Kaimartin - 
Placement tools use pin 1 as a part origin (0,0), and some use the package 
center as a special location, primarily for part pickup handler? I think the 
pin 1 end of the wire used as origin would give a reliable point for schematic 
drawing placement use, as then adding pins, etc in the symbol would not move 
the part or disconnect the part Nor would adding properties / text, etc in the 
symbol. Since part disconnects can be invisible to the naked eye, this can be a 
very dangerous critter to play with. Imagine placing a symbol, and then fixing 
a typo by adding a small l in some small text at the symbol drawing bottom, 
causing the whole symbol to shift in the schematic by a little bit. Not good. 
Having a method that obviously just works, like the pin 1 wire end as symbol 
origin, to avoid this kind of breakage, gives us the time to look out for all 
the other errors that creep into our schematics. Yes, this does require 
negative coordinates in the symbol and sch. 

Ross Amans
Hardware Design Engineer
Biometric Access Co
512-426-9252 (cell)

-----Original Message-----
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Kai-Martin Knaak
Sent: Tuesday, May 27, 2008 7:07 AM
To: [EMAIL PROTECTED]
Subject: gEDA-dev: origin in gschem

Hi, 
I just got hurt by a relocated symbol again. The whole business of symbol 
translation is a pain and a likely source of error. Why not introduce the 
notion of an explicit origin in the definition of symbols? You put the 
origin at some outstanding point of the symbol and keep it there. This 
would allow to put the symbol somewhere in the middle of the canvas. No 
need to relocate the symbol just because of some additional graphics at 
the bottom. No more broken schematics because of such cosmetic shifts.

At first glance, an origin seems to imply additional offset calculation 
for each and every coordinate. In reality this has to be done only once 
when a symbol is loaded. The coords relative to the origin of the symbol 
would be cached. The rest of the code could remain unchanged. 

A different approach would be to allow negative coords in *.sch format. 
In that case the origin remains always remains at (0,0) but the viewport 
might be shifted to extend into the negative.

Any thoughts? 

---<(kaimartin)>---
-- 
Kai-Martin Knaak                                  tel: +49-511-762-2895
Universität Hannover, Inst. für Quantenoptik      fax: +49-511-762-2211 
Welfengarten 1, 30167 Hannover           http://www.iqo.uni-hannover.de
GPG key:    http://pgp.mit.edu:11371/pks/lookup?search=Knaak+kmk&op=get



_______________________________________________
geda-dev mailing list
[email protected]
http://www.seul.org/cgi-bin/mailman/listinfo/geda-dev


_______________________________________________
geda-dev mailing list
[email protected]
http://www.seul.org/cgi-bin/mailman/listinfo/geda-dev

Reply via email to