Rather use gitlab: https://gitlab.com/kicad/code/kicad/issues/2558
There seems to be some acceptance, but not for your most straightforward solution.
1 Like
Maybe it was because I used an autorouter.
Worked fine, saved me a lot of work. 
One of the advantages of being naive. Picked the obvious switch, picked an appropriate footprint, and it all worked.
Actually now that I think of it, both pins of the same number were joined by a rats nest line so this would have come to the attention of a human router.
KiCad cannot be changed the way you suggest, because it would break a lot of existing use cases.
For example, what happens if you have an IC with multiple GND pins that need to be connected? You want DRC to flag an error if one or pins are not connected, and you don’t want to use the IC as a jumper.
Pretty sure we went over this stuff in another thread, unfortunately people tend to have a singular view of their own use case without considering related issues.
1 Like
Maybe you didn’t read carefully. Piotr suggested interpreting the pads with the same number being internally connected. That wouldn’t affect any existing IC footprint (unless made in a nonstandard way, violating datasheet).
Sorry that is nonsense. Not everyone uses the official libraries or their conventions. You can’t tell users their existing footprints are wrong “because they didn’t make a footprint in (your newly defined) standard way”. Ridiculous!
Existing behaviour is that KiCad expects all pads with the same number to require external connection. You are proposing to break that behaviour.
Unfortunately, if you don’t understand the implications of a proposed change, it will never be adopted by KiCad, and you will be forever wondering why.
Wasting my time again, giving up.
The 
at the end of my sentence (you bothered to cut out) was to signal that that sentence is not to be understand exactly as was said.
Ok, for the benefit of non-mindreaders can you in future write what you mean. I cut the smily because it was not important or relevant to what you wrote.
I think footprints with few pads having the same number and really requiring that pads to be connected externally are extremely rare, but questions about 4 pin tact switches are frequent.
At that moment I can’t imagine any such footprint but of course my imagination is limited.
Do you have any example?
No I’m not. As you can see in the gitlab bug report. I’m not proposing that KiCad should behave that way.
What is needed/wanted is a way to tell KiCad “these pins are connected outside the board; connecting them in the board is allowed but not necessary”. If you have something constructive to say, you can continue in the given bug report.
I think I have written exactly what I mean:
- There is no acceptation for my concept (it is simply true)
- I have no problem with it (I on it).
Even if I would use 
it wouldn’t mean that everyone except me are stupid but that I had so genius concept which felt down when confronted with reality.
I really dislike settings doing double duty. The pin and pad numbers already define which pad is connected to which pin. Adding another functionality (encoding of internally connected) on top of that functionality just screams dirty hack (think about how stupid the double duty of the hidden power pins is).
So in my opinion the “internally connected” info should be provided by some separate information (i would prefer on the symbol level as this is more flexible.)
1 Like
Really? I would say schematic level is more flexible because it can be used only when needed without modifying the symbol itself, so the symbol is fitting for more purposes.
Fully specified symbols 
(I am not against also having the option to define it at the schematic level but really this is a property of the part so it must be set able in the library)
Thanks a lot for your replies. They helped see where I was wrong. When I browsed the footprints, I did not noticed that for some of those 4 pin switches, that they had 2xpin1 and 2xpin2.
With the datasheet in mind my brain printed 1, 2, 3, 4. So I was looking for a four pins symbol.
Maybe I’m not the first one having this " misunderstanding" but my short term issue is solved now.
Thanks for the help !
JMF
1 Like
As I understand you, functionality you describe is reached by having separate number per symbol pin and per footprint pad (may be I don’t follow you right). Using the same pad number to mark internal connections seemed for me the simplest method to be used, and I don’t understand it as encoding the next functionality with the same.
I don’t understand the problems there were discussed in my bug report because I know too little about KiCad internal working. I’m not sure if using another method of marking internal connection helps to solve any of those problems. If it helps I am certainly for such solution and not using the same pad number for it, but if not I just think the simplest and easiest to understand for anyone is that if symbol pin has the same number then several footprint pads then that pin is just connected to all of them.
This topic was automatically closed 90 days after the last reply. New replies are no longer allowed.