kicad-developers team mailing list archive

[Chris Pavlina <pavlina.chris@xxxxxxxxx>]

Ah, but what of the possibility to make more than two connections on a 
wire? You could connect both endpoints and still put a label on the 
wire!

Interesting to think about, but it seems to me that if you go for 
logical connections instead of purely graphical ones, you're opening up 
a huge can of complex worms. Suppose I have a wire that is disconnected 
at both ends and then I place a label in the middle. Which target should 
disappear? Suppose I have a wire that goes to a pin at one end, and I 
place a label in the middle. The target at the other end disappears, but 
now I have nothing indicating whether the /third/ connection that I 
intended to make (which the connection algorithm didn't know about) was 
successful.

The targets indicating connection only for that exact graphical point is 
the way most software packages would handle this, and I think the way 
most users will expect.


On Tue, Jun 09, 2015 at 07:44:37PM -0400, Wayne Stambaugh wrote:
> On 6/9/2015 4:45 PM, Chris Pavlina wrote:
> > Okay, I see what you're talking about in the video demo. Here's a quick 
> > screenshot for those who want to follow along:
> > 
> > http://misc.c4757p.com/video-demo-block.png
> > 
> > Simply put, those wires ARE being displayed correctly. The *ends* *are* 
> > dangling. Nothing is connected to them. There is no possible way I could 
> > write a test that detects that a label is on the wire and somehow 
> > decides which end to associate it with, hiding the target.
> 
> This depends on your definition of dangling end.  If you mean there is
> no graphical item connection, then wires shown are dangling.  However,
> if by dangling end you mean no electrical connection (net), then these
> wires are not dangling.  They are connected by net to another
> wire/label/pin combination which yields a valid net.  In this case the
> test code is the same code that generates the net list.  It's fine for
> now and I'm not looking for you to make this change.  I just wanted to
> get you thinking about what connection really means in a schematic.
> 
> > 
> > The point of hiding the targets is to show that *that* connection point 
> > is correctly in use. These are not in use, so why hide the target? The 
> > equivalent here is the target on the label itself, which has always 
> > hidden when connected.
> > 
> > If you're going to use labels to make connections, and want the targets 
> > to disappear, then you'll have to place the label target on the wire 
> > target. This is the same as for pins, if you stick a pin in the middle 
> > of a wire and connect it with a junction, the *end* of the wire is still 
> > not connected.
> > 
> > I think this is correct behavior.
> > 
> > On Tue, Jun 09, 2015 at 03:33:56PM -0400, Wayne Stambaugh wrote:
> >> How does this bus connect to another bus without a label that looks
> >> something like A[19..32] on other open ended buses?  Unless the bus is
> >> directly connected I believe the label is required.
> >>
> >> FYI.  I need to get my eyes checked.  If a wire doesn't end on the label
> >> connection point, then open ended wires are shown as unconnected.  It
> >> looks the dangling end test is in need of some improvement.  Again, see
> >> the video demo.  There are examples of both cases.
> >>
> >> On 6/9/2015 3:24 PM, Chris Pavlina wrote:
> >>> No, I just mean things like this:
> >>>
> >>> http://misc.c4757p.com/bus.png
> >>>
> >>> The bus is very frequently drawn past the end of a row of connections to 
> >>> emphasize the fact that it is a bus.
> >>>
> >>> (Imagine the bit going off to the right goes somewhere else.)
> >>>
> >>> On Tue, Jun 09, 2015 at 03:17:52PM -0400, Wayne Stambaugh wrote:
> >>>> Yes, but typically in open ended bus use case they have a label attached
> >>>> to them that makes the connection.  The same thing can be done with
> >>>> wires which does show the connection status correctly when a wire is
> >>>> open ended but has a label attached to it.  The video demo project is a
> >>>> good example of both behaviors.  It's not that important but some users
> >>>> may wonder why the difference between wire and bus connections.  It's
> >>>> something to think about.
> >>>>
> >>>> On 6/9/2015 2:30 PM, Chris Pavlina wrote:
> >>>>> I'm not sure about that one - it's common to draw buses in a manner that 
> >>>>> leaves their ends open.
> >>>>>
> >>>>> On Tue, Jun 09, 2015 at 01:43:56PM -0400, Wayne Stambaugh wrote:
> >>>>>> I've tried this patch and it works as expected.  Good work Chris.  Does
> >>>>>> anyone have any objections to this patch before I commit it?
> >>>>>>
> >>>>>> One other thing I noticed is the bus wires (and I'm guessing bus entries
> >>>>>> as well) do not have any segment end point connection indicators.  It
> >>>>>> would be nice if they supported them as well.
> >>>>>>
> >>>>>> On 6/8/2015 5:38 PM, Chris Pavlina wrote:
> >>>>>>> Hi,
> >>>>>>>
> >>>>>>> I already sent in this patch but as a reply to my existing pin targets 
> >>>>>>> thread; I think it may have got lost. This patch enables pin "targets" 
> >>>>>>> on bus entries as well, as Wayne suggested, and hides them when the bus 
> >>>>>>> entry is connected.
> >>>>>>>
> >>>>>>> --
> >>>>>>> Chris
> >>>>>>>
> >>>>>>>
> >>>>>>>
> >>>>>>> _______________________________________________
> >>>>>>> Mailing list: https://launchpad.net/~kicad-developers
> >>>>>>> Post to     : kicad-developers@xxxxxxxxxxxxxxxxxxx
> >>>>>>> Unsubscribe : https://launchpad.net/~kicad-developers
> >>>>>>> More help   : https://help.launchpad.net/ListHelp
> >>>>>>>
> >>>>>>
> >>>>>> _______________________________________________
> >>>>>> Mailing list: https://launchpad.net/~kicad-developers
> >>>>>> Post to     : kicad-developers@xxxxxxxxxxxxxxxxxxx
> >>>>>> Unsubscribe : https://launchpad.net/~kicad-developers
> >>>>>> More help   : https://help.launchpad.net/ListHelp
> >>>>>
> >>>>> _______________________________________________
> >>>>> Mailing list: https://launchpad.net/~kicad-developers
> >>>>> Post to     : kicad-developers@xxxxxxxxxxxxxxxxxxx
> >>>>> Unsubscribe : https://launchpad.net/~kicad-developers
> >>>>> More help   : https://help.launchpad.net/ListHelp
> >>>>>
> >>>>
> >>>> _______________________________________________
> >>>> Mailing list: https://launchpad.net/~kicad-developers
> >>>> Post to     : kicad-developers@xxxxxxxxxxxxxxxxxxx
> >>>> Unsubscribe : https://launchpad.net/~kicad-developers
> >>>> More help   : https://help.launchpad.net/ListHelp
> >>>
> >>> _______________________________________________
> >>> Mailing list: https://launchpad.net/~kicad-developers
> >>> Post to     : kicad-developers@xxxxxxxxxxxxxxxxxxx
> >>> Unsubscribe : https://launchpad.net/~kicad-developers
> >>> More help   : https://help.launchpad.net/ListHelp
> >>>
> >>
> >> _______________________________________________
> >> Mailing list: https://launchpad.net/~kicad-developers
> >> Post to     : kicad-developers@xxxxxxxxxxxxxxxxxxx
> >> Unsubscribe : https://launchpad.net/~kicad-developers
> >> More help   : https://help.launchpad.net/ListHelp
> > 
> > _______________________________________________
> > Mailing list: https://launchpad.net/~kicad-developers
> > Post to     : kicad-developers@xxxxxxxxxxxxxxxxxxx
> > Unsubscribe : https://launchpad.net/~kicad-developers
> > More help   : https://help.launchpad.net/ListHelp
> > 
> 
> 
> _______________________________________________
> Mailing list: https://launchpad.net/~kicad-developers
> Post to     : kicad-developers@xxxxxxxxxxxxxxxxxxx
> Unsubscribe : https://launchpad.net/~kicad-developers
> More help   : https://help.launchpad.net/ListHelp