Bug: Insight shape fontSize and strokeWidth

[wmoore]

Bugs reported in ​https://www.openmicroscopy.org/community/viewtopic.php?f=6&t=7834

If I create a Shape that has font size set like this

    fontSize = omero.model.LengthI(50, UnitsLength.PIXEL)
    rect.fontSize = fontSize

then Insight fails to show ALL the ROIs on an image and Insight log says

2015-06-18 12:28:42,738 ERROR [  o.o.s.a.m.v.MeasurementViewerComponent] (entQueue-0) Cannot convert server ROI into UI objects:java.lang.NullPointerException 
2015-06-18 12:28:48,973 ERROR [      o.o.s.a.m.view.MeasurementViewerUI] (entQueue-0) Problem while handling ROI No ROIShape on coord T: 0 Z: 0 

If I set strokeWidth like this:

    strokeWidth = omero.model.LengthI(5, UnitsLength.PIXEL)
    rect.strokeWidth = strokeWidth

this is ignored in the display of shapes on the image in Insight, although it is shown in the "Inspector" tab (see screenshot).

Also attached is the script for creating tested shapes and screenshot of how they look in web.

[wmoore]

strokeWidth not shown on image

[wmoore]

Script to create shapes

[wmoore]

ROIs and Shapes in Web

[jburel]

The font size should be in point not pixels.

[jburel]

The font size should be in point not pixels.
cf forum post
insight is also converting to point units

[wmoore]

Units of POINT or PIXEL are both broken, but should both work the same.

[jburel]

if the shapes are created from insight, the values seem to be correctly read and displayed.

[wmoore]

That's great, but doesn't help with the display of shapes from scripts.

[jburel]

I know. one step at a time!!!

[wmoore]

Examples on trout (user-3).

StrokeWidth in PIXELS:
https://trout.openmicroscopy.org/merge/webclient/img_detail/9461/

FontSize in PIXELS
https://trout.openmicroscopy.org/merge/webclient/img_detail/9460/

FontSize in POINTS
https://trout.openmicroscopy.org/merge/webclient/img_detail/9463/

[jburel]

I have improved the error handling in insight and pojo so the shapes with fontsize are visible.
A NPE is thrown during the conversion from Pixels to Point so we will to have a look at it.

Remaining issue is the stroke width
Cc Josh and Dominik who have both been working on the conversion

[jburel]

All the problems in insight/pojo have now been fixed. see ​https://github.com/openmicroscopy/openmicroscopy/pull/3890

The fact that the stroke width (same for fill color) was displayed in the web was due to the fact that the alpha component of the color is not handled in web.
Passing ticket to Will to fix that issue.

[jburel]

see also #12394

[jburel]

Will, did you have a chance to look at that one?

[wmoore]

I think alpha is handled in web unless it is zero. This is the same logic you have in Insight now, right?
So I don't think there's anything else that needs doing.

[jamoore]

Splitting 5.1.4 due to milestone decoupling