Kerning read one pair of parentheses around its text only while it
showed an automatic value as (12), so typing (10+5) over a manual value
committed nothing. spin_kerning now reads that pair in both states;
the field still shows parentheses only for an automatic value, and
(10+5)*2 is still refused.
The opening tests waited 400 or 500 ms of wall-clock time for files read off
the UI thread, and up to 30 s in loops that polled until the queue emptied.
Under load a read could take longer, and
a_file_that_opens_and_asks_holds_back_the_next_one_ready failed.
wait_read now blocks on each pending file's own result channel and keeps the
result for poll, and open_ready waits and then polls once. The background save
in long_operation_busy_while_a_file_opens_or_a_background_save_runs waits with
background::wait_all, and its job waits on the release channel without a
timeout. Shipped code is unchanged.
On macOS the menu bar is native and is read again only when its state
hash changes. Type > Font now lists the loaded fonts until the installed
ones are scanned, so the hash takes the font generation and whether the
scan has ended. The tests that wait for process-wide generations to
hold still wait for the fonts too.
The menu bar builds Type > Font in every frame, the first included, and
its family list waited for the scan of every installed font. When that
scan took longer than the 20 s first-frame watchdog (thousands of fonts
on a slow disk, or each file checked by antivirus), the app took the
graphics adapter for hung and started again without it, until none was
left and no window ever showed.
The menu now lists the fonts already loaded until the scan ends, then
all of them (FontDb::menu_family_list_now, installed_fonts_ready).
text.tabs.set and text.tabs.clear return without an edit when every
targeted paragraph already holds the stops. A Ctrl-drag that the 0 pt
bound keeps from moving, or a drag released where it started, no longer
adds an undo step that changes nothing. When one targeted paragraph
holds other stops, the command still sets them all.
appearance.addFill copies the active fill row only when the call has no
`ids`, no `id` and no `target: "contents"`; the params text now names
`id` too. appearance.addStroke gives the new stroke at least 1 pt, so a
0.5 pt top stroke gives a 1 pt stroke; the params text now says so.
Tests pin both: `id` alone leaves the active row alone and puts the new
fill on top, and Add New Stroke on a 0.5 pt stroke adds a 1 pt one.
OK in Shape Builder Tool Options sent highlightColor to tool.setOption,
which keeps only #rrggbb, so #ff000080 was applied as #ff0000 with no
error. OK now gives the same error as the Live Paint tool options for a
hex color with an alpha part (#rrggbbaa or #rgba), the dialog stays open
and the tool's options do not change. The tool.options params text,
docs/mcp.md and the progress doc say so.
A command bound to plain Escape no longer runs when that Escape
cancels a floating Tools panel move or a ruler-origin drag, nor on the
key repeats while Escape stays held. The next separate Escape press
runs it as before. A test binds Select None to Escape and holds Escape
through both cancels.
File > Package stored its status line in the interface language, so after
a language switch the status bar kept the old language, and the control
channel and MCP read a translated message. The status line and its notes
are now English, and only the "Package created" dialog is translated when
it is shown. Each catalog gets message rows for the notes, so the status
bar translates them too. On the desktop, where the dialog names every font
left out, a long font list is cut short in the status line so the line
stays short enough to translate.
A number typed into the Kerning field after Optical keeps the Optical
flag on the type, and the manual value sets the spacing. The Kerning
menu still checked Optical. It now checks neither Auto nor Optical
while a manual value is set.
A freeform gradient whose field has no points of its own takes its
colors from the stops, so spot-linked stops in it are written in
process colors too. The export now warns in that case, as it does for
the other gradients drawn as images.
VECTORCRAFT_RENDER_THREADS can make default_threads() return more than 64,
the largest value prefs.set accepts for renderThreads. The test now caps
the value at 64, so it passes with the variable set to 100.
A Stop or a document switch sent through the control channel in the
frame a step comes due now takes effect before that step runs. An
action played through the control channel at Pause For starts its first
pause in the next frame, as the panel's Play does.
Step by Step with the only document closed, and Pause For with the playing
document closed while another stays open: when the next step is due the
playback ends with its status and no further step runs.
A Step by Step playback now waits after each step until a UI pass that the window shows has drawn it. A logic-only iteration, a pass in a minimized or covered window and a discarded pass do not count, so a hidden window no longer runs the steps unseen and does not ask for frames while it waits.
Step by Step and Pause For remember the document that was active after each step. When the next step is due and a different document is active (another one was made active, or that one was closed), the playback ends with a status and the rest of the steps do not run. A step that makes another document active itself does not end the playback.
Step by Step holds the frame after Play, so the first step's result is drawn before the
second step runs, also when the panel's Play button runs after the canvas drew.
An action step recorded as command.batch (an action played while recording, including one
playing at Step by Step or Pause For, or a panel operation that runs as one batch) now plays
as its own steps in Play and Batch..., instead of failing with "batches cannot nest".
The Actions panel menu's Playback Options... sets how actions play (ui.action.playbackOptions {speed?, seconds?}). Accelerated, the default, plays every step at once as one undo step, as before. Step by Step plays one step per frame so the canvas draws each step's result, and Pause For waits 1 to 60 seconds between steps; each paced step is its own undo step. Stop Playing/Recording ends a paced playback, and another action does not play while one is paced. The setting is kept with the UI preferences. Batch... plays at full speed.
Batch Trace Folder replaces a script, so it is now a Document Tools
add-on command, addon.document.batchTraceFolder, under Object ›
Plug-ins › Document Tools (as Export All Open Documents is), and no
longer in Object › Image Trace. Behaviour, caps and the refusal of a
folder too big to list in full are unchanged; the tests move with it.
The status messages have @msgplural rows in every catalog, as does the
menu label. When some images trace and some don't, the status is one
counted message ("Traced {n} images; not traced: {failed}") so it
translates as a whole.
list_dir cuts its listing before sorting, so a cut listing could skip any image. The batch now
stops with an error instead of tracing part of the folder.
imageTrace.batchFolder opens each image in a folder, traces it with a
preset (optionally expanded), saves it into a destination folder under
its own name (vectorcraft, svg, pdf, ai or eps; taken names numbered so
nothing is replaced) and closes it again whether it traced or not; the
active document stays active. Object > Image Trace > Batch Trace Folder
asks for both folders. Not on the web.
A Copy of ruler guides alone left the system clipboard as it was, so the
next Paste took whatever another app had copied before instead of the
guides. The guides-only Copy now empties the system clipboard (on the
web, empty text through egui with Copy As SVG on), as any Copy replaces
what the clipboard held, and drops an earlier Copy's formats not written
yet. What another app copies afterwards still pastes.
Copying only ruler guides published an empty SVG and text to the system clipboard, replacing
what other apps put there. The clipboard now offers other apps nothing without art. Tests cover
two pastes adding guides of their own on the current layer (following its visibility) and the
SVG carrying no guides.
With anchors and ruler guides selected, Cut deletes the anchors and the
selected guides in one undo step (edit.clear's guides: true reaches the
anchor-deleting path).
Cut with objects and ruler guides selected now removes both, in one
undo step: edit.clear takes guides: true beside ids to delete the
selected guides with the objects it names.
Edit > Copy and Cut put the selected ruler guides on the clipboard,
with the selected objects or on their own. Pasting puts them on the
layer new art goes to, selected, in the same undo step: with objects
they move as the objects do; alone they keep their places, while Paste
in Place / in Front / in Back put them on the active artboard as they
lay on theirs, and Paste on All Artboards on every artboard.
text.areaOptions given for several area type frames (ids, or a selection of
several) copied the first frame's options onto all of them, so setting one
option (say the vertical alignment) also replaced the others' columns, gutter,
inset and fit. Each frame now takes only the options given over its own, and
fit resolves against each frame's own setting, so fitMinPercent alone changes
only frames that already shrink text to fit. Options that change nothing add
no undo step and keep redo.
A PDF-compatible .ai (or a PDF saved with the editing data) now opens
from its layers without its PDF part being imported, and without being
compared with its page, when:
- its text document makes the type of every text object, hidden ones
included, at least one text object shows, and none is drawn by the
appearance of the layer or group it is in;
- its layers hold no unread object, no placed file with a second
placement and no live envelope with drawn art to fall back on, and
no image too large to read;
- it declares no large canvas, and its artboards are its pages' trim
boxes one for one, in points or in the first page's UserUnit.
A file whose layers draw differently from its page then opens from its
layers instead of its PDF part. The bleed comes from the pages' bleed
boxes. The check runs only when no pages are picked, layers and the
editing data are on, the crop box is asked for and type comes in as
text. Files saved without PDF compatibility, files whose PDF part
draws only type or nothing, and files whose PDF part's content can't
be read take the same path when they qualify: their bleed comes from
the pages' bleed boxes, pages in units larger than a point draw them
at their size in points, and the note that the PDF part couldn't be
read is not given, since its content is not read.
Files that don't qualify open as before, and the editing data decoded
and read for the check is reused after the PDF part is imported
(layered_ai_from).
When a file whose PDF part can't be read is scaled up to its size in
points, its text wraps are now laid out again at that size, and the
default graphic styles of a file without graphic styles of its own are
no longer scaled. The same applies on the new path.
#5139 (in 0.21.0) made a positive Twist angle turn clockwise, but
documents saved by earlier releases stored their angles with the other
sign and reopened twisted the other way.
The document version (FORMAT_VERSION) goes from 1 to 2. Opening a
version 1 document negates its Twist angles (in layers, opacity-mask
art, Shaper edits, symbols, graph designs, pattern swatches, graphic
styles and brush definitions), unless the file is of the current file
version and its generator is VectorCraft 0.21.0 or later. The document
opens and saves as version 2. A save for an older file version, which
only apps before 0.21.0 read, writes a version 1 document with the
angles negated back. Graphic style libraries (.vcstyles) go to version
2: a version 1 library's Twist angles are negated when it is read.
Readers do not check either version, so older builds open the new files
as before.
Add New Fill again copies only the active fill row's paint; Duplicate
Item copies a whole row. The params text of appearance.addFill and
appearance.addStroke now says the new fill or stroke takes the paint
(and for a stroke, the weight) of the row it starts from, and a test
pins the paint-only copy.
With a fill row active, Add New Fill copied only the row's paint and
reset its opacity, blend mode, visibility, overprint and effects. The
new row is now a copy of the whole active row, as the command's
description says.
The `method` text of `paint.editGradient` and the Method paragraph in
docs/mcp.md said SVG and PDF exports draw a Perceptual gradient as an
image. PDF writes an eligible one natively: tints of one spot ink as a
Separation shading in that ink, and spot inks mixed with other spot,
CMYK or gray inks as a DeviceN shading. Gradients without spot stops,
with RGB or Lab stops, dithered fills and curves with too many stops
still become images there. Both texts now say so.
text.setStyle and text.setRangeStyle collected the fonts for their
warning after Missing Glyph Protection had run. A character the new font
has no glyph for keeps its earlier font, and when that font was not
available the warning named it, although the command did not set it.
protect_missing_glyphs now returns the byte ranges that kept their font,
and both commands leave those ranges out of the warning's fonts.
Reduce to Basic Appearance and Add New Fill matched the Appearance
panel's active row only when every selected object had an equal row.
With three objects whose third had a different lower fill, the second
object lost the match too: Reduce kept its top fill instead of the
active one, and Add New Fill put the copy on top instead of above the
active row.
items_per_target now matches each object against the first one
separately, so an object keeps the match whenever it has an equal row.
eframe's web build sends each key's text before its press. When several
keys arrive in one frame, the editor paired the backtick press with the
text typed by the key after it, so a shortcut on that key could show the
next key's character. The editor now pairs each text with its press
through `shortcuts::typing_presses`, which shortcut dispatch already
uses, for both event orders.
Numeric fields dropped every parenthesis before reading a value, so
(10+5)*2 typed into a field committed 20. Only the Kerning field, which
shows its automatic value as (12), now drops one pair of parentheses
around the whole text. Parentheses anywhere else are refused, and fields
without that display refuse them too.
The raster-effect cache key now includes the setting, so a renderer that
draws the same art with it on and then off no longer reuses a drop
shadow or other raster effect made with the other setting.
The Navigator panel menu's Panel Options… is enabled and opens a dialog with
the view box color (a preset or a custom color; Light Red, today's color, is
the default) and Draw Dashed Lines as Solid Lines, off by default, which draws
dashed strokes without their dashes in the panel's thumbnail only. The
options are kept with the UI state, and ui.navigatorPanelOptions sets them.
RenderOptions::solid_dashes draws dashed strokes as solid lines, including
type strokes and the silhouettes of stroke effects.
layer.setProps, layer.new, layer.newSublayer and livePaint.gapOptions read
an 8-digit hex color such as #ff000080 as #ff0000 and set it without its
alpha. They now refuse it with the error a bad color already gets, and
change nothing. #rgba was already refused. #rrggbb, #rgb, a preset index
and a preset name are read as before.
`Color::from_hex` reads `#rrggbbaa` and keeps only its first six digits, so
a `color` set to `#ff000080` through `ui.dialog.set` was applied as
`#ff0000` without an error by Tile Edge Color, Live Paint Selection Tool
Options, Live Paint Bucket Options and Gap Options.
- A new `checked` next to `resolve` in `dialogs/tile_edge_color.rs` gives
the existing "is not a colour" error for an 8-digit hex color, and
otherwise reads the color as `resolve` does. `resolve` is unchanged, so
the swatches the dialogs draw with it show such a color by its RGB as
before.
- Each of the four dialogs' OK checks the color with `checked` first, and
leaves the dialog open with nothing changed. Tile Edge Color and Gap
Options already refused `#rgba`, which `from_hex` does not read. Live
Paint Selection and Live Paint Bucket used to ignore a color they could
not read, such as `#f008`, and apply their other fields; their OK now
refuses it.
- With Preview on, Gap Options sends no color for a `#rrggbbaa` value, so
the groups keep their own color, and the previewed gaps draw in black.
Docs: the params text, docs/mcp.md, docs/control-protocol.md and
docs/target-app-parity.md.
The params text of select.same.symbolInstance, docs/mcp.md and
docs/target-app-parity.md now say that the object compared with can be
of any kind (a group member selected on its own, or an object that is
not a symbol instance, which leaves the selection empty), and that new
art on a locked artboard is not passed over while it stays selected.
A test covers the three cases.
dotCooCoo