kicad-developers team mailing list archive

[Strontium <strntydog@xxxxxxxxx>]

His third criticism is with net connectivity.

What he did was he has a component with 4 GND pads in a row.  He ran a 
track across them (in the middle) and finished it at the last pad.  The 
Track clearly crosses all pads, however the DRC shows them as 
"unconnected" when they are really connected electrically.

His complaint is that this is tedious and unintuitive to lay, and wrong 
from a physical model of the board perspective.  That any copper track 
crossing a pad at any point should be considered connected, because it 
is, it shouldn't need to touch the centre of the pad, which is an 
arbitrary anchor.

Its also not consistent with a track that crosses a pad that is not its 
net.  That track, if it touches the wrong pad at any point throws a DRC 
error, but if it connects to a pad with the correct net (but doesn't end 
on the pads centre) is considered not connected.

There is an edge case where a track only slightly touches a pad, and is 
electrically connected, but Not connected sufficiently for the design, 
but that is not the same thing as being "unconnected" and isn't handled 
by DRC now in any event.

He IS a new user and he naturally has numerous complaints, most of which 
result from frustration with learning a new tool, and I talk him through 
those.  But, I think these three points are valid criticisms that the 
developers should consider in the future.  I will make some entries in 
the bug tracking system to track these, but I wanted feedback first, are 
these "wont fix" or "feature requests", "bugs", etc.

Steven