OMEdit does not honour line thickness in connecting lines

[massimo ceraolo]

Modelica diagrams can be more effective if some connecting lines are twice the normal thickness. For this purpose one could use the is parameter “thickness”
For instance, when connecting expandable connectors Dymola uses thickness=0.5, that means 0.5mm, which, on usual dpi’s is 2 pixels, i.e. twice the other connecting lines.

When reading models created by Dymola, OM does not honour this thickness and leaves also the connector-connecting (yellow) lines one-pixel wide.

MY PROPOSALS:

1) I propose to modify it so that the line thickness parameter is used in determination of actual line thickness.

2) Moreover, I propose to adopt Dymola’s technique, i.e. to use thickness=0.5 when connecting expandable connectors, since it makes clearer diagrams.

3) Point 1) should be applied considering actual logical DPI’s. For instance with 168 DPI (a win system with 175% magnification) lineThickness=0.25 implies two-pixel lines. This will be useful for 4k display systems

4) Finally, I propose use lineThickness=0.5 also when making array connections, i.e. "connect constructs referencing arrays of components (having size >1)” as they are defined in definitions and examples of sect. 9.1 of Modelica specifications.

Points 1) and 3) above are actions needed to comply with specifications; the other two imply, IMO, a significant enhancement of the quality of our drawings.
In case of need, I can supply screenshots graphically depicting what I mean.

Note that when I say that thickness=0.5 means 0.5 mm I refer to Modelica specifications 18.6.1 and 18.6.5.1. I assume that mm’s used in those specs refer to the logical DPI: 25mm=1in=96 pixels at 96DPI, so 0.25mm is one pixel at 96 DPI

[massimo ceraolo]

Just to show what I mean, I enclose two png's showing the difference. They are created on a 4k screen.
If the ticket is accepted, lines will probably look-like those in the "thickLines.png" file. Tosay they are as per thinLines.png.

Moreover, thick lines would be generated also in case of array connections, not shown in the png's.

[Adeel Asghar]

I have fixed everything you have requested in d3844f3/OMEdit.

[massimo ceraolo]

I've checked today.
The result is amazingly good!
Thanks.

[massimo ceraolo]

Now also Modelica issue 2188 has been fixed, so we will soon see correct thicknesses also in models affected by comment 4 (they are several).

[Adeel Asghar]

Great!