Is this patch still needed?

I must say it didn't work at all anyhow so far.

I think the problem is that for many ways I don't add oneway tag, but still need the reversal to work as in 2013. Here is a list of direction dependant tags - that also if no oneway present MUST make sure that road is not reversed or old behaviour kept:

incline
bicycle:left
bicycle:right
oneway:bicycle
bicycle:oneway
cycleway:left
cycleway:right

(and maybe some others I right now don't recall).


So far I haven't been able to render a correct map, with or without patch. But I noticed that the patch doesn't apply clean anymore and conflicts - so is there something for the better?
Far too often the bicycle way has been on the wrong side of the road, or arrows showing difficult to bike uphill (incline & mtb:scale:uphill>1) way arrows showing wrong direction!


(P.S. I'm basically still detached from fast internet and can only start playing around from next Monday).

On 20.04.2014 13:45, Gerd Petermann wrote:
Hi WanMil,

> Ok, anyhow I will slowly start with some tests with the tag reverser. I
> think most of it will be static so it won't hurt if other things are
> refactored.
Great. If you can implement it in HighwayHooks  it should not create
problems.

>
> >
> > Maybe you can look at it:
> > I plan to evaluate all tags used in RoadMerger once
> > in ConvertedWay and place them
> > in fields (most of them are booleans).
> > RoadMerger should then call ConvertedWay.isSimilar()
> > which will implement most of the compare routines now implemented
> > in RoadMerger.Road.
> > Does that sound okay for you?
>
> I don't see the big advantage now but I am sure you have a good idea why
> it helps :-)
> One small comment: I have some doubts that it is a good idea to put the
> isSimilar() method to the ConvertedWay class. There is a similar
> approach in MapElement. But the implementation does not really help.
> Maybe there are also different views at different stages of the
> processing if two elements are similar or not. So better leave the
> isSimilar method at the RoadMerger who knows better if the two elements
> can be merged or not.

Good point.
My idea is that RoadMerger should use the ConvertedWay instances
instead of creating new Road instances, but most of the code in RoadMerger
is in the Road class.

Gerd



_______________________________________________
mkgmap-dev mailing list
mkgmap-dev@lists.mkgmap.org.uk
http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev