"Single pad nets delete" option of Read netlist is unnecessary
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
KiCad |
Fix Released
|
Low
|
Jeff Young |
Bug Description
I can't see the reason for "Single pad nets delete" option of Read netlist.
Eeschema always includes single pad nets (unconnected pins, even if they are flagged not-connected in schematic) in netlist.
If this netlist is read into pcbnew, these single-pin nets will be loaded.
When Reading netlist next time, nothing happens if Single pad nets option is set to Keep.
On the other hand, if this option is set to Delete, every run of Read netlist
generates pairs of
Changing component "D2:/54AFB4C9" pin "3" net name from "" to "Net-(D2-Pad3)".
Remove single pad net "Net-(D2-Pad3)" on "D2" pad '3'
messages for each of unconnected pins.
If you don't want to be flood by messages and filter out Infos and Warnings, these messages keep displaying as they are type of Action. Actions and Errors disabled is really bad idea.
My wish is opting out "Single pad nets delete" option from Read netlist (it should always work like set to "Keep"). Rather may have an option in eeschema Generate netlist, "Eliminate single pad nets from netlist"
Changed in kicad: | |
status: | New → Opinion |
Changed in kicad: | |
status: | Triaged → In Progress |
milestone: | none → 5.1.0 |
Changed in kicad: | |
status: | In Progress → Fix Committed |
Changed in kicad: | |
status: | Fix Committed → Fix Released |
There are 2 items here -
1) Novak wrote " Rather may have an option in eeschema Generate netlist, "Eliminate single pad nets from netlist"
That wishlist item I agree with, it would be a good idea to have the choice to not generate single-pin-net info, coming out of Sch.
2) Novak wrote "My wish is opting out "Single pad nets delete" option from Read netlist (it should always work like set to "Keep"). "
That one is trickier, as NET import does not _always_ come from EESCHEMA, and operations like Append, may _require_ single pin nets to be present, to correctly join another reference.
I think the _choice_ to remove single pin on import is quite useful, but your point about spurious
messages is valid - perhaps a [Quiet] tick choice, next to the Single-Pin choice ?
Related to all this, I think I saw a design with some no-net PADS, that got a little confused when I generated a Pour outline with Net Name of None or ""
IIRC PcbNew seems to think a Pad with "" and a Fill with "" are the same net.
That detail may need checking, before wholesale strip of Single-Pin Names is undertaken.