is what makes `tab_bar_padding 0` possible at all --
names flush against the frame, which was not previously a thing anyone could
ask for. The statuses, the `+ new tab` row and the pane count follow the same
indent, since they were each carrying their own copy of that space.
It is air *inside* the plate: a tab's fill still spans the whole row and the
whole row is still the hit, because the row is what you click and a highlight
that stopped short of the frame would read as a highlighted word rather than a
chosen row. The bottom rows are given up when the strip is short of them, like
every other piece of air in here.
tab_bar_pad is the older name for the top component alone. It still works,
still means exactly that, and wins where a config writes both -- it is the
more specific of the two. The dump writes tab_bar_padding and mentions it.
|
|
`1:api` puts a number at the start of every row, and in a sidebar that is a
column of digits in front of the thing you are actually reading -- the names
start one, two or three cells in depending on how many tabs you have, and the
numbers are only wanted at the moment you reach for C-a 2.
tab_bar_index "right" moves them:
│ tests 1 │ │ 1:tests │
│ verylongproj 2 │ vs │ 2:verylongproje│
│ api 3 │ │ 3:api │
Every name starts in the same column and the numbers line up in one of their
own, which costs a column of sidebar and buys the list a shape. "prefix" stays
the default, because it is what the strip has always drawn and the strip is
where most people meet it.
Sidebars only, and not by omission: a tab in the top strip is exactly as wide
as its label, so there is no far end to align against and a number floating
after a name it does not belong to would be worse than the prefix it replaced.
The setting is ignored there rather than half-honoured. A row too narrow to
spare the columns (under the number plus five) also keeps the prefix, since
the name is the part you came for, and the rename editor keeps the whole cell
as it always did.
A name long enough to reach the number stops one cell short of it, because
`verylongproje2` is a tab called verylongproje2 as far as any reader can tell.
The cell comes back as padding, so the plate is unbroken and the hit is still
the whole row -- clicking the number is clicking the tab, and the number
brightens with the label under the pointer rather than leaving half a row lit.
The bell mark keeps its place between them, budgeted first as before, so a
long name cannot eat either of the two things that are not the name.
|
|
`theme_name "sl0p"` with all seven themes sitting in /usr/share/slosh/themes:
"no theme named sl0p". The lookup knew two places -- your own themes directory
and the PREFIX this build was compiled with -- and on that machine the first
did not exist and the second was /usr/local, because that is the Makefile's
default and the package was built with PREFIX=/usr. Neither was where the
themes actually were, and nothing in the search could ever have found them.
A compiled-in prefix is a guess about where the binary will end up, and the
binary is the one thing that knows where it really is. So the search now asks
it: path_self() (/proc/self/exe, _NSGetExecutablePath, GetModuleFileName --
lifted out of server.c, which was already doing this for SLOSH_BIN) gives
<bin>/../share/slosh/themes for an installed slosh wherever it was installed
to, and <repo>/contrib/themes for one being run out of a build tree, which
makes a checkout able to name a theme without installing anything first. The
compiled-in prefix stays as the last candidate, for a binary sitting somewhere
that says nothing about its data, and duplicates collapse so the usual install
does not report the same directory twice.
Your own directory is still first and still shadows everything, which is the
property that matters; the rest is about the shipped set being findable at all.
The tests that assumed a closed world of two temp themes now narrow to their
own names by typing, since the shipped ones are (correctly) always in the list
-- and the "no themes anywhere" toast lost its test, because with the shipped
set reachable that state takes a binary with nothing beside it. The case that
started this is a test of its own.
|
|
`theme_name` is a line in a file, and a name in a list tells you nothing about
a colour scheme. So the picker previews by *applying*: moving the selection
puts that theme on, which is the only honest preview of a palette meant to
dress a whole session -- a swatch in a list would be a small lie about it, and
the session is already on screen behind the box.
It opens on the theme being worn rather than at the top, so the first arrow
press moves off it instead of silently changing the session, and the row your
config names says so, because "which one is this" and "which one comes back on
restart" are two different facts the moment you switch.
Escape puts back what was worn, since a picker that changed your session by
being looked at is one you stop opening. Enter keeps it for the session and
says as much. Only C-s writes `theme_name` into your config, which is the one
place in the UI where a keystroke edits it, so it is a deliberate chord and
the footer says so: "C-s keeps it, esc undoes".
Reuses the picker the finder, the palette and the projects list already wear
-- one machine with a subject, so the query, the selection, the mouse, the
frame and the scroll behaviour are the ones already tested. The list is
scanned on the keystroke, like the projects are: the moment it opens is the
moment it has to be right, and a theme written since the session started is
exactly the one being looked for. With nothing to list it toasts the directory
it looked in rather than opening an empty box.
`t` was free: tabs cycle on tab, and new-tab has had `c` since the first day.
|
|
Wearing a theme meant pasting its `theme { }` block into config.kdl, or
including a path to a file you had copied there yourself. Both make the
palette a literal in your config, and a literal is the one thing you cannot
change cheaply: not from a script, not from a running session, not by a list
of what is available, and never again from upstream -- the colours you pasted
are the colours that theme had the day you pasted them.
So a theme is a file in a directory now, and a config names it:
theme_name "phosphor"
theme { frame_focus "#00ff88" } // ...but that one colour is mine
theme_name is applied after the includes and before everything the file says
itself, which is the rule includes already follow and for the same reason: the
loader reads a document by asking it for names rather than walking it in order,
so "here" is not a position it could honour, and "what you asked for is the
base, what you wrote is yours" is the rule worth having.
Two directories, yours first: `themes/` beside the root config -- beside the
file that was loaded, so a config somewhere else (SLOSH_CONFIG, or one handed
to --check) takes its themes with it -- and then <prefix>/share/slosh/themes,
compiled in from PREFIX. install(1) has been putting themes there all along
and nothing could read them; now a shipped theme is nameable without being
copied, and a file of yours with the same name shadows it, which is how a
root-owned theme gets edited.
A name, not a path. The directory is what makes a theme switchable, listable
and shadowable, and a path silently does none of those, so `theme_name
"../x.kdl"` is refused with a line pointing at `include`, which still works
and is still the answer for a file somewhere a directory of names cannot
reach. A theme that is not there is one complaint and no more (D9): the rest
of the config applies, because losing your keybindings over a mistyped colour
scheme would be the wrong trade. A theme file is read by the same loader, so
it is watched like any other: saving one reloads the session wearing it, and
the derivations (tab_status_fg, tab_idle_bg) run over the theme's own colours
exactly as before.
Switching without editing anything, over the socket a session already has:
slosh -s main cmd theme # amber default* ...
slosh -s main cmd '{"cmd":"theme","name":"amber"}' # now
slosh -s main cmd '{"cmd":"theme","name":"amber","save":true}'
The session holds the name over later reloads, so saving your config for an
unrelated reason does not put the old colours back; `save` writes the
theme_name line into the file, replacing the one already there and leaving
every other line and comment alone -- dumping the whole config back over
itself would reformat a hand-written file and drop its comments, which is not
a fair price for changing a colour scheme. A name is tried on a throwaway
config first, so a typo costs nothing. An empty name gives the answer back to
the file.
--dump-config stops writing colours: it writes the theme's name and the
colours a config overrode by hand, so a dumped config keeps following its
theme instead of freezing it. --dump-theme is the other half and is new: the
resolved palette as a theme file, which is how you start one from the session
in front of you rather than from an empty file. It is also what keeps
contrib/themes/default.kdl honest -- that file is now the annotated reference
(the commented palette moved there out of config.kdl, where it was 119 lines
of colour in the middle of the prose), and a test compares it against the
compiled-in one.
While in there: tests/test_themes.py's "every theme sets every colour" guard
had been grepping for `&c->`, the spelling before the colour table moved to
offsetof, so it had been finding zero colours and passing for months. Fixed,
with the derived and opt-in colours exempted for the reasons they are derived
and opt-in, and with a check that the guard found a table at all.
|
|
Seen next to the accent, the first plate was too polite to do the job it was
added for: a 30% lift off the background under labels in the same #45454a the
idle frames use. From another monitor the strip still read as one dark band
with a blue block in it.
Both ends move. tab_idle leaves the ambient chrome grey and becomes #8a8a95,
because a tab you are not in is a place to go and has to be legible at a
glance, where a dimmed pane border only has to be findable. The plate follows
it to 43%, which is not a taste constant but a contrast ratio: the same step
taken from a dark background upward and a light one downward lands at about
1.9:1 either way, so paper's #bdbebe on near-white and the default's #3b3b40
on black are the same amount of plate.
The six themes in contrib named their idle tab colour as their frame grey, so
each gets the same move in its own family -- 40% of the way from that grey to
the colour the theme writes text in -- and their plates follow by derivation:
amber #a67c26, mono #909090, paper #666a6e, phosphor #4a9c5d, slate #70789c,
sl0p the house #8a8a95. Nobody who named a colour loses it; the derivation
only ever fills in a theme's silence.
|
|
The active tab has been filled with the accent since the strip existed; every
other tab was words on the background. That leaves the strip with one shape on
it and nothing to compare it against -- where a tab ends, how far the row you
are about to click reaches, and in the sidebar which rows are tabs rather than
statuses, all had to be read off the text.
tab_idle_bg fills them. Barely off the background and in the label's own grey,
because this is not a second highlight: it gives a tab edges, and it gives the
accent something to be brighter than, which on a strip of short labels is most
of how the eye finds the tab it is in. Hover keeps it for the same reason the
active tab keeps its fill -- losing the plate under the pointer would read as
the tab going away -- so hovering still lands on the text, and the bell mark
sits on the plate rather than punching a hole in it.
Derived a third of the way from default_bg to tab_idle when nobody names it,
the same bargain tab_status_fg and tab_status_stripe already make: both ends
are colours every theme in contrib defines, so each gets a plate in its own
family -- paper lightens, phosphor greens, mono stays grey -- without a line
added to any of them. Naming it wins, and naming your own background is how
you ask for the bare tabs back.
The `+ new tab` button and the gaps between tabs stay unpainted: the button is
the one row in the list that does not mean "go here", and tab_gap is air
between plates, which is what makes them read as separate tabs at all.
|
|
A tab carrying three rows of wrapped status and the tab under it carrying
three more read as one column of text: the labels are in there, but the list
has stopped looking like a list. tab_gap is the blank rows between them, 0 by
default, which is exactly what the sidebar has always drawn.
It is the space *between* tabs and nothing else. Above the first tab is
tab_bar_pad's and before the button is newtab_pad's, so the gap is applied
before each label except the first and none of the three is ever paid twice.
Given up a row at a time when the list is short of room, ahead of every label
and the button, which is the same order the pad and the statuses already
follow: a gap that pushed a tab off the bottom would be spacing costing you
the thing it spaces. Paint rather than layout, so tuning it repaints without
reflowing any pane. Sidebars only -- the top strip is one row, and the space
between tabs there is the label's own padding.
|
|
Sixteen cells is not enough for a sentence, and a sentence is what a pane
has to say: "add a summary…" is most of a status and none of its meaning.
tab_bar_status_lines 3 word-wraps instead of cutting, and the same column
carries forty-eight cells rather than sixteen.
Wrapped, the rows under a tab stop being a list of panes and become a
paragraph each, so the shape no longer says where one pane's status ends and
the next begins -- both are italic text down one column. Alternate panes are
drawn on alternate bands to say it instead: tab_status_stripe, derived from
the theme like tab_status_fg, so a theme that has never heard of any of this
still gets a band in its own family. Nothing is painted at one line, where
there is no ambiguity to resolve and a band would take a translucent
terminal's background away for nothing.
1 is the default and is the old behaviour exactly, down to cutting at the
column rather than the last word before it: with no next row to move the word
to, breaking early would spend cells on nothing. The reason to prefer a word
boundary only exists once there is somewhere for the word to go.
Rows count against tab_bar_status as before, so wrapping trades panes for
words within the same allowance, and the allowance itself still gives way to
the tabs. A status that will not fit whole is dropped whole and reported with
the ellipsis, rather than shown as a fragment that reads like the whole
thing. Every row of a status is the same `find:` door.
tabstatus's buffer goes to 256 to match pane_status's own slot: at one line
the drawing cut it anyway, but wrapped it is the whole string that has to
survive the trip.
|
|
The `+` is the one entry in the strip that does not mean "go here", and it
sits at the end of a list where every other entry does. Flush against them it
is what you hit when you meant the last tab -- most easily in a sidebar, where
the target is the full width of the row and the pointer only has to be one row
low. newtab_pad is the nothing in between: blank rows in a sidebar, blank
columns on top, one by default, which is enough to make the slip a miss.
It is paint rather than layout, like tab_bar_pad: the panes' rectangle is the
same at 0 as at 5, so tuning it repaints without reflowing anyone. And it is
given up before the button is -- a strip short of room drops the spacing one
cell at a time and still draws what the spacing was protecting, because
spacing must never cost you the thing it spaces.
newtab_button false removes the button outright, for people who make tabs from
the keyboard and would rather have the row (or the columns) back. Only the
place to click it goes: the chord and the palette entry still work. Setting
newtab_mark "" already did this by omission; this says it on purpose, and says
it in --dump-config.
|
|
Three columns of indent is a fifth of a sixteen-column sidebar, spent
saying 'this belongs to the row above'. The slant says it in none: a
status now starts in the same column its tab's name starts in, and is
set in italic. That also survives the monochrome themes in contrib,
where a colour-only distinction would have had nothing to spend.
tab_status_fg / tab_status_bg are there for themes that do have colour
to spend. The fg follows tab_count when a theme does not name it -- the
same declared/derived split apply_scrolled makes -- so the six themes
in contrib get coherent status rows without a line added to any of
them. The bg is unset by default, because a band is opt-in and 'none'
is what a translucent terminal wants; the row is padded so a band that
*is* set covers the row rather than the words on it.
Two things fell out. The padding loop existed three times and is now
pad_cells. And cb_color wrote #000000 for an unset colour, so
quietly pinned attach_bg -- 'gray
ink on the terminal's own background', says its comment -- to a black
band. Unset colours are now written commented out, the same trick the
states block uses for what it wants re-derived rather than pinned.
|
|
(tab_bar_chrome)
A list of tabs sitting in bare columns beside a framed layout reads as
text that wandered in from outside. So the sidebar gets a frame of its
own -- and in compact mode not its own: its inner vertical lands on the
tab area's ring column, and because the lines are strokes that union
with whatever is already in the cell, the corners become tees without
anyone computing a junction. The sidebar and the layout come out as one
figure, which is the whole point.
Classic mode gets a free-standing frame in the idle frame colour,
rounded or square with everything else. The frame costs a column each
side and a row top and bottom, taken from the sidebar's own budget --
and dropped entirely, before the list is, when the sidebar is too small
to afford it: chrome must not displace what it is chrome for.
The content box (cx/cw) now feeds the labels, the status rows, their
hits and the truncation budget alike, so the frame cannot disagree with
where the clicks land.
|
|
Panes already announce a one-line status over OSC 5577, and it rendered
in exactly two places -- both inside the pane's own frame, which is the
one place you are not looking when you most want it. The sidebar has
the vertical room the top strip never had: under each tab, up to
tab_bar_status rows of what the panes inside said, in tree order, dim
like the pane count. A dead pane's row says how it died instead
(exit_words), the same substitution its own frame makes.
The cap's last row is spent on an ellipsis rather than one more status
pretending the list is complete, and a status cut to the sidebar's
width ends in one too -- 'building 3' is a different fact than
'building 3/7'. Each row is a find: hit, the verb that already means
'go to this pane', so clicking a status jumps to the pane that said it,
across tabs. Freshness costs nothing new: an OSC status marks the pane
dirty, the same path the bell indicator rides out of background tabs.
|
|
One blank row above the first tab by default: a list starting in the
very corner reads as cramped, and in compact mode the row lines the
first tab up with the pane content inside the ring rather than with the
ring's top line. The pad is where the drawing starts, not a layout fact
-- the panes' rect is identical at pad 0 and pad 5, so tuning it
repaints without reflowing anyone.
The bottom indicators (pane count, prefix badge) move in one cell to
sit under the labels' own leading space; flush at the edge column they
read as misaligned on the left and collide with the compact ring's
corner on the right.
|
|
A sidebar lists one tab per row down the chosen edge, tab_bar_width
columns wide; the panes trade the strip's row for the sidebar's columns
in tab_area(), the one place that arithmetic lives. The per-tab cell
drawing is extracted from draw_tab_strip (label, rename editor, bell,
hover, drop states, hit) and shared, so every mouse verb -- click,
double-click rename, drag to a tab, drop on + -- works in both shapes
because the hits move with the cells.
The strip's indicators (pane count, prefix badge) take the sidebar's
bottom rows, reserved first under the same budgeting rule the strip
applies from the right. The status line below keeps the full width. A
terminal too narrow to give up the columns falls back to top, re-derived
per frame, so a resize moves the bar without anyone storing state.
|
|
C-a / opens an inline search bar on the focused pane's status row, on
lib-vt's GhosttySearch (D4's pattern: the scrollback walk, reflow and
alt-screen survival are the library's, the UI and the wiring are ours).
Type to search live; the viewport follows the current match; up/down
(or C-n/C-p) step through them with wraparound; Enter keeps your place,
Escape snaps back. A k-of-n counter reads on the right, or "no matches".
Every match on screen is tinted, and the one the bar is on is tinted
apart -- warm gold by default, painted after the shader pass so the
scrolled wash does not dilute a match you are hunting for. Four themeable
colours (search_fg/bg, search_cur_fg/bg).
Bare / took the key from help, which keeps ?; that needed the legacy
decoder to read the shift out of a punctuation byte the way it already
did for capitals -- without it ? and / decode identically (SLASH, no
mods) and cannot be two keys on a client without the kitty protocol.
tests/test_search.py drives it end to end (39 checks): the bar, the
counter, stepping and wrap, all-match vs current highlighting, Enter/
Escape, new output updating the count, case-insensitivity, the keyboard
belonging to the bar, click-away, and that / and ? no longer collide.
|
|
Overlays are painted over the panes and the compositor needs no more
than that: painted later is on top. A placement is not painted, though
-- the client's terminal draws the image over the cells, in whatever
order they arrived -- so an overlay that is only drawn is an overlay an
image sits on top of. The cheatsheet came up under a picture and was
unreadable through it, in the terminal and in the mirror both.
So a frame now records where its overlays landed, and the graphics pass
takes them out of a placement exactly as it takes out a float: the
modal's box (the cheatsheet and the picker wear the same frame), each
toast, and the splash. The list is rebuilt every compose, so a rect that
is not registered again is one that is no longer on screen.
Only the opaque things are registered. The scrim behind a modal tints
what is behind it rather than hiding it, and an image cannot be tinted
by us at all -- so a picture beside the cheatsheet stays bright while
the text around it is pushed back, which is wrong-looking but a great
deal better than the image disappearing while a modal is up.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01CRz3zwZJvBvzsR9k4cj6dd
|
|
Occlusion for placements is clipping: the cell compositor gets it free
from paint order, but images are sent after the diff and drawn over it,
so what a float covers has to be taken out of the placement itself. A
clean edge came out as a smaller rectangle; anything else -- a float
across the middle, a float on a corner, a float wholly inside -- was
refused, on the grounds that the remainder is an L and a placement is a
rectangle. The image vanished, all of it, the moment a floating pane
crossed it, on the terminal and in the mirror alike, and came back when
the float moved off.
The remainder is not one rectangle, but it is always a few: the strips
above and below what is covered, and what is left beside it between
them. So rect_subtract() cuts the visible remainder into rectangles and
each is placed in its own right, with the source rectangle the same
arithmetic the pane edge already used -- generalised to a cut on any
side, and to a first piece that keeps the sub-cell offset while the
others start on a cell boundary. Eight pieces is the ceiling (two
floats' worth); past it the old refusal stands.
The pieces of one placement need an id each, and the program's own
placement id is one number that cannot be several, so the model now
hands out the placement ids the client is told and keeps the program's
as the identity it was. `graphics` reports the piece alongside it.
Checked against a float parked over an image: the picture is cut around
the float, the strips tile exactly the image minus what the float
covers, and their source rectangles land where their cells do (a float
inside a 40x10-cell natural-size image gives 320x32 above, 320x32
below, 56x96 left and 72x96 right).
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01CRz3zwZJvBvzsR9k4cj6dd
|
|
Every frame re-emitted `a=p` for every visible placement. On a terminal
that is invisible: a kitty placement is a standing instruction, not a
per-frame one ("the other commands to erase text must have no effect on
graphics"), so a repaint of the cells under a picture does not disturb
it and saying it again says nothing new.
It is not free anywhere else. A mirror cannot know that the identical
placement is identical: shellglass re-decodes and re-encodes the image
on every one, holds the placement back while that runs, and -- because
an image change rides only in a full frame -- republishes the whole
screen each time. Measured on one image and a pane printing text for
twenty seconds, the mirror published 430 full frames and re-sent the
image list 215 times; 215 of those fulls carried no image at all,
because the placement was still mid-decode. That alternation is what
was visible in the browser: a picture that flickered while anything on
screen moved and vanished when it stopped. The same twenty seconds now
costs one full frame, 412 diffs and one image list.
So a placement is emitted only when it is new or has actually changed,
and the client's state records what it is holding. Three things have to
be right for that to be safe: a dropped frame un-marks the placements it
carried (they are owed again, exactly as transmissions and deletions
already were); a retransmit of the pixels invalidates the placement,
because kitty deletes an image's placements when its data is replaced;
and a full repaint clears the screen, which takes the images with it, so
every placement is owed again -- gfx_repaint(), not gfx_reset(), because
the pixels the terminal was sent survive a clear and a deletion still
owed from a dropped frame is still owed.
The live check drives a picture that does not move while the screen
around it repaints: one placement across 121 frames, where the old
behaviour emitted 120.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01CRz3zwZJvBvzsR9k4cj6dd
|
|
An updated GhosttyRenderState retains a full copy of the pane's viewport
(rows, cells, per-row arenas, styles) so renders can read it without the
terminal lock. Measured with bpftrace and a lib-vt micro-benchmark, that
copy is ~140KB for an 80x24 pane -- more than twice the terminal it
mirrors (~57KB) -- and every pane kept one forever, composed or not.
Only the current tab is ever composed, so app_compose now drops the
snapshot of every pane on the other tabs (a pointer test once dropped),
and pane_new no longer creates one at all: pane_compose and
pane_select_word build it on first use. Both already begin with
ghostty_render_state_update(), so a dropped snapshot is rebuilt, never
missed, and the tab switch already forces a full repaint.
Server memory per empty 80x24 terminal drops from 154K to 66K anon
(Superlogical reports 68K on macOS; tmux 4K, zellij 1.2-6.2M). Fifty
background tabs: 8.2M -> 3.8M, reclaimed on close, and cycling through
tabs does not ratchet. What remains is the lib-vt terminal itself.
|
|
The lookup miss now writes the embedded entry to ~/.terminfo and
uses it, so a bare box gets the full TERM with no step at all. An
entry that exists anywhere is respected -- possibly newer, certainly
somebody's -- and a HOME that cannot be written leaves the
xterm-256color fallback standing. --install-terminfo stays as the
by-hand spelling; the writer moved to pty.c where the panes are.
|
|
The +x+y offset goes -- 'N clients' says the thing that matters. The
inverse-video chip goes with it: the tag now draws in attach_fg /
attach_bg from the theme, defaulting to gray ink on the terminal's
own background. Standing information, not an announcement.
|
|
A small tag in each client's own top-right corner when several
terminals share the session: the viewer count, and on a view that is
a crop being panned by somebody else's cursor, the crop's offset into
the shared screen ("2 clients +12+4") -- which is exactly the moment
a screen moving on its own needs explaining. Stamped per client after
projection, because every terminal has its own count of others and
its own offset; the shared status bar is composed once and can be
cropped out of a panned view anyway. attach_indicator false hides
it.
|
|
The patch made shared attach unconditional; these make it a policy.
multi_attach false restores the classic displacing attach
(EXIT_REPLACED, which old clients already understand). size_follows
picks whose size the shared screen takes: active (the patch's rule),
smallest (the tmux rule: filler, never a crop), or largest. The size
fold lives in apply_canvas, so any attach, resize or departure can
move the canvas under the folding policies -- including a non-active
client leaving, which under smallest may have been the very client
holding the floor down. A config reload re-applies both knobs to the
running session.
|
|
Last input wins, canvas size is adjusted to active connection.
|
|
The static musl release builds cannot dlopen -- musl compiles it out
-- so loadable shader plugins were a feature most shipped binaries
could not use. Nothing we ship used it either: every shader in
contrib/shaders is an expression. Gone: the dlopen/glob loader in
shader.c, shader_dir in the config, the plugin ABI table and entry
macro, the Windows dl*/glob shims that existed only for it, the
contrib/shader-plugin skeleton, its tests, and every doc mention.
shader_abi.h stays as the shader model (cell_t, shade_ctx_t,
shader_t) with a note about what it used to be.
|
|
docs/layouts.md always said an absent coordinate takes the centred
default; the loader said 0, and a layout float with only w= h= landed
in the top-left corner. Now an axis the file does not name is
FLOAT_CENTRE, a sentinel no real coordinate can reach, and float_clamp
answers it with the middle -- every frame, at whatever size the
terminal is by then, until an explicit move or resize writes a real
number over it. The webdemo's welcome float is the customer: it must
sit centred on a terminal whose size the layout file cannot know.
The dump writes a still-centred axis by omission, because the number
it happened to land on would pin to this screen's size the one intent
that deliberately has none. Every mutating verb already writes back
from the drawn rect, so the sentinel survives nothing it should not.
|
|
A pane image is re-emitted to the client as decoded RGBA, which for one
screenshot is ~4MB of base64 in a single indivisible message -- and
conn_send silently dropped anything that would overflow the 4MB outbox,
while push_graphics ignored the refusal and the model advanced anyway.
Everything stateful in that dropped frame was lost for good: the image
was marked sent (so its placements drew nothing, forever), and the
deletions rode out with it (so a placement the client had was never
revoked). Reported as screenshots in a recovered pi transcript parked
at a fixed spot on the screen, immune to scrolling and pane swaps --
those were placements whose deletion left with a dropped frame. The
1000x800 one could never be delivered at all: its transmission alone
was bigger than the cap.
So the frame is now transactional. Deletions queue and are repeated
every flush until gfx_commit() confirms the client heard them;
transmissions recorded in a frame are unwound by an undelivered commit
so they are sent again. A flush nobody commits counts as undelivered at
the next gfx_begin(), which is what the `graphics` control command is:
bytes rendered for a CLI, not the client. The server reports the
verdict and keeps the paint pending until it lands.
The outbox cap now bounds the backlog rather than backlog-plus-message,
so one big transmission on a drained queue always fits -- and that
surfaced a second way to lose: a queued image legitimately holds the
backlog over the cap until the client swallows it, and push_frame read
that as a dead client and closed the connection. The session
auto-detached every time a screenshot crossed the wire. A full outbox
is a client that is *behind*, not gone: push_frame now marks the screen
for a full repaint (the dropped diff means the client's cells are
unknown) and retries as POLLOUT drains the queue; a client that truly
left still shows up as a write error there.
The headless suite pins the owed-until-committed contract; the new
tests/live_graphics.py drives the real socket with a
bigger-than-the-cap image and a reader that stalls, and would have
caught both bugs: the flood must not detach, the transmission must
arrive whole, and a placement scrolled away must be deleted -- then
come back on scroll without a second copy of the pixels.
|
|
watch_config() rebuilds the set of watched directories after every
reload, because a config that just changed may include different files
than it did a second ago. The inotify and kqueue branches unwatch and
re-add. The Windows branch did neither: unwatching was a comment saying
there was nothing to do, and sl_watch_add() appended to an array the
pump thread had already finished reading.
The pump captured `int n = w->ndirs` once, at the top, and armed one
ReadDirectoryChangesW per directory it found. Everything added after
that was stored and never watched. So an include added to a running
session loaded fine and then sat there: editing it reloaded nothing,
while editing the main config still worked, which is the shape of bug
that gets reported as "sometimes".
It was worse than deterministic. `w->ndirs` was written by the server
thread and read by the pump with nothing between them, so whether a
directory made it in depended on which won the start-up race --
_beginthreadex returns before the new thread runs, so the includes
present at startup usually did, and the ones added later never could.
And because nothing was ever closed, each reload re-added the same
directories: handles accumulated until the cap of 8, after which
sl_watch_add() failed quietly and no new directory could be watched at
all for the life of the session.
So the pump owns the handles and the server owns the wish. An
overlapped read can only be cancelled by the thread that issued it,
which settles where the boundary goes: the server writes a directory
list under a lock, bumps a generation and pokes an event; the pump
notices the generation moved, disarms what it had and arms what is
wanted now. Cancelling waits for the cancellation to land before the
slot is reused, because the OVERLAPPED and its buffer live in that
slot. Duplicate directories are dropped on the way in -- a config and
the theme beside it are usually in the same one -- and the compare is
case-insensitive, as the filesystem is.
sl_watch_close() also stopped every watcher in the process rather than
the one it was given, and freed nothing. It now finds its own, joins
the thread, and releases what that leaves behind.
Verified on the Windows box, with a fully backslashed SLOSH_CONFIG:
editing the config reloads; adding an include and then editing it
reloads; repointing the include at an unrelated directory reloads;
editing the file the include used to name does *not*, which is the
rebuild working rather than an append; and after twelve reloads in a
row it is still watching, where the old cap would have given up after
eight.
|
|
Two places split a path by hand, both with strrchr(path, '/'), both
falling back to "." when there was none: include_path() in config.c and
watch_config() in server.c. On Windows "none" is the ordinary case.
`SLOSH_CONFIG=C:\Users\you\slosh\config.kdl` is how a person writes it,
and a path that has been near a native API comes back all backslashes
whatever it started as.
So an include resolved against "." -- `cannot open ./themes/t.kdl` --
and, worse and quietly, the config watcher watched "." too. Not the
config's directory: the current directory of whatever started the
session. Hot reload did not fail, it watched somewhere else and waited.
Saving your config did nothing at all, with no error to go looking for.
The default path escaped this by accident. slosh builds it as
`C:\Users\you/.config/slosh/config.kdl` -- mixed, because a literal
joined to $HOME always is -- and that one embedded '/' is what
strrchr found. It worked for the path slosh writes and not for the one
a user does.
Both callers now use path_dir(), which learns to stop at a backslash
too, and server.c takes path_base() for the other half rather than
keeping its own copy of the arithmetic. Windows-only: a backslash is a
legal character in a POSIX file name, and splitting on it there would
invent a directory out of a file called `a\b`.
path_dir() also keeps the separator when what is left would be a bare
drive, because `C:\config.kdl` lives in `C:\` and `C:` names the
current directory on drive C, which is a different place.
Verified on the Windows box with a fully backslashed SLOSH_CONFIG:
--check passes, the include loads, and editing the config reloads the
running session. Watching a second directory is still unreliable; that
is the watcher's own bug and the next commit.
|
|
`include "C:/themes/t.kdl"` on Windows did not fail to find the file. It
found a different one, or rather looked for it: absoluteness was tested
as `path[0] == '/'`, a drive letter is not that, so the path was read as
relative and joined to the including file's directory. The result was
cannot open C:/Users/you/cfgA/C:/Users/you/cfgB/t.kdl
which at least says what happened. A layout's `cwd="C:/srv/api"` took
the same turn through path_resolve() and said nothing at all -- the pane
started in the wrong directory, which is the failure this file's own
header comment is about.
Windows spells absolute three ways and all three now count: rooted on
the current drive (`\dir`), rooted on a named one (`C:\dir`, `C:/dir`),
and a UNC share. A drive letter has to be followed by a separator to
qualify, because `C:file` is relative to the current directory *of that
drive* -- a real thing, and not one we can resolve for anyone.
Windows-only, deliberately. `C:` is an ordinary file name on POSIX and a
path beginning with one is relative there, so reading it as a drive
would break the platform that works in order to fix the one that does
not.
|
|
Windows has no VEOF. The console driver does not know the character, so
^D arrives at cmd.exe and PowerShell as an ordinary keystroke and they
ignore it -- it even echoes, which is how you find out: the line reads
`echo hello^D`. The one key that closes a pane on every other platform
does nothing at all, and closing a pane becomes a word you have to type.
That you can type `exit` instead is the tell: the intent was never
ambiguous, only unspoken. So slosh speaks it. What goes to the pane is
the word, not a signal and not a close from outside -- the shell runs
its own cleanup, returns its own status, and keep_dead decides what is
left behind, all three of which reaching past it would skip. It is also
what the user would have typed, so there is nothing new to explain.
The knob is what makes it honest, because a wrong guess costs a pane.
It fires only with the shell itself in the foreground (a program that
is running may want the EOF for itself -- `cat` reading to end is the
whole job), only on the primary screen (in an editor or a pager ^D is
half a page down, and the alternate screen is the pane stating that
rather than us matching process names), and only on an empty line
(POSIX sends EOF there; what it does not do is discard what you typed).
The line is an estimate. The shell's editor is on the far side of a
pty, so pane_send_key counts what it forwarded since the last Enter.
Editing it does not model -- a recalled history entry, ^U, completion
-- leaves the count too high, which forgets to exit. That is the
direction to be wrong in; the other one throws away a half-typed
command.
Default true on Windows, false on POSIX, where the line discipline
already does this and does it properly: ^D there is also EOF for `cat`
and every REPL, none of which is ours to intercept. A knob rather than
an #ifdef so the guards are the same code everywhere and can be tested
where the test suite runs -- test_ctrl_d.py turns it on and drives all
four decisions, plus the case that does exit.
Telling trapped from forwarded needs a program that reports what it
received: on POSIX a forwarded ^D also ends an idle shell, and looking
for the word `exit` cannot separate them either, because bash prints
`exit` itself on EOF.
Verified on the Windows box as well, where it is the default: ^D at an
idle cmd.exe prompt takes the pane (and with it the session, having no
panes left), `more` in the foreground keeps its own EOF, and a
half-typed line survives with the line still on screen.
|
|
slosh builds and runs on Windows 10/11, x86-64 and ARM64, as one exe
with no runtime dependency beyond the OS: panes, splits, tabs,
detachable sessions, the config file and its live reload, the control
socket, the mouse.
The port is a compatibility layer rather than a fork. compat/win holds
stub <poll.h>, <sys/socket.h>, <termios.h> and friends that redirect to
include/compat_win.h, which is force-included ahead of every translation
unit -- so the existing sources keep their POSIX includes and their
POSIX call sites, and the platform differences are answered in one
place.
Two decisions carry most of it.
Everything the event loop waits on is a socket. On POSIX a pty master,
an inotify descriptor and a unix socket are all file descriptors, so one
poll() covers them; Windows has no such union, and WSAPoll waits on
SOCKETs and nothing else. Rather than rewrite server.c's loop around
WaitForMultipleObjects -- which caps at 64 objects and cannot express
POLLOUT -- each non-socket source gets a pump thread that copies it into
one end of a loopback socketpair. The loop is then the POSIX one
unchanged, and only what sits behind the descriptor differs.
Socket descriptors are tagged rather than guessed: a wrapped SOCKET
comes back as SL_SOCK_BASE + slot, so read, write, close and poll can
tell a socket from a CRT descriptor with certainty instead of inferring
it from the numeric range. That is why socket() and accept() are shimmed
too -- an untagged descriptor escaping into the loop is the one failure
mode this scheme has, and it is worth closing completely.
Three things bit hard enough to be worth naming.
A pseudoconsole does not stop the child inheriting your stdio.
CreateProcess copies the parent's standard handles into the child when
STARTF_USESTDHANDLES is absent, and those win. With slosh's own stdout
redirected -- a pipe under ssh, a file under --script, a log under the
daemon -- the shell wrote its banner into slosh's output instead of into
the pane, and the pane had a correct title (ConPTY emits that itself)
over an empty body. The three handles are cleared for the duration of
the call, leaving the child nothing to inherit.
A ConPTY never reports EOF. The output pipe belongs to the
pseudoconsole, not to the child, so it stays open after the child exits
and pane_pump()'s read() -- the entire mechanism by which a pane learns
its program is gone -- never returns 0. A pane whose shell had been told
`exit` sat there looking alive forever. So the exit is waited for on the
process handle, and ClosePseudoConsole is what then produces the EOF.
That third thread is also why the per-pty state is reference counted: a
pump blocked on a dead pipe must not be waited for, or pty_close would
hang the session on the one program that misbehaves, so whoever finishes
last frees.
cmd.exe does not use CommandLineToArgvW's quoting. A literal quote is
\" to an ordinary program and a backslash-then-end-of-quoted-section to
cmd, so building `cmd /c notepad "C:\...\config.kdl"` the first way
handed the editor \"C:\...\config.kdl\", backslashes and all, and
notepad answered "Not a valid file name" about a path that existed.
cmd /c now gets the rest of the line verbatim -- the caller composed a
complete command with its own quoting, exactly as it does for sh -c.
Paths are mixed, and only some programs mind. slosh joins with '/' while
the environment supplies '\', so the config landed at
C:\Users\you/.config/slosh/config.kdl. The CRT opens that without
complaint, which is why --check reported it fine, but a GUI file dialog
rejects it outright. Anything handed to another program now goes through
sl_path_native() first; the internal representation is left alone.
The rest is small. Panes run %ComSpec%, sessions live in
%LOCALAPPDATA%\slosh rather than $XDG_RUNTIME_DIR because a session
socket must not follow a roaming profile to another host, there is no
SIGWINCH so a resize is noticed by comparing the console size on a
bounded poll, no fork so the daemon is a fresh detached process that
reconstructs its arguments on the command line, and no process group on
a pty so the foreground program is the youngest descendant of the pane's
shell. C-a e prefers whichever console editor is actually installed --
notepad is a GUI program, so a pane running it draws nothing and looks
broken -- and says so when it has to fall back.
make -f Makefile.windows ARCH=x86_64 (or aarch64) cross-compiles from a
mac or Linux, and builds natively on the box with zig on PATH.
|
|
Text taller than the viewport was not selectable in one gesture: the
drag stopped dead at the pane's border. Two faults, one visible and one
waiting. The visible one: DRAG_SELECT motion subtracted the content
origin in uint16_t, so a pointer above or left of the pane underflowed,
the grid lookup failed, and the selection froze. The waiting one: the
anchor was a pair of *viewport* coordinates re-resolved on every extend,
so the moment anything scrolled mid-drag the anchor would have drifted
with the viewport, re-anchoring to whatever slid under the press.
The anchor is now a tracked grid ref -- the thing lib-vt documents as
being for 'long-lived anchors such as selections' -- so it follows its
cell across scrolling, pruning and reflow, and only the near end of the
selection moves. If a program flooding output prunes the anchored row
out of scrollback entirely, the selection keeps what it had; there is no
cell left to anchor a new answer to.
Motion clamps into the content rect (a pointer past an edge is pointing
at the nearest cell), and holding past the top or bottom edge scrolls on
a clock rather than on motion, because a resting pointer sends no events
and 'keep going' is exactly what resting it there means. app_tick() does
the step, app_next_deadline_ms() wakes the poll loop for it -- the same
mechanism toasts and hover guides already use -- and both front ends
call the tick, so the behaviour exists under the server and the headless
harness alike. Rows per step is how far past the edge the pointer sits:
one row past creeps, further hurries, so the hand already dragging is
also the throttle. The metronome is select_scroll_ms, a knob so the test
can configure the clock down instead of sleeping.
The test drags to one row above the pane and simply holds: the viewport
walks to the top of scrollback (clamping makes the endpoint exact), the
copy starts at a line that was never on screen with the press, and still
ends at the pressed row -- that last check is the tracked anchor's.
|
|
An audit of every claim in docs/config.md against the parser and the
dump found the page complete -- all 47 top-level settings named, no
ghosts, defaults and error strings as printed, 48 theme colours exactly
-- and three statements wrong or imprecise, all fixed here and in the
two comments that repeated one of them:
- compact does not switch off gap_aspect. The divider ignores it, but
padding is still aspect-corrected through it and the float nudge step
still uses it; saying it 'stops applying' was wrong in config.md,
config.h and config/config.kdl alike.
- 'Two of those take more than one value' miscounted: padding,
min_pane/min_split (cols=/rows= properties, whose syntax the page
never showed) and project_roots all do.
- a compact boundary saves four *columns* or three *rows*; 'four' with
no axis was the columns number wearing a general claim.
|
|
Everything the server said went out through msg_send(), which loops on
sendmsg() until the kernel has taken all of it. That is a blocking write
to whoever is attached. A display whose reader has paused -- C-s at the
terminal, a stopped ssh, a client being swapped -- fills its socket
buffer, and then the *server* sits inside that loop: every other pane,
the config watcher, the control socket and every other connection stop
with it. A session that is supposed to outlive its client hangs on one.
macOS is where this showed up, because its socket buffers are a fraction
of Linux's 64K and it wedges within a screenful rather than after a few
hundred. `make test-live` could not get past the forty-connection probe
in test_session.py: the server was blocked writing to a pty nobody was
draining, so `slosh cmd panes` timed out and took the suite with it.
So connections are non-blocking and output is queued, flushed when
poll() says the socket will take more. The queue is capped: past 4MB we
are not buffering a slow reader any more, we are storing a screen nobody
is looking at, and the connection is dropped -- a display that comes
back reattaches and gets a full repaint, which is what happens after any
detach anyway.
Measured, with a pane running `yes` and a display that never reads:
before, the first control command hung forever; after, 40 of 40 are
answered, the session stays listed, and the server's RSS is flat.
|
|
One knob, `compact true`: panes pack flush against 1-cell divider lines
that meet in real junctions, one frame rings the tab, and a pane's title
rides the line above it -- a boundary costs one cell where the default
look spends four. The lines are strokes: each write says which edges of
its cell it needs, unions with what is there, and never clobbers text,
so ├ ┬ ┼ fall out of paint order and the focus ring recolours without
inventing arms. The divider *is* the gap (eff_gap, one answer for the
layout, split_fits and the resize clamp), so gap hits, hover hints and
corner drags work unchanged. Interior edges give up click-to-split --
one cell cannot hold two verbs -- and a pane's status row overlays its
last content row. Floats, zoom and the D6 flatten keep classic frames
through place_hidden: zero new tree cases.
Underneath it, mode or no mode: a border drag now moves the boundary on
that side, converted on the first motion so the press that never moves
keeps its old meaning -- the handle's click still splits, the rim's is
still deliberately nothing (drag.rim). Which is also the fix for gap 0,
where flush panes leave no gap cell to grab and the resize gesture had
silently vanished with it.
|
|
A dumped session quietly re-tiled its floats, which was the one data-loss
gap left in D22. dump-layout now records `floating=true x= y= w= h=` -- the
*wanted* rect, because intent is what survives a session and the next
screen's layout clamps it like any frame -- layout files declare the same
(linted: floating is a bool, x/y/w/h mean nothing without it), and a file
whose every pane floats self-heals through ensure_tiled: the top float
lands, the same rule closing the last tiled pane follows, no lint required.
save-workspace inherits all of it from the shared dumper.
`float` over the API takes x/y/w/h and then *places* rather than toggles --
a placement is a statement about where, not about whether -- floating the
pane first when it is tiled, re-placing when it already floats, with
omitted fields meaning "keep". The intent is stored exactly as asked: the
write-back rule is for hands, whose drags need the clamp to agree with the
eye; a script means what it typed.
And kitty graphics respect the stack. The cell compositor gets occlusion
free from paint order, but placements are sent after the diff and would
paint an image straight over a float covering it. Each pane's placements
are now clipped by the float rects above it -- every float for a tiled
pane, the higher-raised for a float -- cropping what one clean edge can
express (the pane-edge source-rectangle arithmetic aimed at another
clipper, including from the top and left, where the source origin moves and
the sub-cell offset is consumed) and suppressing what it cannot: a strip
across the middle or a corner overlap leaves a shape one placement cannot
say, and the image returns whole when the float moves. No z= needed --
clipping is the z-order, since nothing is ever allowed to overlap.
D22's "still owed" paragraph is paid and rewritten; M10 is fully checked.
|
|
lib-vt tracks a hyperlink per cell and slosh dropped every one of them: the
client's terminal can only offer a link it was sent, so `ls --hyperlink`,
gcc diagnostics and friends silently lost theirs inside the session.
The link rides beside the cell -- cell_t is the shader ABI and a link is
not a colour -- as a plane of ids into an interned URI table, append-only
so an id in prev_link still means the same URI next frame (which is what
lets the diff compare ids instead of strings), and capped so a hostile pane
cannot grow it forever. URIs with control bytes are refused outright: they
are emitted verbatim inside an OSC, and a control byte in one is a way to
write escape sequences into the client's terminal. The diff wraps emitted
runs in OSC 8, closes what it opens before the frame ends, and treats a
link-only change as a change. Chrome painted over a link takes the link
with it, so a frame, modal or float across a URL is never clickable with
the pane's link; links survive scrollback with the text that carries them.
Plain-text URLs stay the outer terminal's own matcher, with the catch
documented in panes.md: under mouse capture Ghostty only offers links with
shift added to the gesture (shift+cmd hover/click), shift being its
capture escape. Wrapped URLs can never be regex-matched through a
multiplexer -- rows repaint as hard lines -- which is exactly what OSC 8
links, passing through whole, are for.
`snapshot format:"bytes"` returns the emitter's own output for a frame (a
second call is the delta): the model and the bytes are two different
things, and this codebase already paid once for asserting on one while the
other was wrong.
|
|
A pane can float: lifted out of the layout, drawn over the tiled panes and
under the modals, freely moved and shaped. The model is a flag and a rect on
the leaf, exactly like `minimized` -- the seat in the tree is kept, layout
skips it, the siblings absorb its share, and un-floating returns it home.
The drawn rect is the wanted one clamped to the tab every frame, never
written back, so a terminal squeeze round-trips exactly; an explicit move or
resize *does* write back, because the hand is the author of the intent.
Floating pops the pane to the centre: the jump is what says it happened.
Painted after everything tiled, so the hit list -- searched backwards --
resolves an overlap to the float you can see: the z-order and the click
order are the same list. It casts a shadow (float_shadow, one number), and
`floating` joins the states table above `unfocused`, so a float is never
dimmed as merely-unfocused.
The hands: the title moves it, any other border resizes it (a bottom corner
on both axes; the grabbed edge pins at walls and floors, never sliding the
window), H J K L move a focused float, the =/+ key grows it about its
centre, `-` shrinks it (its split was a refusal anyway), and hovering shows
which edges a grab would take, from the same derivation the press uses.
Equalize moved to `0` to free the key; `C-a f` floats (finder went to `s`,
where it reads as search); `C-a F` opens a new floating shell -- the
throwaway terminal; the function row f1..f12 became nameable in configs
along the way. Guards throughout: the last tiled pane cannot float or
minimize away, closing it lands the top float, splits are refused with
their own words, floats are not swap targets.
Dropping got a grammar: while a pane is carried, the pane under the pointer
subdivides into a swap centre and four edge bands meaning "insert me beside
this pane, on this side" -- detach_leaf + place_beside, the surgery
cross-tab moves already had. Bands are thirds (the first map of quarters
showed the swap owning half of each axis), the hovered band fills the half
the drop would hand over, the release resolves against the painted frame's
hits so it cannot disagree with the fill, and a drop between siblings along
the axis is a reorder that always fits -- refusing it was refusing the most
ordinary drag there is.
Also: rename-pane/rename-tab as palette actions (renaming was a
double-click and nothing else, the one verb the palette could not reach),
and C-u clears the rename editor like it clears the picker's query.
Design log in PLAN-FLOATING-PANES.md; to fold into DESIGN.md as D22/M10-11.
|
|
The old name argued that the extension should say how to read the
file. The argument lost to a collision: zellij speaks the same KDL
syntax under that same extension with a different schema, and to
anyone arriving from there .kdl said *whose* file it is -- naming the
wrong owner is worse than not naming the parser. So a layout is
x.layout, a project's is slosh.layout, the name says what the file is
and --check says how it was read. (It is still KDL underneath; point
your editor's highlighting at *.layout.)
Nobody's checked-in workspace vanishes on upgrade: the project scan
still finds slosh.layout.kdl, never writes it, and where both
spellings exist the new one wins, because two files disagreeing needs
one answer. The unparseable-file heuristic in --check learned the new
suffix -- it matched '.layout.' with a trailing dot, which the new
name no longer has.
The webdemo overlay moves with it (demo.layout), rebuilt and
boot-checked.
|
|
Every non-space glyph of the logo is a particle that flies to its cell
over the first two fifths of splash_ms, then *is* the logo for the
rest. There is deliberately no particle state anywhere: a position is
a pure function of (target, elapsed, seed), recomputed every frame
like the layout and the pane states and for the same reason -- physics
that is stored can drift, physics that is derived cannot. The seed is
the splash timestamp, so every attach flies differently and one splash
is coherent frame to frame; a per-particle hash staggers departures,
and the delay rescaling lands every straggler exactly at the end,
because an almost-assembled logo reads as a bug.
Five motions, picked by the clock and pinned by {"cmd":"splash",
"motion":N}: rain (fall in, column by column), scatter (converge from
everywhere), slide (rows enter from alternating sides), vortex (spiral
in around the centre, orbit shrinking and unwinding), shuffle (the
glyphs trade places, then sort themselves). The colour chains keep
playing over the box throughout, so particles catch their hue as they
cross into it.
expr_sin() is the expression language's own sine exported for C-side
animation: a motion sketched in the repl and one compiled into the
binary cannot disagree about what sin(90) is.
The tests learned two things the flights taught: the logo's opening
run appears three times per line and assembly is staggered, so a
middle occurrence can land first -- "assembled" is the row whose first
block glyph is where the match begins -- and windows need margins that
survive a parallel suite on a busy box.
|
