Is it specific to “label not connected to anything” errors? Also, does it matter what type of label it is?
I believe the check for local labels actually being connected is somewhat new, so it would not shock me to find out there was something strange going on.
As a basic check, I can’t replicate this on Mac with a wire connected to nothing and a unique local label.
EDIT: I am a fool, with a label actually connected to nothing, I do replicate this with all label types on Mac.
Yes. Also for v6.99. Affects all label-types (local label + global label + hierarchical label) which are connected to really nothing (even no wire-stubs).
Ah, yeah, did not mention that the label was connected to nothing. But then I saw a issue report where people are discussing things like that. Maybe a fix is coming then.
Maybe the system is looking for both ends? If so, I’d say the fix is worse than the problem as it could rely on some coding gymnastics to get around. I remember when I was learning to code you fixed one error at a time because fixing one could take care of lots of others.
I think I confused things slightly here. Does this give context? In my opinion each of these cases should give one error, but the case where the label is truly totally alone gives two errors. (KiCad 6.0.6 on Mac)
And we agree with you @scone, but they are already “Errors”. See it on your screenshot. Now you should see 2 items there, and not 3, since one of those is duplicated.
No. I understand. I think the software is looking for both ends. In one case one end is connected. In one case both ends are disconnected. Two ends are implied with labels. This is no small distinction in terms of the check. “Fixing” the duplication runs the risk of breaking the check.
I hope that makes sense.
EDIT: This is just my best guess on why it happens. The decision on how well/easily this can be remedied is up to the developers of course.