textureSize() is not available in opengl 2.1, so we calculate the xray error image scale outside the shader (which is also a theoretical performance improvement because now the scale does not get computed over and over for each pixel)
These numbers end up beautifully on round numbers if you were to just scale them to their original size and apply these mysterious translation operations.
Also do away with all of the Inkscape and Sodipodi metadata. They don't make it any easier to read.
Contributes to issue CURA-7262.
QImage's bytes are aligned to memory words per column of pixels. That means that one of these columns contains 99% valid image data, but with several bytes of unassigned noise at the end. How many of these padding bytes there are would depend on the image size, i.e. Cura's window size. In the end, the total number of bytes in the image ends up slightly more than w*h*3. As a result, Cura would crash because it couldn't reshape the image. Reshaping was completely unnecessary anyway, but this random noise was giving false positives also.
But how do you then get only the actual pixels from each column of data? We can't just go iterating over this array, as that would be an iteration of thousands of columns which is prohibitively slow in Python. No, we're going to do some Numpy magic. We're going to create a class that pretends to be a Numpy array. Give this class some data and say that this data has a certain pixel size but also a certain STRIDE LENGTH. This stride length can be the length of the actual pixel data. As a result when Numpy sees this object it will read out the data using these strides, all done efficiently within the C code of Numpy.
Framerate is fantastic on my computer. No problems at all. Pretty powerful computer though. But also a big 5k screen. Still no problem for Numpy. Seems to be decently efficient.
Took me quite a while to figure all of this out.
Contributes to issue CURA-7262.
That's what that function is meant for. It's executing essentially the same, since before it was executed right after the call to _checkSetup() and now it's at the end inside that function.
Contributes to issue CURA-7262.
It was using a bit of programmer jargon there that was very hard to understand. Initially I was coming in here to remove the space at the end of the translated string but I'm now changing all of it to make it more clear to the user as well.
Done as a 5 minute fix.
Reproduction steps:
* In the Material Manager, click on 'Import material'.
* Change the file type in the file dialogue to 'All files'.
* Select any .def.json file, for example from Cura's own resources/definitions folder.
Previously this crashed the application. Now it shows the user an error message instead.
Fixes Sentry error CURA-D4.
This happens when the printer is changed, such as when changing the Machine Width in the machine settings dialogue. It updates these menus then and if not all extruders are defined it'll give the following errors:
2020-03-20 16:56:57,839 - WARNING - [MainThread] UM.Qt.QtApplication.__onQmlWarning [406]: file:///home/trin/Gedeeld/Projects/Cura/resources/qml/Menus/SettingsMenu.qml:51: TypeError: Cannot read property 'isEnabled' of undefined
2020-03-20 16:56:57,840 - WARNING - [MainThread] UM.Qt.QtApplication.__onQmlWarning [406]: file:///home/trin/Gedeeld/Projects/Cura/resources/qml/Menus/SettingsMenu.qml:44: TypeError: Cannot read property 'isEnabled' of undefined
2020-03-20 16:56:57,841 - WARNING - [MainThread] UM.Qt.QtApplication.__onQmlWarning [406]: file:///home/trin/Gedeeld/Projects/Cura/resources/qml/Menus/MaterialMenu.qml:26: TypeError: Cannot read property 'isEnabled' of undefined
2020-03-20 16:56:57,841 - WARNING - [MainThread] UM.Qt.QtApplication.__onQmlWarning [406]: file:///home/trin/Gedeeld/Projects/Cura/resources/qml/Menus/MaterialMenu.qml:28: TypeError: Cannot read property 'material' of undefined
Done during Turbo Testing and Tooling.
This then updates in a fraction of a millisecond later before things even appear on the screen, so it's not so bad to not have a tooltip or to have it show 0 then.
Done during Turbo Testing and Tooling.
This prevents a QML warning when the ListModel is created before the property provider is. This QML warning looked like this:
2020-03-20 15:04:30,012 - WARNING - [MainThread] UM.Qt.QtApplication.__onQmlWarning [406]: file:///home/trin/Gedeeld/Projects/Cura/cura/../resources/qml/MachineSettings/ComboBoxWithOptions.qml:74: TypeError: Cannot call method 'match' of undefined
Done during Turbo Testing and Tooling.
This widget can't set its width to the width of its children, since its children anchor to its right-hand side which depends on the width.
This fixes the following warning in the log:
2020-03-20 14:25:20,351 - WARNING - [MainThread] UM.Qt.QtApplication.__onQmlWarning [406]: file:///home/trin/Gedeeld/Projects/Cura/plugins/PrepareStage/PrepareMain.qml:16:5: QML ActionPanelWidget: Binding loop detected for property width
Done during Turbo Testing and Tooling.
Fixes the following QML warning in the logs:
2020-03-20 14:02:13,482 - WARNING - [MainThread] UM.Qt.QtApplication.__onQmlWarning [406]: file:///home/trin/Gedeeld/Projects/Cura/resources/qml/PrintSetupSelector/Recommended/RecommendedSupportSelector.qml:125:13: Unable to assign [undefined] to int
Instead it now assigns the first extruder before a printer has been created. The first extruder also doesn't exist yet at this point, but this doesn't really matter because there is no interface to show yet at this point.
Done during Turbo Testing and Tooling.
This removes a warning from our log that this image couldn't be found. Instead it now draws a completely transparent image there.
The warning that is no longer in the log now looked like this:
2020-03-20 13:52:27,980 - WARNING - [MainThread] UM.Qt.Bindings.Theme.getImage [133]: No image header_pattern defined in Theme
Done during Turbo Testing and Tooling.
To do this, I'm giving more power to the NumericTextFieldWithUnit QML element, to allow an arbitrary minimum and maximum. Enforcing this minimum and maximum is fairly simple with a JavaScript hook. This hook is necessary because the DoubleValidator allows intermediary values which defeats the purpose, essentially allowing any number as long as it has the correct number of digits.
Printers larger than 2km would start to give overflow errors in its X and Y coordinates. Z is okay up to about 9 billion kilometres in theory, since we don't need to do any squaring math on those coordinates afaik. In practice I'm doing this because at very high values the Arranger also gives errors because Numpy can't handle those extremely big arrays (since the arranger creates a 2mm grid).
Fixes Sentry issue CURA-CB.
Given that:
* stair stepping is intended to be used to reduce the adhesion between support and the model, where the support rests on the model, and to reduce scarring, and
* PVA doesn't suffer from scarring or adhesion issues due to its water-solubility, and
* our current stair stepping algorithm has some bugs, causing support to disappear sometimes
we've decided to set stair stepping to 0 for all PVA profiles for UM3, UMS5 and UMS3.
Contributes to issue CURA-7311.
