aboutsummaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* Fix QQmlImportsPrivate::resolvedUri()v5.10.0-rc3v5.10.0J-P Nurmi2017-11-301-4/+6
| | | | | | | | | | | | | Qt Quick Controls 2 styles have a version in the middle of the path/URI (e.g. qml/QtQuick/Controls.2/Material). The resolvedUri() helper method, which is used by addFileImport() for implicit imports, was never taught to deal with this type of versioning. It was only attempting to remove the version from the end. Change-Id: Ibdf23dc6c3b0794527d5f9330602858291c23e01 Task-number: QTBUG-64868 Reviewed-by: Simon Hausmann <simon.hausmann@qt.io> Reviewed-by: Lars Knoll <lars.knoll@qt.io>
* Fix issue with circular singleton instantiationsv5.10.0-rc2Michael Brasser2017-11-285-1/+39
| | | | | | | | | While a recursion check exists and works, it can lead to instanting the same singleton multiple times (leaking all but one copy). Change-Id: Icf342aad71c5cb225488262341517d95786e1f84 Reviewed-by: Erik Verbruggen <erik.verbruggen@qt.io> Reviewed-by: Simon Hausmann <simon.hausmann@qt.io>
* blacklist tst_qquickflickable::rebound on macOSv5.10.0-rc1Shawn Rutledge2017-11-231-1/+2
| | | | | | | | | It's not only on 10.10 Task-number: QTBUG-26696 Task-number: QTBUG-36804 Change-Id: I6652cc3c32cc0219e58f7bf6d9584e3041ad34e8 Reviewed-by: Shawn Rutledge <shawn.rutledge@qt.io>
* never skip updating passive grabbers, or ungrab, on mouse releaseShawn Rutledge2017-11-231-8/+11
| | | | | | | | | | If you drag a ListView with the right mouse button, it was reacting, but not rebounding after release. And if a PointHandler is used to show feedback about mouse position, it also did not get the release. This was due to returning early from deliverMouseEvent. Change-Id: I24b39e4769d6824d3bd1f400dbf1f973bb29fbb6 Reviewed-by: Jan Arve Sæther <jan-arve.saether@qt.io>
* Add changes file for Qt 5.10.0Antti Kokko2017-11-211-0/+188
| | | | | | | Change-Id: I49d734212a4d9b4002db012c2149e1614292adeb Reviewed-by: Shawn Rutledge <shawn.rutledge@qt.io> Reviewed-by: Alex Blasche <alexander.blasche@qt.io> Reviewed-by: Laszlo Agocs <laszlo.agocs@qt.io>
* Merge remote-tracking branch 'origin/5.9.3' into 5.10.0Liang Qi2017-11-214-5/+89
|\ | | | | | | Change-Id: Id2b11dccb04fc99d10608f690f5827bfa9b2114c
| * Fix crash when accessing pixmap on default QSGSoftwareInternalImageNodev5.9.3Allan Sandfeld Jensen2017-11-171-4/+5
| | | | | | | | | | | | | | Task-number: QTBUG-64562 Change-Id: I52e07b0d8b7a5d1cc960431dcbd1a90dd3e7e518 Reviewed-by: Jüri Valdmann <juri.valdmann@qt.io> Reviewed-by: Laszlo Agocs <laszlo.agocs@qt.io>
| * Fix crash with dangling context object pointersAndy Shaw2017-11-141-0/+2
| | | | | | | | | | | | | | | | | | | | | | | | This is a regression introduced by commit e22b624d9ab1f36021adb9cdbfa9b37054282bb8, where the object that owns the QML context would destroy the context upon destruction. Now the context may live longer and thus the context->contextObject pointer would become a dangling pointer. Task-number: QTBUG-64166 Change-Id: I1df631fa11187abdeff735d8891ad7907e8d4a3d Reviewed-by: Lars Knoll <lars.knoll@qt.io>
| * Blacklist tst_qquickwidget::enterLeave() on MacJ-P Nurmi2017-11-141-0/+2
| | | | | | | | | | | | | | Task-number: QTBUG-64397 (cherry picked from commit 5f16aa795d39969d93b520861a1e86729c7db90e) Change-Id: I28268ea87b81dd1f7dbf8bb5a8eb421962cc5f31 Reviewed-by: Jani Heikkinen <jani.heikkinen@qt.io>
| * Add changes file for Qt 5.9.3Jani Heikkinen2017-11-121-0/+73
| | | | | | | | | | | | | | Task-number: QTBUG-64181 Change-Id: I85a6150e00143b379143c353a37e844cb9708627 Reviewed-by: Shawn Rutledge <shawn.rutledge@qt.io> Reviewed-by: Simon Hausmann <simon.hausmann@qt.io>
| * Improve encapsulation of the the IR de-serialization from QtQuick CompilerSimon Hausmann2017-11-121-1/+9
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The code used by QQC to deserialize the IR requires setting the javaScriptCompilationUnit member in order to connect the generated C++ code. Knowledge of the QmlIR::Document data structure layout on the side of the generated code (thus application) has its downsides though (see referenced bug). We can avoid that dependency easily by doing the entire de-serialization on the QtQml library side. The old "API" (load member function) is still around until the qqc change is also in. Task-number: QTBUG-63474 Change-Id: I239838afacc71474c86114b5b05679ff36e4c2e2 Reviewed-by: Lars Knoll <lars.knoll@qt.io>
* | Merge 5.10 into 5.10.0Oswald Buddenhagen2017-11-2087-1391/+1672
|\ \ | | | | | | | | | Change-Id: I68e1137c57cdb4c585dd44b0528744bde7df047a
| * | Fix outdated FDL license headerKai Koehne2017-11-151-4/+4
| | | | | | | | | | | | | | | Change-Id: I9ea98ccf98c7421818c40fe0128a435db0fa28e7 Reviewed-by: Liang Qi <liang.qi@qt.io>
| * | Fix outdated BSD license headerKai Koehne2017-11-1550-1146/+744
| | | | | | | | | | | | | | | Change-Id: Ib1fe267c23ea9fce9bcc0a91ed61081260338460 Reviewed-by: Liang Qi <liang.qi@qt.io>
| * | add PointHandlerShawn Rutledge2017-11-154-0/+219
| | | | | | | | | | | | | | | Change-Id: Iad9ba3f8121ff22136e85413ec610bc1ee361e37 Reviewed-by: Jan Arve Sæther <jan-arve.saether@qt.io>
| * | a PointerHandler can veto overtaking of grab by an ItemShawn Rutledge2017-11-141-7/+7
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | In order to make sure of it, we have to do the permission check in the legacy call chain too: QQuickFlickable::filterMouseEvent() -> QQuickItem::grabMouse() -> QQuickWindowPrivate::setMouseGrabber() -> QQuickEventPoint::setGrabberItem() -> QQuickPointerHandler::approveGrabTransition() Change-Id: Ice3499bc56e7c89fb43b054ddcd0098fea94ba2a Reviewed-by: Jan Arve Sæther <jan-arve.saether@qt.io>
| * | pinchHandler manual test: more PinchHandlers; TapHandler for z-orderShawn Rutledge2017-11-141-58/+101
| | | | | | | | | | | | | | | | | | | | | grabPermissions are OK by default. Change-Id: I1e9564415a12e4cd7a9213d11e5374d619239099 Reviewed-by: Jan Arve Sæther <jan-arve.saether@qt.io>
| * | QQuickWindowPrivate::deliverToPassiveGrabbers: localize the eventShawn Rutledge2017-11-141-1/+3
| | | | | | | | | | | | | | | | | | | | | | | | Always localize the pointer event to the handler's parent's coordinate system before sending it to a handler. Change-Id: I3006329a07cc9439b472a475444bcd4a0336ad0c Reviewed-by: Jan Arve Sæther <jan-arve.saether@qt.io>
| * | Add a manual test quickwidgetviewerFriedemann Kleint2017-11-142-0/+95
| | | | | | | | | | | | | | | | | | | | | | | | Add a simple command line viewer allowing for showing QML in QQuickWidget. Change-Id: Ie77e287e3fed928ab4f148e1f250ce59936048ca Reviewed-by: Shawn Rutledge <shawn.rutledge@qt.io>
| * | PointerHandler: add grabPermissions, enforce in setExclusiveGrabShawn Rutledge2017-11-144-19/+160
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | As soon as we enable the concept that PointerHandlers can use passive grabs to lurk, monitor all movements, and then steal the passive grab, they can fight over the grab. For example if there are two items with PinchHandlers, and two or more touches occur within bounds for both, then each update event can cause the other PinchHandler to steal the grabs and become active. So we replace stealing with negotiation: the handler which wants to take over the grab checks its own flags to see whether that's allowed, and the handler which is about to lose its grab also has the right to approve or deny the takeover (just as QQuickItem has had keepMouseGrab and keepTouchGrab for a long time.) Additionally, if one handler wants to cancel another handler's grab without taking over (simply set the grabber to null), it must be approved. A single-point handler can simply call setExclusiveGrab, with the expectation that permission may be granted or denied. A multi-point handler only wants to grab all points if grabbing all of them will be allowed, otherwise grab none; so it calls canGrab on each point to check beforehand. Thus, when two handlers are competing for the same grabs, one or both can be prevented from stealing from each other, or from Handlers in general, or from Items, or some combination. Change-Id: I5c733b2b8995ce686da0be42244394eeee82a268 Reviewed-by: Jan Arve Sæther <jan-arve.saether@qt.io>
| * | Let passive-grabbing PointerHandlers see all point updatesShawn Rutledge2017-11-145-96/+111
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | even if all points are accepted or grabbed. A passive grab isn't much good if there are cases where the handler is prevented from monitoring. This enables e.g. the PinchHandler to steal the grab when the right number of touchpoints are present and have moved past the drag threshold, and enables completion of a couple of autotests. Change-Id: I78dc6fc585f80bfb3c13e0c6e757ef815fb94afe Reviewed-by: Jan Arve Sæther <jan-arve.saether@qt.io>
| * | blacklist tst_QQuickFramebufferObject::testInvalidate on macOSShawn Rutledge2017-11-141-0/+3
| | | | | | | | | | | | | | | | | | Task-number: QTBUG-64470 Change-Id: I0119b2fc6a4110c21d1cf083516d32302506aae4 Reviewed-by: Shawn Rutledge <shawn.rutledge@qt.io>
| * | QQuickEventPoint: make ungrab/cancel notifications more consistentShawn Rutledge2017-11-142-22/+20
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | setGrabberPointerHandler is now implemented more similarly to setGrabberItem. One improvement is that in tests/manual/pointer/pinchDragFlingMPTA.qml the MPTA deactivates when the PinchHandler takes over its touchpoint grabs. Change-Id: I0bd4f143b5f25f1b393839f86c2a7802f1fa1886 Reviewed-by: Jan Arve Sæther <jan-arve.saether@qt.io>
| * | Merge "Merge remote-tracking branch 'origin/5.9' into 5.10" into ↵Liang Qi2017-11-1319-48/+214
| |\ \ | | | | | | | | | | | | refs/staging/5.10
| | * \ Merge remote-tracking branch 'origin/5.9' into 5.10Shawn Rutledge2017-11-1319-48/+214
| | |\ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Conflicts: src/quick/items/qquickwindow.cpp src/quick/items/qquickwindow_p.h tests/auto/quick/quick.pro Change-Id: Ia12f20e95fb151bbbc75dbf187364a924cd0bc7a
| | | * | Make sure we remove stopped animators from the list of rootsGunnar Sletta2017-11-101-1/+3
| | | | | | | | | | | | | | | | | | | | | | | | | Change-Id: I89b36ee7d03ac6b8d07b24c656d3311728e8f9c3 Reviewed-by: J-P Nurmi <jpnurmi@qt.io>
| | | * | Blacklist tst_qquickwidget::enterLeave() on MacJ-P Nurmi2017-11-091-0/+2
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Task-number: QTBUG-64397 Change-Id: I28268ea87b81dd1f7dbf8bb5a8eb421962cc5f31 Reviewed-by: Gatis Paeglis <gatis.paeglis@qt.io>
| | | * | Make tst_qquickframebufferobject pass on macOSJ-P Nurmi2017-11-091-2/+4
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The test has not been run in the CI, so the problem went unnoticed for a long time. Change-Id: I42a44a5fb89c0bd78e8997d4841e85672c73acdb Reviewed-by: Laszlo Agocs <laszlo.agocs@qt.io> Reviewed-by: Shawn Rutledge <shawn.rutledge@qt.io>
| | | * | Fix tests/auto/quick/quick.proJ-P Nurmi2017-11-091-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The following auto tests have not been run in the CI at all: - tst_qquickanimatedsprite - tst_qquickframebufferobject - tst_qquickopenglinfo - tst_qquickspritesequence - tst_qquickshadereffect Change-Id: Iacc832563fd2c002eef480fa4d42469d852adc0f Reviewed-by: Frederik Gladhorn <frederik.gladhorn@qt.io>
| | | * | Fix tst_qquickshadereffectJ-P Nurmi2017-11-092-5/+5
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The expected signal counts were not matching. Since the test has not been run in the CI, it went unnoticed. Furthermore, the test crashed due to a missing null pointer check. Change-Id: Iff80a2ea17832eb7bc531ac9eb2fc482f2c69e38 Reviewed-by: Laszlo Agocs <laszlo.agocs@qt.io>
| | | * | docs and example: clarify the usage of Flickable contentX/contentYShawn Rutledge2017-11-082-3/+17
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | It's such a common mistake to observe that they normally go to zero at the top-left corner, but fail to realize that they won't always do that. Task-number: QTBUG-64219 Task-number: QTBUG-22894 Task-number: QTBUG-27884 Change-Id: I6bc81d4761debdaff8fb3366bf1e944241207157 Reviewed-by: J-P Nurmi <jpnurmi@qt.io>
| | | * | stabilize and optimize tst_QQuickListView::QTBUG_34576_velocityZeroShawn Rutledge2017-11-081-8/+9
| | | |/ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | - don't use QTRY_VERIFY or QTRY_COMPARE if there's nothing to wait for, because it will always wait a short time and add needless delay to the test - QVERIFY(QSignalSpy::wait())'s should perhaps not be done in sequence, in case the second signal already happened while we were waiting for the first. QTRY_VERIFY(QSignalSpy::count() > 0) should be more reliable in such a case, as long as we are sure the count started at zero before the behavior which was supposed to make the signal be emitted. - 1000ms is probably not long enough to wait for ListView velocity change on a slow CI system. - according to the comment "verify that currentIndexChanged is triggered" we need another spy for that signal. Change-Id: I99d93a849b674ce6c81acbe91639f03933025117 Reviewed-by: Jan Arve Sæther <jan-arve.saether@qt.io>
| | | * Fix ListView::positionViewAtIndex with ListView.Contain modeJoni Poikelin2017-11-073-11/+91
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Sticky headers and footers weren't accounted for when calculating new view position causing the requested item to be left behind them. Task-number: QTBUG-63974 Change-Id: Id69579643a942e8558960b2c8b0fee980fa86947 Reviewed-by: Shawn Rutledge <shawn.rutledge@qt.io>
| | | * QQuickWindow: cleanup pointer event instancesJ-P Nurmi2017-11-071-0/+2
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Pointer event instances are QObject-children of QQuickWindow, meaning that they remain alive after ~QQuickWindow(), until execution reaches ~QObject(). Make sure the pointer event instances are cleaned up in ~QQuickWindow() to avoid accessing them later during the destruction phase. Task-number: QTBUG-61434 Change-Id: Icd4576e7581524773a3eb33864fdd64df821e0e8 Reviewed-by: Jan Arve Sæther <jan-arve.saether@qt.io> Reviewed-by: Shawn Rutledge <shawn.rutledge@qt.io>
| | | * Never create pointer events from mouseGrabberItem()Jan Arve Saether2017-11-072-15/+24
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Normally, this was not a problem, but it is problematic during QQuickWindow destruction: The list of pointer event instances are destroyed, but later the grabber is removed (through call to removeGrabber()). This queries the mouseGrabberItem(), which would create a new pointer event instance. It also has the benefit that d->pointerEventInstances are now only populated due to actual incoming events. Task-number: QTBUG-61434 Change-Id: I4e7b6f5643f3b971138a1f7c7237ee734d29783c Reviewed-by: Shawn Rutledge <shawn.rutledge@qt.io>
| | | * QQuickWidget: pass enter and leave events to the offscreen windowAndy Shaw2017-11-063-0/+46
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | By passing the enter and leave events to the offscreen window it will enable mouse areas inside a QQuickWidget to know when the mouse has actually entered or left the area if it is covering the whole item. Task-number: QTBUG-45557 Change-Id: I670ebe30e367e919c73fed449bf2bed7ca42b5fd Reviewed-by: Gatis Paeglis <gatis.paeglis@qt.io>
| | | * Fix coverity reported warningLars Knoll2017-11-061-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The variable wasn't initialized in any constructor. This didn't cause issues in practice, as it always got set in optimizeRenderList(). Change-Id: I37459bb2d51bbe2bb881aaefffd6972a5345c75d Reviewed-by: Andy Nichols <andy.nichols@qt.io>
| | | * Prevent the QML engine from registering circular dependenciesErik Verbruggen2017-11-062-1/+10
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Change-Id: Ic4fd2bde745e7dfaf0909e8cc575441bb04cefa3 Task-number: QTBUG-64017 Reviewed-by: Lars Knoll <lars.knoll@qt.io> Reviewed-by: Marco Martin <mart@kde.org> Reviewed-by: Bhushan Shah <bshah@kde.org>
| * | | Doc: Update the list of highlighted examplesTopi Reinio2017-11-131-0/+2
| |/ / | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Update the list of highlighted examples for modules in qtdeclarative. Currently highlighted examples/demos for Qt Quick are in relatively good condition, so keep the highlighting for them. Add highlighting for 'Writing QML Extensions with C++' tutorial in Qt QML. Task-number: QTBUG-60629 Task-number: QTBUG-60630 Change-Id: If7fe8705cf8614d5a811d88f44a0c00ecf3348f0 Reviewed-by: Nico Vertriest <nico.vertriest@qt.io>
| * | Remove deprecated grabber()Jan Arve Sæther2017-11-081-1/+0
| | | | | | | | | | | | | | | | | | | | | We already have grabberItem() that serves the same purpose Change-Id: I137ebab8ecc2c3924e55015e28735eefc79c2ae8 Reviewed-by: Shawn Rutledge <shawn.rutledge@qt.io>
| * | Check for duplication with url rather than uriMichael Brasser2017-11-081-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | We might have duplicated uris with different paths. This comparison is meant to check for multiple imports of the same path. Change-Id: I2d99728d32116453ae92a152337d2627c2653b79 Task-number: QTBUG-64237 Reviewed-by: Tim Jenssen <tim.jenssen@qt.io>
* | | Prevent crashes when profiling translation bindingsUlf Hermann2017-11-114-4/+39
|/ / | | | | | | | | | | | | | | | | The m_v4Function member can now be a nullptr, which means we cannot use it as ID and we cannot retrieve the source location from it. Change-Id: Ibb49a3e68cf961f9ffe2a83b2a0b83f7d04e149e Reviewed-by: hjk <hjk@qt.io> Reviewed-by: Simon Hausmann <simon.hausmann@qt.io>
* | Merge remote-tracking branch 'origin/5.9' into 5.10v5.10.0-beta4Liang Qi2017-11-0624-86/+759
|\| | | | | | | Change-Id: I8ede7e36592cd21f3e4a0a9b30dbe26bb40fe69b
| * QQmlEngineDebugService: Check QML contexts for validityUlf Hermann2017-11-052-12/+49
| | | | | | | | | | | | | | | | | | We should not operate on invalid QML contexts and once we have established a context to be valid we don't have to check the result of QQmlContextData::get anymore. Change-Id: I9106115ddf925c3572048f1fd334bdfd9a9cfca7 Reviewed-by: Lars Knoll <lars.knoll@qt.io>
| * Clean up handling of opaqueness in the SW rendererLars Knoll2017-11-032-26/+15
| | | | | | | | | | | | | | | | And check whether nine patch pixmaps are opaque as well. Change-Id: I23f2cb675b923eace849a1c0ad71efe1446c86c4 Reviewed-by: Eirik Aavitsland <eirik.aavitsland@qt.io> Reviewed-by: Volker Krause <volker.krause@kdab.com>
| * Mark layers in the SW rasterizer as opaque if possibleLars Knoll2017-11-034-10/+22
| | | | | | | | | | | | | | | | | | | | If the content of a layer completely covers every pixel of it, mark that layer as opaque so that we can avoid alpha blending where possible. Change-Id: Ia0be4e7a96ecddd31a26f353509de976bcc9e397 Reviewed-by: Eirik Aavitsland <eirik.aavitsland@qt.io> Reviewed-by: Volker Krause <volker.krause@kdab.com>
| * stabilize tst_QQuickListView::incrementalModelShawn Rutledge2017-11-031-9/+4
| | | | | | | | | | | | Task-number: QTBUG-30716 Change-Id: I0c6829ae496850d6a2cdcc349c496dfbf4e8adbb Reviewed-by: J-P Nurmi <jpnurmi@qt.io>
| * Add benchmark for event deliveryFrederik Gladhorn2017-11-036-1/+255
| | | | | | | | | | | | | | Change-Id: Ic433a832190ff3e89fae696e1eabf0c7dfa57cec Reviewed-by: Frederik Gladhorn <frederik.gladhorn@qt.io> Reviewed-by: Robin Burchell <robin.burchell@crimson.no> Reviewed-by: Shawn Rutledge <shawn.rutledge@qt.io>
| * Fix providing correct velocity when using index based scrollingPasi Petäjäjärvi2017-11-034-5/+156
| | | | | | | | | | | | | | | | | | Index based scrolling does not enable moving and flicking properties. Task-number: QTBUG-34576 Change-Id: Ief06d37115ca389027670c97ce6c0457a74d4872 Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org> Reviewed-by: Shawn Rutledge <shawn.rutledge@qt.io>
| * QQuickItemView: always honor the removeDisplaced animationAlberto Mardegan2017-11-033-4/+45
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The animation was not being performed if the delayRemove attached property was changed by the handler of the remove() attached signal. We need to run the delayed transitions not only if we have an animation for the target item, but also if we have an animation for the items being displaced. (The flag variables can safely be obtained outside of the for loop, given that their value should not change during the loop iteration) Task-number: QTBUG-57225 Change-Id: I8c138677d7dcdf63e0932ec5cf7738c0caeb2ab8 Reviewed-by: J-P Nurmi <jpnurmi@qt.io> Reviewed-by: Shawn Rutledge <shawn.rutledge@qt.io>
generated by cgit v1.2.3 (git 2.25.1) at 2024-05-04 04:53:50 +0000