This patch unifies the Gamma node across editors:
- Compositor now uses the Shader Gamma node implementation.
- Geometry Nodes also uses the node as a new addition.
- UI-wise in the Compositor, only socket labels change to "Color".
Forward compatibility is broken for the compositor, and the python API
changed, since the node's ID name is now ShaderNodeGamma as opposed to
CompositorNodeGamma.
Pull Request: https://projects.blender.org/blender/blender/pulls/142414
Follow up to fix for #132460, slow startup from file file-system access
has been removed as it causes problems, then added back since it seems
like a fix if we assume all file-systems are local & online.
Note why this causes problems to prevent the checks being re-introduced.
Changes the overload from accepting `const AssetWeakReference *` to
`const AssetWeakReference &` to more strongly indicate proper usage. We
should not have a usecase for setting a null `AssetWeakReference`
Pull Request: https://projects.blender.org/blender/blender/pulls/144821
Prior to this commit, determining the default brush asset used for a
particular mode was based on the object's mode. This is slightly
incorrect for the Image Editor, since it may be in Paint Mode no matter
the underlying object type. To fix this, use the runtime `PaintMode`
enum for determining these default values instead of `eObjectType` and
store the `PaintMode` on `PaintRuntime` for easy access.
Additionally, inside the toolsystem, prevent accidentally unsetting
the default brush by checking for the presence of the asset when
loading an `AssetWeakReference`.
Pull Request: https://projects.blender.org/blender/blender/pulls/144765
This allows to control the type of data inside the create infos for
easy variation of the same shader.
Also this wraps the last global shader resource usage for the future
SRD.
Pull Request: https://projects.blender.org/blender/blender/pulls/144769
Two formatted reports introduced in e239c7f43c used a type defined at
compile time using the PRId64 macro. Translation failed for these
messages, because the message was extracted as:
"No keyframes removed from % strip(s)"
and the type was lost at the time formatting happened.
This commit instead uses fmt to format the message, which deals with
using the proper int type depending on the platform.
Pull Request: https://projects.blender.org/blender/blender/pulls/144374
Moving the python call itself into a dedicated sub-scope allows to
ensure local variables (arguments) do not risk leaking into the
rest of the code after being released.
Also use helper macro `PyTuple_SET_ITEMS`.
Pull Request: https://projects.blender.org/blender/blender/pulls/144798
This patch lowers the priority of menu sockets from the maximum priority
to the lowest priority. That's because menu sockets are nearly never the
main socket.
The current state is blocking the adoption of menu sockets in the
compositor, since most of the compositor nodes have color and float main
sockets, which currently have lower priority than menu sockets, making
node link insertion fail.
Pull Request: https://projects.blender.org/blender/blender/pulls/144779
So far only active color spaces were queries from the OpenColorIO, which
made sense from the list of spaces displayed in the menus. However, some
logic requires knowing more details about color spaces used by displays,
for example to check if it is data, or whether color space matches the
display one.
This change makes it so our OpenColorIO integration handles inactive color
spaces by storing them in a separate array. Config::get_color_space will
now work for both active and inactive color spaces.
Pull Request: https://projects.blender.org/blender/blender/pulls/144780
`asset_shelf::regiondata_duplicate()` first creates a shallow copy of
the `AssetShelf`, including its `AssetShelfSettings` member. So the
contained pointer point to the same memory.
While this is a rather unusual case for a copy assignment operator to
consider, I think this is fine since the API allows these shadow copies.
This is a bit of a consequence of mixing C and C++ style memory
management.
Pull Request: https://projects.blender.org/blender/blender/pulls/144613
Passing negative precision to internal units functions made the code
difficult to follow, especially when the precision is being
adjusted/clamped.
While the public API remains unchanged, detect the negative precision
early and assign it to a boolean which is passed as an argument.
Also improve doc-strings.
For Todo: #138764
This follows a fix that tried to address this issue, but got negative feedback #139398
At certain zoom levels the animation editors in 4.5 would draw lines on subframes.
The previous patch fixed that by drawing less lines, but that got pushback from artists.
This patch restores the previous behavior, but ensures that major lines are *always* drawn
on even frame numbers, thus allowing minor lines to be drawn between
them on full frames.
Pull Request: https://projects.blender.org/blender/blender/pulls/142858
Use a thread local buffer to gather info about the custom normal space
for each corners group in threading to concatenate them later. This
avoids need to lock a mutex to write into a buffer shared betwen threads
in the corner groups traversal hot loop.
The performance improvement depends on the mesh size and the number of
CPU threads. In some files there might be no change, in other files an
improvement of over 2x was observed, mostly because we can now remove
the compromise from c8a4026984.
Pull Request: https://projects.blender.org/blender/blender/pulls/144660
Before, trying to add multiple strips would always create an image
sequence. This patch allows images to be added in bulk, just like movie
& sound strips.
A new "Image Sequence Detection" property has been added which works
like the image editor. Toggling it off will always import individual
images and never create image sequences.
If it's on, you can also optionally "Use Placeholders" to fill in gaps,
which exhibits the same behavior as before (placeholders will show as
pink in the preview until you have the files on your system in the right
folders).
You can even add both image sequences and regular images at the same
time. Only caveat is that this doesn't work if the regular images are
interspersed within sequence frames in the file browser view when you
add them. I.e., all frames of a sequence should be listed without any
regular frames in between (but the sequence frame numbers don't have to
be in order). This can usually be ensured on all sort types (like date,
name).
NOTE: This patch also replaces the "end_frame" property on image /
effect strips with "length," to support behavior when multiple strips
are involved. This is a breaking change for the API, but is a fairly
small change and will be included in the release notes.
The "length" property allows one to adjust all image strips' lengths at
once. It does nothing for image sequences, however, since their lengths
are defined by their image count.
Pull Request: https://projects.blender.org/blender/blender/pulls/143974
- Deduplicate `scale_fit_methods` enum
- Reorder functions, polls, and code to locations that make sense
- Rename some functions to remove `seq` terminology in favor of `strip`
- Renames `len` in `LoadData` to `count` for images
- Add clarifying comments and improve descriptions
With the aim of removing `seqbasep` to remove the complicated logic for
repairing pointers within the `Strip` struct when loading files and undo
steps, this commit just moves access of the variable behind a function.
In the future the function will retrieve the list from a Strip pointer,
for now it just returns the existing pointer.
Overall motivation is that blend file pointer manipulation is incompatible
with the changes required for #127706.
Pull Request: https://projects.blender.org/blender/blender/pulls/144624
This makes the new parser code output correct errors.
Also simplify error logging.
Simplify a few things inside the parser interface.
Add debug string to the `Token` and `Scope`.
Pull Request: https://projects.blender.org/blender/blender/pulls/144708
VK_KHR_maintenance4 is enabled when available. VMA isn't aware that we
enabled it and could still use less optimized code-paths. This PR will
inform VMA that it can use the optimized code-paths.
The improvement is that we can decide in which memory area a specific
resource will be allocated, without the need of allocating the resource
header (VkImage/VkBuffer).
Pull Request: https://projects.blender.org/blender/blender/pulls/144552
Dragging a non-Blend file onto the Blender executable currently
silently fails with Blender just aborting. With this PR any files used
as file argument will load Blender and show a warning if the file does
not exist or is not a blend file. Blend files that cannot be read are
treated as they are now.
Pull Request: https://projects.blender.org/blender/blender/pulls/139128
We have a set of corners around some vertex. They have arbitrary
connectivity between each other and form a set of linked lists through
shared smooth edges. We have to linearly traverse each of these lists to
mix the normals. And as a way to iterate over all these lists mixed in
some index space we use a mask of already visited corners. So we take
the first non-checked corner by mask lookup. And this lookup starts from
the beginning every time even if we know there is no way to find any
elements before current one. In a synthetic example with 20k fan fill
of circle this gives a 8x speed up.
Pull Request: https://projects.blender.org/blender/blender/pulls/144649
The semantics of checking "has_value()" (etc.) are much better than
checking for an empty span when dealing with the result of an attribute
lookup. This mainly affects the Bezier curve handle position attributes
currently. Plenty of places assume those attributes exist now. In a
couple places the code is a bit safer now, otherwise it's just a bit
more obvious.
Pull Request: https://projects.blender.org/blender/blender/pulls/144506
There are many accumulated unused properties in theme. Generally, it's a mess,
lot of copy-paste errors and leftovers between editors, shared functions with no
actual use, etc. Some were just added and never used, like Pattern and Layout nodes,
which were meant to be used with Texture Nodes I believe, but it never happened.
Overall 19 properties removed:
- 3D Viewport: Edge UV Face Select
- 3D Viewport: Last Selected Point
- Dope Sheet/Timeline: Value Sliders
- Dope Sheet/Timeline: View Sliders
- Nonlinear Animation: View Sliders
- Video Sequencer: Window Sliders
- Graph Editor/Drivers: Window Sliders
- Graph Editor/Drivers: Vertex Bevel
- Graph Editor/Drivers: Vertex Group Unreferenced
- Image/UV Editor: Vertex Bevel
- Image/UV Editor: Vertex Group Unreferenced
- Image/UV Editor: Freestyle Face Mask
- Image/UV Editor: Face Retopology
- Image/UV Editor: Face Orientation Front
- Image/UV Editor: Face Orientation Back
- Node Editors: Selected Text
- Node Editors: Pattern Node
- Node Editors: Layout Node
- Video Sequencer: Draw Action
---
Details:
- Made Vertex Bevel and Vertex Group Unreferenced colors in `rna_def_userdef_theme_spaces_vertex`
optional, so that they don't appear in Graph Editor and UV Editor theme properties, where they're unused.
- Same thing in `rna_def_userdef_theme_spaces_face` for face attributes (Freestyle Mark, Retopology),
so that they don't appear in UV Editor, where they're unused. Also did same for face orientation colors,
but used different argument for them, since they could be used in future (remember seeing PR for this).
Pull Request: https://projects.blender.org/blender/blender/pulls/143782
This PR adds a "Mirror" menu to the Strip menu within the VSE preview.
This menu is similar to the "Mirror" menu in the view3d "Object" menu.
It allows the user to access the mirror operator through a menu, rather
than via a shortcut.
Pull Request: https://projects.blender.org/blender/blender/pulls/142506
This PR just updates the error message shown when the selected backend
is Vulcan but that fails and uses OpenGL instead. This currently uses a
bespoke popup alert block, but this PR replaces that with new UI_alert
function with the same wording.
Pull Request: https://projects.blender.org/blender/blender/pulls/144676
Its generally preferable to use `edge_other_vert` if you have a set of
2 ints and know some of them and want to select other. This is
preferable since it will make a check that assumption is met and at
least one of the ints is the same as the known one. This function itself
is also made to work with topological indices so has invariant for
positive ints and can work without branches and even test instructions.
Pull Request: https://projects.blender.org/blender/blender/pulls/144631
Reported on chat, confirmed with the python docs which lists
cursor_position as "float array of 3 items in [1.17549e-38, inf]"
Root cause is a fun fact about floats, FLT_MIN is the smallest positive
value possible, not the largest negative value (like INT_MIN)
Pull Request: https://projects.blender.org/blender/blender/pulls/144657
Tree view elements can be deleted by the remove operator located on the
right side with `-` icon. But these operators are different for every
type. Now introduced `UI_OT_view_item_delete`. With that it is possible
to delete elements with X key when mouse is over tree view. In future
maybe we could remove type specific operators (for example: `GREASE_PENCIL_OT_layer_remove`).
See video in PR description
Pull Request: https://projects.blender.org/blender/blender/pulls/144045
