[mkgmap-dev] feedback on sharp-angles-v4.patch?

[Gerd Petermann]

Hi all,

I am still waiting for feedback from Felix, but maybe other are willing to contribute?

See http://gis.19327.n5.nabble.com/Patch-v4-sharp-angles-tp5854127.html

Gerd

[Felix Hartmann]

Hi Gerd,

well the problem is that I only test on long distances. So my opinion -
v4 and not giving --cycle-map is what I found best.
As for the differences between v2-v3-v4, it's so small in the end result
that I really cannot objectively classify one or the other as better.

Overall it's clearly an improvement.

On 13.10.2015 09:39, Gerd Petermann wrote:
> Hi all,
>
> I am still waiting for feedback from Felix, but maybe other are
> willing to contribute?
>
> See http://gis.19327.n5.nabble.com/Patch-v4-sharp-angles-tp5854127.html
>
> Gerd
>
>
> _______________________________________________
> mkgmap-dev mailing list
> mkgmap-***@lists.mkgmap.org.uk
> http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev

--
keep on biking and discovering new trails

Felix
openmtbmap.org & www.velomap.org

[lasemax]

Hi Gerd,
you can upload http://files.mkgmap.org.uk/download/282/mkgmap.jar again. I
can´t find the file.




--
View this message in context: http://gis.19327.n5.nabble.com/feedback-on-sharp-angles-v4-patch-tp5856851p5856860.html
Sent from the Mkgmap Development mailing list archive at Nabble.com.

[Steve Ratcliffe]

Hi

> you can upload http://files.mkgmap.org.uk/download/282/mkgmap.jar again. I
> can´t find the file.

I've made the original file re-appear.

..Steve

[Minko]

Hi Gerd
I've tried to run it but cycle-map is not a valid option?

[Minko]

Sorry my fault, I have used the wrong/older mkgmap jar file, now it is running...

> Hi Gerd
> I've tried to run it but cycle-map is not a valid option?

[Minko]

Hi Gerd,
I have tested it a bit and noticed some improvements in routing at sharp corners.
Haven't seen any major differences or issues yet, i think it's a good improvement.

[Gerd Petermann]

Hi Minko,

thanks for testing. The open question is wether the
--cylcle-map option changes something to the better.

Gerd

________________________________________
Von: mkgmap-dev-***@lists.mkgmap.org.uk <mkgmap-dev-***@lists.mkgmap.org.uk> im Auftrag von Minko <***@online.nl>
Gesendet: Samstag, 17. Oktober 2015 10:55
An: Development list for mkgmap
Betreff: Re: [mkgmap-dev] feedback on sharp-angles-v4.patch?

Hi Gerd,
I have tested it a bit and noticed some improvements in routing at sharp corners.
Haven't seen any major differences or issues yet, i think it's a good improvement.

[Minko]

I have tested it with --cycle-map
What is the difference vs without this option?

[Gerd Petermann]

Hi Minko,

if you can't find any, that is also telling me that the option is rather useless.

With --cycle-map mkgmap might produce larger changes when the
sum of the road_speed values of two arcs is rather high.

The corresponding part of the patch:
+ if (cycleMap){
+// if (sumSpeeds >= 14)
+// maskedMinAngle = 0x80;
+// if (sumSpeeds >= 12)
+// maskedMinAngle = 0x70;
+// if (sumSpeeds >= 10)
+// maskedMinAngle = 0x60;
+ if (sumSpeeds >= 8)
+ maskedMinAngle = 0x50;
+ else if (sumSpeeds >= 6)
+ maskedMinAngle = 0x40;
+ else if (sumSpeeds >= 4)
+ maskedMinAngle = 0x30;
+ }

Does that help?

Gerd
________________________________________
Von: mkgmap-dev-***@lists.mkgmap.org.uk <mkgmap-dev-***@lists.mkgmap.org.uk> im Auftrag von Minko <***@online.nl>
Gesendet: Samstag, 17. Oktober 2015 15:17
An: Development list for mkgmap
Betreff: Re: [mkgmap-dev] feedback on sharp-angles-v4.patch?

I have tested it with --cycle-map
What is the difference vs without this option?

[Minko]

Hi Gerd,
My highest road_speed value is 3.
Maybe this is why this option doesn't make very little difference, if any at all?


> Hi Minko,
>
> if you can't find any, that is also telling me that the option is rather
> useless.
>
> With --cycle-map mkgmap might produce larger changes when the
> sum of the road_speed values of two arcs is rather high.
>
> The corresponding part of the patch:
> + if (cycleMap){
> +// if (sumSpeeds >= 14)
> +// maskedMinAngle = 0x80;
> +// if (sumSpeeds >= 12)
> +// maskedMinAngle = 0x70;
> +// if (sumSpeeds >= 10)
> +// maskedMinAngle = 0x60;
> + if (sumSpeeds >= 8)
> + maskedMinAngle = 0x50;
> + else if (sumSpeeds >= 6)
> + maskedMinAngle = 0x40;
> + else if (sumSpeeds >= 4)
> + maskedMinAngle = 0x30;
> + }
>
> Does that help?

[Gerd Petermann]

Hi Minko,

you may see a difference when two roads with road_speed=3 build a sharp angle.
Maybe the only difference is the calculated travel time, in rare cases a different
route will be used because of this difference.

Gerd

________________________________________
Von: mkgmap-dev-***@lists.mkgmap.org.uk <mkgmap-dev-***@lists.mkgmap.org.uk> im Auftrag von Minko <***@online.nl>
Gesendet: Samstag, 17. Oktober 2015 16:04
An: Development list for mkgmap
Betreff: Re: [mkgmap-dev] feedback on sharp-angles-v4.patch?

Hi Gerd,
My highest road_speed value is 3.
Maybe this is why this option doesn't make very little difference, if any at all?


> Hi Minko,
>
> if you can't find any, that is also telling me that the option is rather
> useless.
>
> With --cycle-map mkgmap might produce larger changes when the
> sum of the road_speed values of two arcs is rather high.
>
> The corresponding part of the patch:
> + if (cycleMap){
> +// if (sumSpeeds >= 14)
> +// maskedMinAngle = 0x80;
> +// if (sumSpeeds >= 12)
> +// maskedMinAngle = 0x70;
> +// if (sumSpeeds >= 10)
> +// maskedMinAngle = 0x60;
> + if (sumSpeeds >= 8)
> + maskedMinAngle = 0x50;
> + else if (sumSpeeds >= 6)
> + maskedMinAngle = 0x40;
> + else if (sumSpeeds >= 4)
> + maskedMinAngle = 0x30;
> + }
>
> Does that help?

[GerdP]

Hi all,

sorry, forgot the more important effect of this option:
With --cycle-map you can tell mkgmap that different
vehicles like car, truck are used as replacements for
racing bike, mtb, normal bike or whatever.

I'll review that part again as it doesn't seem to work
as expected.

Gerd



GerdP wrote
> Hi Minko,
>
> if you can't find any, that is also telling me that the option is rather
> useless.
>
> With --cycle-map mkgmap might produce larger changes when the
> sum of the road_speed values of two arcs is rather high.
>
> The corresponding part of the patch:
> + if (cycleMap){
> +// if (sumSpeeds >= 14)
> +// maskedMinAngle = 0x80;
> +// if (sumSpeeds >= 12)
> +// maskedMinAngle = 0x70;
> +// if (sumSpeeds >= 10)
> +// maskedMinAngle = 0x60;
> + if (sumSpeeds >= 8)
> + maskedMinAngle = 0x50;
> + else if (sumSpeeds >= 6)
> + maskedMinAngle = 0x40;
> + else if (sumSpeeds >= 4)
> + maskedMinAngle = 0x30;
> + }
>
> Does that help?
>
> Gerd
> ________________________________________
> Von:

> mkgmap-dev-***@.org

> &lt;

> mkgmap-dev-***@.org

> &gt; im Auftrag von Minko &lt;

> ligfietser@

> &gt;
> Gesendet: Samstag, 17. Oktober 2015 15:17
> An: Development list for mkgmap
> Betreff: Re: [mkgmap-dev] feedback on sharp-angles-v4.patch?
>
> I have tested it with --cycle-map
> What is the difference vs without this option?
>
> _______________________________________________
> mkgmap-dev mailing list

> mkgmap-***@.org

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

> mkgmap-***@.org

> http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev





--
View this message in context: http://gis.19327.n5.nabble.com/feedback-on-sharp-angles-v4-patch-tp5856851p5857337.html
Sent from the Mkgmap Development mailing list archive at Nabble.com.

[GerdP]

Hi all,

okay, it took a while because I am busy cleaning up wrong highways...

My thinking was this:
1) We want to enlarge sharp angles because we know that a cyclist
doesn't care that much about them and typically they are not as sharp
in reality.
2) I don't want to enlarge sharp angles at junctions which are only used by
cars, esp. I don't want to introduce time penalties at motorway junctions
and I don't want to remove time penalties at junctions where a turn
restriction
is missing and the Garmin algo avoids routing through the sharp angle.
3) The --cycle-map option allows to ignore the problem cases in 2)
and (with v4) it also allows to enlarge the angles even more.

I guess that this later part is not a good idea, so I've commented that
part of the code and changed the documentation of the option a little bit.

See also svn log for r3649.

Gerd


GerdP wrote
> Hi all,
>
> sorry, forgot the more important effect of this option:
> With --cycle-map you can tell mkgmap that different
> vehicles like car, truck are used as replacements for
> racing bike, mtb, normal bike or whatever.
>
> I'll review that part again as it doesn't seem to work
> as expected.
>
> Gerd
>
> GerdP wrote
>> Hi Minko,
>>
>> if you can't find any, that is also telling me that the option is rather
>> useless.
>>
>> With --cycle-map mkgmap might produce larger changes when the
>> sum of the road_speed values of two arcs is rather high.
>>
>> The corresponding part of the patch:
>> + if (cycleMap){
>> +// if (sumSpeeds >= 14)
>> +// maskedMinAngle = 0x80;
>> +// if (sumSpeeds >= 12)
>> +// maskedMinAngle = 0x70;
>> +// if (sumSpeeds >= 10)
>> +// maskedMinAngle = 0x60;
>> + if (sumSpeeds >= 8)
>> + maskedMinAngle = 0x50;
>> + else if (sumSpeeds >= 6)
>> + maskedMinAngle = 0x40;
>> + else if (sumSpeeds >= 4)
>> + maskedMinAngle = 0x30;
>> + }
>>
>> Does that help?
>>
>> Gerd
>> ________________________________________
>> Von:

>> mkgmap-dev-***@.org

>> &lt;

>> mkgmap-dev-***@.org

>> &gt; im Auftrag von Minko &lt;

>> ligfietser@

>> &gt;
>> Gesendet: Samstag, 17. Oktober 2015 15:17
>> An: Development list for mkgmap
>> Betreff: Re: [mkgmap-dev] feedback on sharp-angles-v4.patch?
>>
>> I have tested it with --cycle-map
>> What is the difference vs without this option?
>>
>> _______________________________________________
>> mkgmap-dev mailing list

>> mkgmap-***@.org

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

>> mkgmap-***@.org

>> http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev





--
View this message in context: http://gis.19327.n5.nabble.com/feedback-on-sharp-angles-v4-patch-tp5856851p5857988.html
Sent from the Mkgmap Development mailing list archive at Nabble.com.