[mkgmap-dev] Fw: Re: Curvy routing support: new function?

Thu Jul 3 21:48:15 BST 2014

Hi Felix,

as far as I understand the Garmin algo, these penalties only take effect if you turn from one street into another (sharp edge > some deg, Garmin says "turn"). For straightly connected roads or road pieces, this does not appear (AFAIK)

In my bicycle maps, therefore I use only road_speeds of 2 or less so that turning penalties show little effect. 

If used with car routing, curviness should only be used for lowering road_speed, e.g. on curvy roads without speed limit. With motorcycle maps, I'd like to increase road_speed for curvy road. Yes, penalties show more effect then. Then it is a question, how often you make use of the function. Of course curviness should not be used that much so that every light curvy road gets a road_speed of 7. It should be used that only very curvy roads get a bonus, and thus preferred better than with standard settings. As long as the other non-curvy or little-curvy roads get no benefit, the curvy ones will be preferred.
It's likely the same with road_class. Only ca. 5% of roads should be road_class 4, so a reduced main routing network exists.

Cheers
Manfred

> Gesendet: Donnerstag, 03. Juli 2014 um 20:39 Uhr
> Von: "Felix Hartmann" <extremecarver at gmail.com>
> An: "Development list for mkgmap" <mkgmap-dev at lists.mkgmap.org.uk>
> Betreff: Re: [mkgmap-dev] Fw: Re:  Curvy routing support: new function?
>
> Oh - actually there is one thing that could be done - but I'm not sure 
> if it is possible and even if might lead to big fuckups..
> 
> That is - display the road as is, but manipulate the counters so it 
> appears shorter on calculations. Maybe this is possible. I mean there 
> must be a way to make it longer - to account for altitude - so is it 
> possible to shorten the road but display it longer?
> That would of course give you wrong kilometers and distances on every 
> possible setting...
> On 03.07.2014 20:32, Felix Hartmann wrote:
> > You are missing a big big point here!
> >
> > The higher you put the speed, the bigger the time penalties for sharp 
> > turns.
> > So the only thing reasonable to do is ramp up the road_class, and keep 
> > or back down the road-speed..
> >
> > But best give it some tries yourself by creating a sample map where 
> > you increase speed for some curvy roads based on name filters. You 
> > will likely notice that the result is rather the opposite. Especially 
> > if there are crossings...
> >
> >
> > Felix
> > On 03.07.2014 19:49, Manfred Brenneisen wrote:
> >> Hi all,
> >>
> >>> Gesendet: Mittwoch, 02. Juli 2014 um 08:46 Uhr
> >>> Von: "Felix Hartmann" <extremecarver at gmail.com>
> >>> An: mkgmap-dev at lists.mkgmap.org.uk
> >>> Betreff: Re: [mkgmap-dev] Curvy routing support: new function?
> >>>
> >>> Well I have thought about this too - also new Basecamp has a curvy 
> >>> roads
> >>> preference - but what should mkgmap do about it?
> >>>
> >>> You can calculate if a road is curvy or not, but what then? The only
> >>> thing you can do until mkgmap knows NT format maps which I think has a
> >>> special "tag" for this, is to increase the road-class.
> >>> Increasing the road-speed would be against your intentions - as the
> >>> higher the road-speed - the more garmin algo will punish you for curvy
> >>> roads...
> >>>
> >>> So until we know how Garmin maps tell Basecamp/the GPS Unit to 
> >>> exclude a
> >>> road from curvyness punishment - the only thing you can do is to
> >>> increase the road-class...
> >>> I'm 100% sure the prefer curvy roads switch, works like avoidance -
> >>> looking for the curvy road indicator inside the map. There is 
> >>> absolutely
> >>> no change in routing in Basecamp if I set it or not with non NT maps...
> >>
> >>
> >>
> >> I devised myself a day to think about it. To be not too enthusiastic 
> >> or conservative, as the future will be NT maps as you state 
> >> correctly. Finally, I think:
> >> mkgmap could translate the physical shape of a road into different 
> >> categories, e.g. by setting mkgmap:road-speed parameters. This is 
> >> more analog than digital "curvy" or "not curvy" setting by just 
> >> setting single bits (which are not widespread yet, NT only).
> >> One might want to differentiate between this way: 
> >> http://www.openstreetmap.org/way/186500650 and that way: 
> >> http://www.openstreetmap.org/way/106597772
> >> Which one is faster? Would you like to add the (highly unwanted!) 
> >> maxspeed:practical tag to OSM data?
> >> Car map designers might want to lower speeds for curvy roads, 
> >> motorcycle map makers would go the other way (knowing that the 
> >> calculated travel time is incorrect)
> >> Adding style rules like
> >> highway=primary & curviness() > 50 {add mkgmap:road-speed-max=5}
> >> would not influence the map, only curvy roads are affected.
> >>
> >> Nobody is forced to use the new function (knowing that it slows down 
> >> map generation if used). But as it is not a really big thing you 
> >> might give it a try. I'll test, and report when it works correctly 
> >> (for me :-))
> >>
> >>
> >> Cheers
> >> Manfred
> >>
> >>> On 01.07.2014 22:02, Manfred Brenneisen wrote:
> >>>> Hi all,
> >>>>
> >>>> Garmin offers "curvy roads" preferences for their zümo 390 and 590 
> >>>> devices.
> >>>>
> >>>>
> >>> -- 
> >>> keep on biking and discovering new trails
> >>>
> >>> Felix
> >>> openmtbmap.org & www.velomap.org
> >>>
> >>>
> >>> _______________________________________________
> >>> mkgmap-dev mailing list
> >>> mkgmap-dev at lists.mkgmap.org.uk
> >>> http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
> >> _______________________________________________
> >> mkgmap-dev mailing list
> >> mkgmap-dev at lists.mkgmap.org.uk
> >> http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
> >
> >
> 
> 
> _______________________________________________
> mkgmap-dev mailing list
> mkgmap-dev at lists.mkgmap.org.uk
> http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev