Local label : Option for graphical goTo nubs (similar to those of hierarchal/global labels)


Unlike local labels, hierarchal labels, global labels, (and even netclass directives) make use of graphical goTo nubs, marking the specific associated wire and further differentiating the label from floating graphical text.

When using the labels to wirelessly connect pins (which can be common), the appearance conveys a wire which goes nowhere rather than a wire end which waypoints to other locations. Use of a hierarchal

Example: goTo Nubs


R801 : Netclass directive
R802 : Local label
R803 : Hierarchal label
R804 : Global label
R805 : Proposed alternative (Note: It’s not the correct colors…)

Example: Present Functionality vs Proposal

In the minimal examples below, use of local labels fails to intuitively signify a functional connection and, depending on the text justification, leaves the text:

Whereas, with goTo nubs:

  • The wire end intuitively signifies a functional connection
  • The text alignment is consistent vertically and horizontally:


I propose offering goTo nub shape options for local labels (with the inclusion of the present option for None); albeit, in the case of local labels the shape would suggestibly either share the color of the wire (green or blue for buses) or the color of the label text (yellow or blue for buses).

I included a rough example of this at the bottom of the goTo Nubs example image, above.


  • Similar discussion from 4 years ago found here.




Additional Example: Present Functionality vs Proposal

[Work in progress]

Further example :


Even just vertically centering net labels at an endpoint which is not a corner would make them much more consistent with other labels.

The vertical alignment-issue (which makes it hard to mix local&global labels on the same symbol/subsheet) is tracked in Eeschema: selectable anchor position for labels' text (lp:#1577324) (#2022) · Issues · KiCad / KiCad Source Code / kicad · GitLab.

Regarding the go-to-nubs I have mixed feelings. I recognize your arguments, but I’m not sure it will improve the schematic appearance. The current situation delivers a automatic distinction between all label types, which maybe get lost if local labels look similar to others.


Gave it +1 and added my input there; thank you.

This topic was automatically closed 90 days after the last reply. New replies are no longer allowed.