From: theom+news@chiark.greenend.org.uk   
      
   Don Y  wrote:   
   > I tend to like assurances that the right connector is   
   > in the right place.  And, I think users -- when faced with   
   > multiple connectors -- like some reassurance that they've   
   > mated them correctly.   
   >   
   > But, if I can deduce the role of the actual connection from   
   > the data observed, is there a downside to NOT keying the   
   > connector(s)?  (aside from the fact that it requires LOOKING   
   > at the signals before you can make that assertion -- if you   
   > haven't looked, do you really *care* about Schrödinger's cat   
      
   Do you mean orientation, or location?   
      
   ie if you have 3 USB-A ports they do care about orientation, but (perhaps)   
   it doesn't matter which one you plug into.   
      
   The main thing I can think of is user confusion.  If the user is expecting a   
   1:1 mapping of cables to sockets, they may be uncertain when presented with   
   3 identical sockets and need to check documentation/techsupport about what's   
   allowed.  Whereas if your connectors are square / triangle / circle shaped   
   they instinctively know the round peg goes in the round hole.  If you give   
   them three round connectors and three round holes, they may freeze because   
   they're worried about plugging into the wrong hole may blow something up.   
      
   For things like US mains receptacles and USB-C sockets, they're familiar   
   enough with the idea that any orientation works and plugging into the wrong   
   socket isn't going to blow anything up, but if you have a proprietary   
   solution then they may not have that assurance.   
      
   Theo   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)