Re: gEDA: Has anyone done a symbol for an Atmel AVR?

[Ales Hvezda <ahvezda-at-NOSPAMseul_dot_org>]

Hi,

[snip]
>I do have a sort of "style" question.  The 90S8515 (and the 8051 for
...
[snip]
...
>When creating the symbol, would it be better to have both functions
>of a pin listed, or only one, and if only one, which one.  For example
>I started modifying the 8031 symbol, and I decided to list both so I
>had something that looked sort of like


	If you look at the current symbol library, you will find both
styles.  Sometimes you find these two different styles on parts from
the same family.  I don't think I could enforce any sort of style like
this (without causing people to give up on contributing their symbols),
so I generally leave them all along (with the exception of some minor
cosmetic things).  I don't really have much of a strong opinion on this,
whatever makes the schematic more readable to you.


>
>etc.  Which is  fine, except that you need to widen the symbol some
>to get all of the special purpose pin names in there.  Also, there
>is the question about what to do when the the pin can be either a
>port pin or an inverting input or output.  In that case should you
>include a bubble or not?  For example, there is PD6 which is also
>/WR.


	Yeah, that's a tough call.  My initial reaction would be to not
put a bubble if the pin can be an input or output.


>
>I suppose this might be an argument for having multiple symbols
>for the same device, but it seems you could go crazy and create
>too many devices.


	Some would argue that the current symbol scheme is already
crazy.  For the most part I agree with them, but until something better
comes along, I guess we are stuck with it. :)  


>Is there a general policy about this?  I didn't find it in


	Nope.  Should there be one? 


>a quick read of the "Component Symbol Guide".  My own preference
>would be to have both functions listed for most of the pins,
>mainly for debugging purposes.  If you see a signal acting
>strangely, and you look at the schematic and see that the
>special function would make it act that way, it might save
>a bit of time.


	Your call is as good as mine (probably better). :-)

								-Ales