X-Authentication-Warning: delorie.com: mail set sender to geda-user-bounces using -f X-Recipient: geda-user AT delorie DOT com Message-ID: <50CF6157.9030405@laserlinc.com> Date: Mon, 17 Dec 2012 13:15:51 -0500 From: Joshua Lansford User-Agent: Mozilla/5.0 (X11; Linux i686; rv:15.0) Gecko/20120907 Thunderbird/15.0.1 MIME-Version: 1.0 To: geda-user AT delorie DOT com Subject: Re: [geda-user] Find rat lines - summary References: <20121204183305 DOT 6b04c0dc AT jive DOT levalinux DOT org> <20121208112649 DOT 388a9d22 AT jive DOT levalinux DOT org> <1355011808 DOT 19390 DOT 8 DOT camel AT localhost> <1355585479 DOT 7067 DOT 26 DOT camel AT monster> <1355587723 DOT 7067 DOT 40 DOT camel AT monster> In-Reply-To: Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Reply-To: geda-user AT delorie DOT com > To me, currently it seems the best approach would be to use the > combination _and_ one of the minimal cut ideas in parallel. This would > allow the user to use tagging as much as (s)he finds reasonable, > between no tagging at all and tediously tagging every piece of copper > ever existed in the history of the board. Whenever tagging can reveal > a short, PCB should use that info and fall back on the minimal cut > otherwise. > +1. I was just about to suggest this myself. Everyplace that has been manually tagged can be added to the graph as a source as if there was a component pin with a net attribute at that point. That way you can retroactively add tags to chase down a short being guided by the minimum cut or other heuristic, or add the tags as you go along if it fits your development cycle. In the retroactive method, the heuristic would probably provide better and better guesses as you supplied it with more and more information by adding tags. ~Joshua