Re: gEDA-user: Gnetlist -g PCB

[Wojciech Kazubski <wk0@xxxxx>]

> 
> On Jun 29, 2008, at 4:49 AM, Wojciech Kazubski wrote:
> 
> > This is a bug in the library. The symbol pot-1.sym should have all  
> > pins declared as passive instead of input/output.
> 
> No, it's a conceptual problem with DRC. It is perfectly reasonable to  
> consider the pins on a pot as inputs or outputs. The classification  
> depends on the *application*, not on the *part*, so it is insane to  
> be classifying the pins on this kind of part.

For passive components any (as resistor or capacitor) any pin can be considered as "input" or "output" depending on circuit diagram, direction of current flow or so. Even for discrete semiconductor it is hard to say which pin is input and which is output. For example if a diode works as a rectifier, the above depend on required polarity of output voltage. The "passive" pintype is specified for such case and  means that the pin can be connected with anything else without error or warning.
Another variable resistor pot-bourns-1.sym does have all pins declared as passive. 

> 
> Pintypes are only sensible in a narrow class of flows: outside of  
> pure digital circuits using a single technology they don't make  
> sense. It is not surprising that they are confusing to newbies: there  
> is no way to make sense of them in general! Yet we grossly oversell  
> drc2, spamming every user of gnetlist with "Remember to check that  
> your schematic has no errors using the drc2 backend", an exhortation  
> that grossly exaggerates drc2's utility. It certainly doesn't "check  
> that the schematic has no errors": it mostly strains out gnats and  
> swallows camels.
>

Agree, DRC check works only for classic digital logic and does not cover all possible design errors. For example it does not check for two i/o pins being set as output in the same time. 
It does not work well for PLD devices or uC, where most of pins have to be specified as io.

> > There should be "pintype=pas" attribute for each pin instead of  
> > "pintype=io".
> > It is probably safe to ignore the warning.
> 
> The basic DRC should be much more limited. Right now, it's pretty  
> useless: it complains about a host of non problems. Its purpose  
> should be to detect problems in the translation from graphics to  
> netlists, not to try to outguess the engineering behind the graphics.  
> You can't do applied physics with clerical methods. It is not sane to  
> attempt to do so.
> 
...
> 
> John Doty              Noqsi Aerospace, Ltd.
> http://www.noqsi.com/
> jpd@xxxxxxxxx
> 
> 
> 
> 
> _______________________________________________
> geda-user mailing list
> geda-user@xxxxxxxxxxxxxx
> http://www.seul.org/cgi-bin/mailman/listinfo/geda-user



_______________________________________________
geda-user mailing list
geda-user@xxxxxxxxxxxxxx
http://www.seul.org/cgi-bin/mailman/listinfo/geda-user