X-Authentication-Warning:	delorie.com: mail set sender to geda-user-bounces using -f
Date:	Tue, 13 Oct 2015 02:32:37 -0400
Message-Id:	<201510130632.t9D6WbG5008643@envy.delorie.com>
From:	DJ Delorie <dj AT delorie DOT com>
To:	geda-user AT delorie DOT com
In-reply-to:	<20151013075427.ebe071d65cbd1277d7f3c714@gmail.com>
	(geda-user AT delorie DOT com)
Subject:	Re: [geda-user] A lesson from gnet-makefile
References:	<1042003D-82E2-40F0-AB60-8186580C46AD AT noqsi DOT com>
	<20151012220005 DOT GC11989 AT visitor2 DOT iram DOT es> <20151013075427 DOT ebe071d65cbd1277d7f3c714 AT gmail DOT com>
Reply-To:	geda-user AT delorie DOT com
Errors-To:	nobody AT delorie DOT com
X-Mailing-List:	geda-user AT delorie DOT com
X-Unsubscribes-To:	listserv AT delorie DOT com

> I also agree about poor usefulness of auto placement. The firts
> problem is the decoupling capacitors. Maybe it would work better
> with autoplacement in several steps:
>   1. Disperse components.
>   2. Place decoupling capacitors.
>   3. As a first approximation place tightly coupled components as
>   groups which of curse may be very hard to define.
>   4. Autoplace this group of components there the tightly coupled
>   components may have to be fine tuned.

Autoplace is one of those things that's easy to experiment with using a plug-in.

It's also something that's very hard to "get right" algorithm-wise.

Heck, even a plugin that auto-selects bypass caps would be a
considerable challenge, but very useful.

I did one project where bypass caps had their own refdes (CB instead
of just C) and were numbered according to the pins they went with, but
that project only had one big IC on it anyway...

- Raw text -