X-Authentication-Warning: delorie.com: mail set sender to geda-user-bounces using -f X-Recipient: geda-user AT delorie DOT com Message-ID: <1444157156.1949.52.camel@ssalewski.de> Subject: Re: [geda-user] GTK3, Glade interface designer (router, auto?) From: Stefan Salewski To: geda-user AT delorie DOT com Date: Tue, 06 Oct 2015 20:45:56 +0200 In-Reply-To: References: <20151003210701 DOT de392b925f54dadb0a5fedd8 AT gmail DOT com> <1443903758 DOT 1873 DOT 13 DOT camel AT ssalewski DOT de> <56104A0A DOT 9020507 AT xs4all DOT nl> <1443909591 DOT 1873 DOT 18 DOT camel AT ssalewski DOT de> <1443975731 DOT 671 DOT 52 DOT camel AT ssalewski DOT de> <20151004191717 DOT bf8223417541a9306bfbd9ea AT gmail DOT com> <1443997480 DOT 2068 DOT 32 DOT camel AT ssalewski DOT de> <1444070851 DOT 1014 DOT 20 DOT camel AT ssalewski DOT de> Content-Type: text/plain; charset="UTF-8" X-Mailer: Evolution 3.16.5 Mime-Version: 1.0 Content-Transfer-Encoding: 7bit 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 Precedence: bulk On Tue, 2015-10-06 at 02:34 +0200, Kai-Martin Knaak wrote: > Stefan Salewski wrote: > > > The real hard part is, that people really do not like autorouters > > generally, > > I don't like auto routers which fail on my boards ;-) > My biggest concern with what the toporouter could demonstrably do is > that > it is/was a game of all or nothing. You could not tell the router to > restrict itself to some routes picked by me. And more troublesome, it > was > not able to play nice with tracks and vias already put in place by me. > Both are hard deal breakers for me. Of course the router should not fail, and you should be able to manually continue. But the toporouter idea is generally starting with a board with no traces, and doing a complete route. You can tell it to use only a subset of nets, but that is generally not helpful. Read the thesis, maybe you have ideas how to improve that. You may think about the connections to ground and power planes, that point is not really covered in the thesis. The thesis is more concerned with chip design. Should be not hard, but we have to do thinking. And you may think about automatically moving elements, for example for shrinking PCB to minimal size. For that point papers may exist. Or another point, which I just remember: Unreachable areas on the board: The layer assignment algorithm does not take that into account when placing vias, because it does not really occur in chip design. But on PCB, there may exists barriers, for example elements may have outer rows of pads, so that putting vias in the inner area for routing side exchange makes not really sense... > > and most really do not like curved traces. > > Anthony already wrote that most people told him that. > > Well, I remember quite some "awsome!" shouts on the list every time > the > toporouter produced some new results. Yes, I can remember that. I think the more negative comments he got by private email. You can read German text, so maybe you have looked from time to time into the mikrocontroller.net forum? I have to admit that I do. Of course the people there are not really smart (with very few exceptions). But there it is common sense that they generally hate autorouters. And they hate curved traces, because it reminds to old handdrawn boards. That people care for PCB board colors and gold platings. And after some thinking, I can somewhat understand autorouter haters. There is the Ikea effect: You want to have build it really yourself. And for professionals: Autorouters may kill your job. Push and shove is more like an action game, people may enjoy it. It is similar like coding in Assembler 30 years ago, that time some people hated C. Here in Germany many people hates automatic gear shifting in their cars! I am sure in twenty years no one will still route manually. But I do not really think they will use our router then?