aboutsummaryrefslogtreecommitdiffstats
path: root/tools
Commit message (Collapse)AuthorAgeFilesLines
* Include what you need: qquaternion.hMarc Mutz2025-07-101-0/+2
| | | | | | | | | | | | | | | | qmatrix4x4.h will lose its qquaternion.h include, so include qquaternion.h explicitly in all files that mention 'QQuaternion', unless, for a foo.cpp, the own foo.h has already included it. Picking all the way back, even though the include removal won't be picked as far back, because it's the correct thing to do and cannot fail. Also add forward declarations, where they suffice. Pick-to: 6.10 6.9 6.8 6.5 Change-Id: Ibfc1e941978c750c05e57e3772f3a68bc921c81f Reviewed-by: Volker Hilsheimer <[email protected]>
* Add lightmapviewer toolJonas Karlsson2025-07-0211-0/+760
| | | | | | | | | | | | | This is a debugging tool for viewing Qt's baked lightmap files. While users can use this tool, it is intended for development only. The tool is both a GUI tool for viewing lightmaps and a command line application that accepts flags for printing info and extracting the content of the lightmap file. Pick-to: 6.10 Change-Id: I95279e248ea82e0e1384bbf880dcecc7af778dfe Reviewed-by: Laszlo Agocs <[email protected]>
* Improve node collection and frame prepChristian Strømme2025-05-301-1/+18
| | | | | | | | | | | | | | | | | | | | | | | | | | Improves the collection of nodes and rendering data during the preparation phase by: - Only traversing and collecting nodes when the node tree changes. - Assigning a unique ID to each node in the tree for fast access to its associated rendering data. - Having a single root node for the entire window. If there are multiple views within a window, they all operate on their respective sections of the tree. - Storing rendering data separately which avoids collecting and copying in addition to make it more feasibly to just uploading the data in bulk when needed. - Proper separation of data allows for the calculation of non-dependent data in parallel, once the global state has been updated. This is merely the initial step towards a more advanced storage solution. However, some adjustments are still necessary to transition a significant portion of the preparation phase to the GPU. Our somewhat contrived benchmark demonstrates a 6x improvement, which is also applicable to the worst case scenario when nodes are dynamically added. Change-Id: I1dd12734f6a043ca8145cb1c667921bb664aa3db Reviewed-by: Christian Strømme <[email protected]>
* XR: Fix issue with effects and multiviewChristian Strømme2025-01-231-2/+1
| | | | | | | | | | | | | | This amends 64821f3552aa7f3826de945141f1e3c16f175bc1 as the logic for adjusting the view count in sub-views was done after the effects was processed causing the queried view count to be out of sync. This change also touches the general state tracking the layer node, to make it more consistent and ensure the dependencies on the layer state is processed in order. Fixes: QTBUG-132811 Pick-to: 6.9 6.8 Change-Id: Idc4c78184f4af3834505c5d6c130e8212ee5837d Reviewed-by: Andy Nichols <[email protected]>
* Add missing qDebug includeVolker Hilsheimer2025-01-171-0/+1
| | | | | | | | | The code evidently relied on a transitive include of the QDebug header. Include it explicitly to follow up on cleaning up changes elsewhere. Pick-to: 6.9 Change-Id: Ifb30262309d599e08557aabd803b1ec378c82f36 Reviewed-by: Christian Strømme <[email protected]>
* Replace deprecated QImage::mirror(ed) with flip(ped)Volker Hilsheimer2024-12-021-1/+1
| | | | | | | | Follow up on change in qtbase:577946c1f05aaaa2a3f9682001aeb4144386b26b. Task-number: QTBUG-129575 Change-Id: Iaec6daeca8299c2f3a2048635112cbdf864b8d56 Reviewed-by: Allan Sandfeld Jensen <[email protected]>
* materialeditor: support instancingInho Lee2024-08-015-2/+157
| | | | | | | | | | | | | | | | | | | When using instanced model's customData, it can be used only with CustomMaterial. It will not fully support instancing's all the feature but customData and color of InstanceEntryList are enough to check the shading effect for current material. They will be mapped to INSTANCE_DATA and INSTANCE_COLOR in vertex shader. When using instacing, Preview will show 5 instances and the center object in position (0, 0, 0) is affected by modified properties.(color and customData) This object's index(INSTANCE_INDEX) is 0 and it can be used in vertex shader, too. Change-Id: Ied2fbf456366d52a41cb6ef2d1dbab266d38112a Reviewed-by: Andy Nichols <[email protected]>
* Compile materialeditor (again)Kai Köhne2024-07-181-1/+1
| | | | | | | | | | | | | There is no feature datastream, so this disabled compilation of the tool on all configurations. Fixes a regression introduced in 4f042b2ebb0bc. Pick-to: 6.8 Fixes: QTBUG-127326 Change-Id: Id330449def1c2f182cc1d357c3a1b507066bc22e Reviewed-by: Jari Helaakoski <[email protected]> Reviewed-by: Alexandru Croitor <[email protected]>
* Spotlight: Use single depth map instead of cubemapJonas Karlsson2024-06-251-8/+8
| | | | | | | | | | | Currently spotlight shadows requires rendering six times. This changes it to only require one depth map. Piggybacks the directional lights texture array usage, so they both go through that path now. Task-number: QTBUG-116885 Pick-to: 6.8 Change-Id: I4a23c8b22a1918d1e0ce2cfe5ac594bb527975c7 Reviewed-by: Jonas Karlsson <[email protected]>
* Correct tools licenseLucie Gérard2024-06-1949-49/+49
| | | | | | | | | | | | According to QUIP-18 [1], all tools should be LicenseRef-Qt-Commercial OR GPL-3.0-only WITH Qt-GPL-exception-1.0 [1]: https://contribute.qt-project.org/quips/18 Pick-to: 6.7 6.8 Task-number: QTBUG-121787 Change-Id: I0883ded9f64fcf8107c397a837149133911f2951 Reviewed-by: Kai Köhne <[email protected]>
* Fix compilation issues when features are disabledJari Helaakoski2024-05-271-2/+6
| | | | | | | | | Commit fixes mimetype, quick_shadereffect, cborstreamwriter, datastream and localserver features. Task-number: QTBUG-122999 Change-Id: If4850f37b2047edb59de198741111e731c0f0669 Reviewed-by: Alexandru Croitor <[email protected]>
* Start enabling postproc effects in multiview modeLaszlo Agocs2024-05-071-1/+2
| | | | | | | | | Won't work still, but for reasons different than before. Plus some qmlxr test assets. Change-Id: Ia7a81b0310d4006b14354e5d897493a73b142537 Reviewed-by: Andy Nichols <[email protected]>
* MaterialEditor: silence -Wunused-private-fieldTim Blechmann2024-04-241-1/+0
| | | | | | | | `m_ready` is apparently not used, causing a clang warning to be emitted Pick-to: 6.7 Change-Id: Ibb56e91ebc7a99c951a12fe52c12b84d1c8f0c30 Reviewed-by: Laszlo Agocs <[email protected]>
* Do not wrap pointers in "const auto &"Ulf Hermann2024-03-211-4/+2
| | | | | | | That's somewhat misleading and makes gcc 12 produce bogus warnings. Change-Id: Ic563bba09a8baec34f07dacf774463399f902d15 Reviewed-by: Giuseppe D'Angelo <[email protected]>
* Enable custom material keywords for multiviewLaszlo Agocs2024-02-191-1/+1
| | | | | | | | | | | | | | | | | | | | Make the following functional in custom materials: MODELVIEWPROJECTION_MATRIX VIEWPROJECTION_MATRIX VIEW_MATRIX PROJECTION_MATRIX INVERSE_PROJECTION_MATRIX CAMERA_POSITION CAMERA_DIRECTION This is done by appending [gl_ViewIndex] after substitution. For example, VIEWPROJECTION_MATRIX[1][2] becomes qt_viewProjectionMatrix[gl_ViewIndex][1][2]. Change-Id: I546f465fb31aa497505f45d0197f564810593fb3 Reviewed-by: Qt CI Bot <[email protected]> Reviewed-by: Andy Nichols <[email protected]>
* Enable multiview for basic Default/PrincipledMaterial usageLaszlo Agocs2024-02-081-3/+3
| | | | | | | | | | | | ...and for 2D content. Thus the xr_simple example is functional with QT_QUICK3D_XR_MULTIVIEW=1 and a 3D API that has multiview support. Apart from the modelview-projection matrix, any other camera-related features are not yet multiview-aware. Change-Id: Ic4f07d6da375ab3ebf57b2984a020d38790ae392 Reviewed-by: Christian Strømme <[email protected]> Reviewed-by: Qt CI Bot <[email protected]>
* AssetImporter: Add name filter for input file dialogTasuku Suzuki2024-02-021-0/+10
| | | | | Change-Id: Ie69a05dde7643d05b5542e1419aa35e59b9bf4b0 Reviewed-by: Jonas Karlsson <[email protected]>
* Per camera storageChristian Strømme2023-12-041-4/+7
| | | | | | | | | | | | | | | | Extends the per extension storage with per camera storage as well, allowing to properly store data specific to the camera being used. Note that this still has some limitation, e.g., it's expected that the camera instance is unchanged within a frame. Since the frustum culling is now done when building the sorted node lists, it's applied more widely (when enabled), which means, e.g., the shadow map will also have its nodes culled. This is a problem due to how the shadow map code is implemented, as it's already doing it wrong, it something that needs to be fixed separately. Taks-number: QTBUG-116570 Change-Id: I9c77648ed82635f14f26ab8260c8b3e7127dd0f1 Reviewed-by: Andy Nichols <[email protected]>
* Move most of QSSGRhiContext to the private classLaszlo Agocs2023-11-241-1/+2
| | | | | | Change-Id: I6016c54dde9564f85469b70034a215f66879fb23 Reviewed-by: Christian Strømme <[email protected]> Reviewed-by: Qt CI Bot <[email protected]>
* Fix export functionality in the custom material editorChristian Strømme2023-11-061-1/+1
| | | | | | | | | | The export functionality didn't work on windows as the path for exporting the component was not resolved correctly. Fixes: QTBUG-118806 Pick-to: 6.6 6.5 Change-Id: I050c6dad23e4bd64fdba03814d961f6e49a683a7 Reviewed-by: Tomi Korpipää <[email protected]>
* Expose the rendererChristian Strømme2023-10-241-1/+1
| | | | | | Task-number: QTBUG-116570 Change-Id: Id8fafb80dd19567bf3d1647d9385003ab0de32f0 Reviewed-by: Laszlo Agocs <[email protected]>
* Expose the render context interfaceChristian Strømme2023-10-242-2/+4
| | | | | | | | | | | | The QSSGRenderContextInterface is the main hub for all the internal sub-components. Note that while the RCI is exposed, it doesn't mean all the sub-components are exposed as well. Depending on the component some might be exposes with a semi-public API while others still will require the use of private headers. Task-number: QTBUG-116570 Change-Id: Iad0d0069fbdfd8b66af6f3d020659395989b49dc Reviewed-by: Laszlo Agocs <[email protected]>
* Move the property table for the shader key into the render layer dataChristian Strømme2023-10-231-2/+3
| | | | | | | | | | | Having this as a member of the renderer is kind of weird, and also leads to interesting, and hard to read code for those not already initiated. The class is due for some modernization, but for now just continue keepng the cached version around. Task-number: QTBUG-116570 Change-Id: If0cf8b7b75319d2f57199bce821e8d68461817ea Reviewed-by: Laszlo Agocs <[email protected]>
* Include what you need: <QPointer>Marc Mutz2023-10-132-0/+4
| | | | | | | | | | | | | | | | | | | | | | | All these TUs relied on transitive includes of qpointer.h, maybe to a large extent via qevent.h, though, given that qevent.h is more or less the only public QtBase header that includes qpointer.h, something else seems to be at play here. Said qevent.h actually needs QPointer in-name-only, so a forward declaration would suffice. Prepare for qevent.h dropping the include. The algorithm I used was: If the TU mentions 'passiveGrabbers', the name of the QEvent function that returns QPointers, and the TU doesn't have qpointer.h included explicitly, include it. That may produce False Positives, but better safe than sorry. Otherwise, in src/, add an include to all source and header files which mention QPointer. Exception: if foo.h of a foo.cpp already includes it, don't include again. Task-number: QTBUG-117670 Change-Id: I7840274c42334e6d2f7baf32a09dbd25bc462fc3 Reviewed-by: Fabian Kosmale <[email protected]>
* Mark the whole repo with QT_NO_CONTEXTLESS_CONNECTAhmad Samir2023-09-043-7/+7
| | | | | | | | | | | | | | By adding it to the default build flags via .cmake.conf. This disables the 3-arg QObject::connect() overload: QObject::connect(sender, signal, functor) For details see: https://lists.qt-project.org/pipermail/development/2023-July/044141.html Task-number: QTBUG-116296 Change-Id: I6ccca69c7cf214e826e35d8c221b067a2587c79a Reviewed-by: Volker Hilsheimer <[email protected]>
* Prevent ShaderBuildMessage from being exported twice to QMLIvan Solovev2023-08-111-1/+8
| | | | | | | | | | | | | | We need to export the type as a value type, and also export its enums in a namespace. This resulted in duplicated entries in the qmltypes file, which confuses Qt Design Studio. Fix it by introducing a derived class, and using it for exporting the enums. Fixes: QTBUG-115362 Pick-to: 6.6 6.5 Change-Id: I7a052592961fcbcecb8b48757a9babaa79d81f77 Reviewed-by: Fabian Kosmale <[email protected]>
* Replace specific material getters from the subset renderableChristian Strømme2023-08-011-2/+3
| | | | | | | | | | | Part of the effort to improve the API of some of our internal class'. This means we don't need to expose the specific material types, this is the callers responsibility to check. All our graph objects have type info. Pick-to: 6.6 Change-Id: I1fe13962920a827e4be7c6d4842d7d36b6bb6fcd Reviewed-by: Andy Nichols <[email protected]>
* Fix qmllint errors in materialeditor toolAndy Nichols2023-06-1214-226/+251
| | | | | | Pick-to: 6.6 Change-Id: I23c6fa951b566ddf87390ec6c14276cf5b63b0e8 Reviewed-by: Christian Strømme <[email protected]>
* Update a tool, meshdebugInho Lee2023-06-061-0/+17
| | | | | | | | A new mesh version has a separate morph target buffer and this information should be printed by meshdebug. Change-Id: I475e54f63e75ef73368012cfb505f981c093f0ca Reviewed-by: Jonas Karlsson <[email protected]>
* Adapt to the RHI API conceptLaszlo Agocs2023-05-222-3/+2
| | | | | | Task-number: QTBUG-113331 Change-Id: I48aa18f2fb4a76d8eabe665d22e1bb3de5b3018c Reviewed-by: Volker Hilsheimer <[email protected]>
* Remove last usage of QSSGRefChristian Strømme2023-05-191-1/+1
| | | | | | Pick-to: 6.5 Change-Id: I889c0c5f45f7b93b9512bd81fb39ae026df7056d Reviewed-by: Andy Nichols <[email protected]>
* Replace QSSGRef usage in the render context interfaceChristian Strømme2023-04-181-14/+13
| | | | | | | | | | | The QSSGRef class is probably going away and as with our internal pass API the RCI, as a central piece of our code, should also move away from using the QSSGRef class. The ownership is also much stricter controlled now. Pick-to: 6.5 Change-Id: I123974a54287b93a7eb99557470501de7d4a2b66 Reviewed-by: Andy Nichols <[email protected]>
* RenderContextInterface improvementsChristian Strømme2023-04-182-9/+13
| | | | | | | | | | | | Move the RCI to the window attachment and get rid of the old global storage. The RCI is per-window, same as the window attachment, so it make sense to keep them together. The releaseCachedResources is also tied to the window attachment, which means we can have a more flexible callback system, for when cached resources should be cleared. Pick-to: 6.5 Change-Id: If32318cdc492ed2303062523a969395706c84a4e Reviewed-by: Christian Strømme <[email protected]>
* Remove data() function from QSSGRefChristian Strømme2023-03-311-1/+1
| | | | | | | | | | | There's no need to have multiple function that return the raw pointer from QSSGRef. The data() function is also not compatible with e.g., smart pointers from the standard library, which we eventually will replace some of the existing QSSGRef usages with. Pick-to: 6.5 Change-Id: If9a3c3cac65123dca7a310501caa5ee942eda7a5 Reviewed-by: Andy Nichols <[email protected]>
* Fix QtQuick style issue in the custom material editorChristian Strømme2023-03-221-2/+6
| | | | | | | | | The qtquickcontrols2.conf file is not picked up when under the URI resource path. Pick-to: 6.5 Change-Id: I980fb01818711a256e01d462af4dba5bd05184d5 Reviewed-by: Kaj Grönholm <[email protected]>
* Asset: fix to handle embedded texturesInho Lee2023-02-281-1/+1
| | | | | | | | | | | | | | | | | | Embedded textures are not always used by "*" + number format. Since GetEmbeddedTextureAndIndex in assimp supports to get index and name at the same time, it will be used instead of previous logic. If an embedded texture has a compressed format, the format hint should be delivered to QImageReader. TextureData has a member, 'fmt', related to the image format, but it was not used. It will be used as the format hint, so the type is changed to QByteArray. Fixes: QTBUG-110918 Pick-to: 6.5 Change-Id: I33cc7871073bca383fc007dd8eaac38ebe7910ab Reviewed-by: Christian Strømme <[email protected]>
* Replace usage of QML settings api from labsChristian Strømme2023-02-224-4/+4
| | | | | | | | | As suggested by the warning from code using the labs settings api, we should from 6.5 use the one in core. Pick-to: 6.5 Change-Id: I0b439dc15a932e0d99b3e5cd76d01f6ca8211b22 Reviewed-by: Andy Nichols <[email protected]>
* Fix setting wrong option key in the balsamuiChristian Strømme2023-02-161-2/+3
| | | | | | | | | | This fixes bug introduced in 44eb90389 Taks-number: QTBUG-111000 Pick-to: 6.5 6.4 Change-Id: I639f1ee2c0e882d898e339ab61d73c235ec5b87d Reviewed-by: Kristoffer Skau <[email protected]> Reviewed-by: Andy Nichols <[email protected]>
* Move toString() functions in base-types out of the global scopeChristian Strømme2023-01-181-3/+3
| | | | | | | | | Moved under the QSSGBaseTypeHelpers scope and moved out of the header as there's no need to inline all these. Pick-to: 6.5 Change-Id: Ibdf298d598d01c6e1ed9d81c0fc5120ccb910fb0 Reviewed-by: Andy Nichols <[email protected]>
* balsam: support asset re-importPaul Olav Tvete2022-12-081-5/+35
| | | | | | | | | | Add support for overriding materials and properties of imported assets. The modifications are read from a JSON file. Task-number: QTBUG-106952 Change-Id: I905e1337bc4bcf9c371152f4f3e43472d2b6d2ef Reviewed-by: Christian Strømme <[email protected]>
* balsam: Allow processing of also large imagesKai Köhne2022-11-301-0/+4
| | | | | | | | | See also https://www.reddit.com/r/QtFramework/comments/yle4im/how_to_change_qt_global_parameters/ Change-Id: I269d12fc34d21322c8848784255ab03600576d18 Reviewed-by: Andy Nichols <[email protected]>
* Remove unneeded custom data typesPaul Olav Tvete2022-11-071-2/+2
| | | | | | Task-number: QTBUG-106952 Change-Id: I36a3fbb756dafe89ebc95ffd43c9e78fd06eaa02 Reviewed-by: Christian Strømme <[email protected]>
* Automatically generate Level of Detail meshes on importAndy Nichols2022-11-011-0/+6
| | | | | | | | | | | | | | | | | | | | | | | | | | | Introduces a new version of the mesh file format that stores additional levels of detail that are generated by using using meshoptimizer to simplify the original geometry of each subset. The simplify process returns a list of new range of index values that should be used when rendering a mesh at a particular level of detail, as well as an ideal "distance" which manifests itself as a ratio of how many pixels will be used to render the mesh. The idea is that if a model will only be rendered as a small ratio of screen space, then the full resolution geometry is not needed to get the same result. Part of the refinement of the simplification process is to add additional vertices to lod levels where the attributes such as vertex normals no longer make sense using the original index. In this case the attributes are recalculated using the new indexes of the simplified level of detail. Currently generation of mesh levels of detail is off by default since some tweaking may be necessary for the best results. This should instead be an explicit choice by the user on import via the tooling. Change-Id: I2a2bc0b75aa30284a40241bdf7b90ff072a0fcac Reviewed-by: Laszlo Agocs <[email protected]>
* Fix build warning when brace-initializing UrlView structVolker Hilsheimer2022-10-271-2/+2
| | | | | | | | | | | | Apple's clang says: error: suggest braces around initialization of subobject [-Werror,-Wmissing-braces] and errors out with the -Werror default configuration. Change-Id: I334a7a0b445d35e2a94cc66200691df840cb3e7e Reviewed-by: Christian Strømme <[email protected]>
* Add a persistent, disk-based cache for the shader generatorLaszlo Agocs2022-10-262-44/+44
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Implemented for default/principled materials and custom materials and effects. For effects some changes are done to avoid the problem of colliding hash keys in the first level effect cache. In a somewhat similar manner, the QQsbCollection is updated to use different kind of keys (like a string with a hex sha, not a uint64 from qHash). Relies on a new shadertools API that allows shader baking to skip target versions that fail in SPIRV-Cross. (because we try to bake everything, but so have to deal gracefully with things like GLSL ES 100 not having textureSize() etc.) The cache location is the QStandardPath::CacheLocation, so e.g. Users\<user>\AppData\Local\<appname>\cache. The .qsbc file is under a directory q3dshadercache-<buildabi>. .qsbc files now store the Qt version they were built with. If there is a mismatch, the file will be ignored, and upon saving it will be replaced with new content. (this is identical to how the low-level pipeline cache works when there is a mismatch) This is an important change for pregenerated materials: previously any .qsbc file was loaded, whereas now even a difference in the Qt patch version will make the prebuilt data to get ignored. This is in theory not great for apps that build against one version and then replace the Qt libs with a newer version without rebuilding the app. Then again there is no compatibility guarantee for the material shader generator, so rejecting something from an older version is the right thing to do. To facilitate this and the key change, the version of .qsbc files is bumped. We do not support loading the old format. Eventually an API on some level to control the file location may get added (just like Qt Quick has QQuickGraphicsConfiguration APIs to control the low-level pipeline cache). This may be interesting in particular because the usual caveats apply with multiple windows: the last window that closes wins, that's the one that gets to write out its cache. (there is only a single file) The story here is identical to Qt Quick: If one wants to have proper control over this, they need to explicitly specify the file location on a per-window basis (not per View3D, though, which may complicate things). The persistent caching can be disabled using the existing mechanisms, meaning disabling the Qt Quick pipeline cache (or the Qt 5 era QOpenGLShaderProgram disk cache) will also disable this one. One can set Qt::AA_DisableShaderDiskCache, or the env.var QT_DISABLE_SHADER_DISK_CACHE or its QSG_RHI_... equivalent. This way we do not need a Quick3D-specific setting at all. For development purposes it can be handy to disable just the loading of a possibly exising cache, while keeping the saving intact (i.e. this allows overwriting any existing cache data when changing the material shader generator, without running into problems due to loading cached data from a Quick3D build without the latest changes). To do this, set QT_QUICK3D_NO_SHADER_CACHE_LOAD to a non-zero value. To confirm what is happening under the hood, set the usual QT_RHI_SHADER_DEBUG env.var. This will cause: - upon QSSGShaderCache construction the printing of something like the following, if there is an existing cache: Attempting to seed material shader cache from <filename> Loaded 15 shader pipelines into the material shader cache - then on each default/custom material lookup it shows one of: * nothing (if the highest level, runtime cache were hit, i.e. we saw the material before during the run of the application), * or it prints something along the lines of "Loading rhi shaders from disk cache for <cache key> (<material key>)". That is when a new QSSGRhiShaderPipeline was created on the cheap path, i.e. when the QShaders are coming from the .qsbc file. * Finally, if it prints the shader source code then it's on the slow path with generating and processing the shaders. A dedicated documentation page will likely need to be introduced separately. That is however postponed until we get an idea if there will be dedicated APIs to control the cache location or not. Change-Id: I166f72c3f2f4ea0fbcf0b0af658117cdc537b241 Reviewed-by: Andy Nichols <[email protected]>
* Improve the renderable typesChristian Strømme2022-10-211-2/+2
| | | | | | | | | | | | | Store the global bounds in the renderable as we'll have use for that later, e.g., culling. Also, the type was removed from the renderable flags, as the type will never be reset or changed during the lifetime of the object. We know the type at creation and is never going to change. Having a fixed type means it's easier to make sensible assumption about the object when it's e.g., created. Task-number: QTBUG-106958 Change-Id: I459734725792b4a7d7fbe4fa88707277e3377b68 Reviewed-by: Christian Strømme <[email protected]>
* Custom material editor: Update the window titleChristian Strømme2022-10-211-1/+1
| | | | | | | "Material Editor" is a bit to generic and conflicts with DS. Change-Id: I716a15c609dd765c37769dba91b6e22ba4530b70 Reviewed-by: Laszlo Agocs <[email protected]>
* Custom material editor: Fix the import dialogsChristian Strømme2022-10-211-4/+4
| | | | | | | | .. for some reason this is needed Pick-to: 6.4 Change-Id: Ib25d9c0a350b5eb367efd491387be452d8e486b4 Reviewed-by: Andy Nichols <[email protected]>
* Make QQsbCollection usable as a runtime cache as wellLaszlo Agocs2022-10-202-29/+15
| | | | | | | | | | | | Move some stuff into a base class (QQsbCollection) then have two subclasses, the existing QIODevice-based one as QQsbIoDeviceColletion and a new QQsbInMemoryCollection. When using the latter's save and load functions, the file format is exactly the same as with the former. Change-Id: I1984f39b10c43d70b91ad341c12b0f8b6beb3b0c Reviewed-by: Andy Nichols <[email protected]> Reviewed-by: Qt CI Bot <[email protected]>
* Remove QSSGSceneDesc::Scene::create()Paul Olav Tvete2022-10-171-3/+3
| | | | | | | | | | After the previous change to a more dynamic data structure, create() just forwarded to new. Using new directly instead makes the code more readable and less error prone. Task-number: QTBUG-106952 Change-Id: Ib9b6e80b8b5f55081b0d07e8e9517c103e917998 Reviewed-by: Christian Strømme <[email protected]>
generated by cgit v1.2.3 (git 2.25.1) at 2025-08-21 02:32:38 +0000