1.0.6 Multiple traces being treated as one

[poetaster]

Current Behaviour

As discussed here: https://forum.fritzing.org/t/1-0-6-issues-with-traces-being-treated-as-one/33493/3
connect a trace from one resistor to an other resistor,
connect a trace from another resistor to the first
attempt to take trace one to the back
both traces went to the back. ???!!! Selecting one trace, selects both!!!

A fritzing file and screen shot are included in the discussion linked above.

Build:
Version 1.0.6
(1.0.5rc1-97-g04e5bb02 2025-10-07) 64 [Qt 6.5.3]

Operating System:
Linux, ubuntu 24

Steps to reproduce:
connect 4 elements ( header -> resistor -> resistor -> header) in series, observe the traces behave as if they were one trace.

Expected Behaviour

Each trace remains independent.

Sometimes after deleting both traces and beginning from scratch, you can get trace one places, directly send it to the back and create trace 2 on the other layer. Sometimes.

[poetaster]

The traces depicted here could not be selected individually.

[KjellMorgenstern]

One workaround is to use "Delete up to bendpoints". This way, the other sections can be modified separately, and only the deleted section needs to be redone.

[paulfer747]

I reported this bug many months ago and I got a reply that the bug could not be replicated. This is a HIGHLY frustrating issue. I sent screenshots and videos. And the issue was simply closed.

It cost me having to re-order boards because I moved one of the resistors wire trace width, and then it simply changed the width for the entire length of the other end of the trace appearing to be unattached by a micron. Even though I checked my boards beforehand, this one was missed.

Having to say "delete wire up to bendpoints" EVERY time is so frustrating and time consuming. For me, this behaviour cropped in, in the latest update.

It appears that wire traces now attach to other wires, and not to the parts anymore.

[KjellMorgenstern]

@paulfer747 I tried really hard to understand #4261. You are saying you posted videos about it, but the only ones I can see are the two videos that I had created. Or are you referring to a different issue?

[KjellMorgenstern]

Wire connects to wire instead of pad:

4290_1.mp4

This also happens with vias or other connectors. With the large pads, it is possible to target the pad or the wire, but during normal operations, or smaller pads, the outcome is random.

Wire to wire connections are not intended here, and cause further side effects. For example, when moving the pad, the wire doesn't follow. It is drawn correctly on the next update, e.g. reloading or modifcations near the wire. But until then it looks separated:

4290_2.mp4

[KjellMorgenstern]

Ground fill does not trigger the wire to jump back after the pad was moved:
So this broken ground fill is created:

Moving Pad1 redraws the wire correctly, but it might be to late:

[KjellMorgenstern]

Workarounds are just suggestions until the underlying issue is fixed:

• Use "delete up to bendpoints" to delete and redraw the affected section.
• When connecting a wire, try to find a spot where the connector (part, pad, via, ...) is highlighted in blue. not other wires

Good connection:

Bug 4290 or 4261:

[paulfer747]

Yes thanks Kjell, thats what I have been doing. Looking fwd to the fix thanks!

[KjellMorgenstern]

Issue also affects Fritzing 1.0.4. and 1.0.5

[poetaster]

Issue also affects Fritzing 1.0.4. and 1.0.5

I never noticed this behaviour in 1.0.4. I'm a bit on the cautious side since I'm in production (ie. costs me money) so I was thinking of stepping back to 1.0.4 which I've produced a number of designs with. I don't think I need to go as far back as 1.0.3 ?

[KjellMorgenstern]

@poetaster This issue happens during edits, and effects are immediately visible: An edit affects an adjacent wire. I don't think it helps to go to back to a previous version. Determining the version it was introduced just helped when looking for a fix.

The fix wasn't difficult:
Version 1.0.7 will only connect wires if you have two loose ends without connectors like vias or pads overlapping. Otherwise the connector will be chosen.

[KjellMorgenstern]

PS: We test that old sketches load fine in newer versions, also compare gerber files for differences.

We don't do that in the other direction, Fritzing might even warn when loading a file from a more recent version.

[poetaster]

PS: We test that old sketches load fine in newer versions, also compare gerber files for differences.

We don't do that in the other direction, Fritzing might even warn when loading a file from a more recent version.

Ok, I'll take one of the two projects to production and just be extra careful when looking at the gerbers.