The OsmAnd development team mentioned many years ago that an hc=1.0 would
deliver the only really "correct" route, as all "paths" will be checked in
the "shortest path" algorithm. Therefore, the only correct route would be
chosen.
So it was not a bd algorithm or bad implementation: it was a design
decision. One I "fought" really hard at that time as at that time I was
more involved in the entire project (as I considered it nonsense. Use
hc=1.0 in a city with lots of possibilties. Use hc=1.2~1.5 for anything
else).
In practice this hc=1.0 is hardly ever necessary, especially on (slightly)
longer routes, but for that you have to read the posts from several people
here (and I do not mean myself), or simply check my pdf and redo the tests
for yourself. Or search on "A* shortest path", "heuristic coefficient"
etcetera. When you start there you will find a wealth of articles. Also
check youtube if you are more visually oriented.

I simply did not test 1.4. If 1.3 and 1.5 deliver the same route, why
should I test 1.4. (You can do the tests yourself. ;) )
I use an hc=1.2 in cities, as I do consider OsmAnd the best in cities (like
explained earlier ;) )
I use an hc=1.5 for longer trips, but only again since 2 months or so since
the profiles option became available, and with roads_only maps (another
option OsmAnd offers, BTW).

The last 2 years I used Magic Earth. Before that and starting from approx.
2014 I used Mapfactor Navigator for car routing. I used OsmAnd for
everything else and now again more and more for car routing as well.
And indeed 3D (or 2.5D or birdview, etc) is nicer for navigation, but
actually I do not care much. I drive a lot of rental cars for my work, and
some also have 2D only or "some other renter" put it back to 2D and/or
"North up". I don't care much. Once in traffic I am not going to fiddle
with it.
Most advanced systems rotate the map back to 2D in case you need to go
right/left in a city, and move back to 3D when going straight ahead again.
So I guess that map makers consider 2D better when you really need to see
which right/left turn you need to take (not talking about motorway exits
here).
Anyway, you have to watch the road, not the screen.


Op za 28 mrt. 2020 om 13:11 schreef zanny <[email protected]>:

> Thanks for the PDF summary, very explicative.
> A couple of questions from my side (I apologise if those have been asked
> and answered somewhere else already).
> 1) why is hc=1 the standard parameter? in 5 of 7 cases the results (in
> terms of distance and travel time) is the same as hc=1.5. And in the
> others, the gap between the results of hc=1 and hc=1.5 is minimal
> (negligible).
> 2) out of curiosity: is there a technical reason for factor 1.4 not being
> tested?
>
> thanks
>
> Il giorno sabato 28 marzo 2020 11:46:12 UTC+1, Harry van der Wolf ha
> scritto:
>>
>> Like I have stated before: Use a different profile with a different
>> heuristic coefficient(See .xmls attached).
>> In another mail thread "heuristic coefficient comparisons" recently, I
>> did comparisons also against Magic Earth. See attached pdf. I don't care
>> about differences in seconds. I do care about differences in minutes as
>> this makes it unusable.
>> I wrote extensively about using these profiles and how they improve
>> OsmAnd a lot for car navigation, next to being it the "swiis knife" for all
>> other navigation/tracking/hiking functionalities with overlay/underlay
>> maps, and you name it.
>>
>> Finally: 3D is in the make. I also posted about that in this mailing
>> group, referring to one of the telegram groups where this is shared.
>>
>> Just do what you want with it.
>> Trying to prove that OsmAnd is "bad" mail after mail, only leads to one
>> conclusion: Use another app and stop wasting your time on such a "bad" app.
>>
>> Harry
>>
>> --
> You received this message because you are subscribed to the Google Groups
> "OsmAnd" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to [email protected].
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/osmand/a0dfee17-b1b9-4569-886a-e329f4d7e39a%40googlegroups.com
> <https://groups.google.com/d/msgid/osmand/a0dfee17-b1b9-4569-886a-e329f4d7e39a%40googlegroups.com?utm_medium=email&utm_source=footer>
> .
>

-- 
You received this message because you are subscribed to the Google Groups 
"OsmAnd" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To view this discussion on the web visit 
https://groups.google.com/d/msgid/osmand/CAGARPpt0igt-kj%2B%3DtfpHCLdhvvvi3qQp2UDq_ncvoMJfjFKpEw%40mail.gmail.com.

Reply via email to