Re: [mkgmap-dev] Could mkgmap be extended to create an empty dem dist layer?

[Gerd Petermann]

Hi Felix,

did you also try to use as many DEM levels as --levels?  Maybe Frank is right 
and there is some relation between them?

Reg. a missing pointer: There may be something in TRE. If I got that right 
there are a lot of unknown sub sections in this file and
I guess that Garmin assumes defaults if they are missing, and those defaults 
might not fit.
I have no idea how to find out :(

Gerd


Von: mkgmap-dev  im Auftrag von Felix 
Hartmann 
Gesendet: Freitag, 20. April 2018 21:49:06
An: Development list for mkgmap
Betreff: Re: [mkgmap-dev] Could mkgmap be extended to create an empty dem   
dist layer?

I've played around a lot with different values - and it behaves very strange.
e.g. 39768,53024 --x-dem-empty-dist=106048 does not work - no DEM shown at all.
--dem-dists=39768 --x-dem-empty-dist=106048 - shows dem at 8-15km - at 20km 
then the Basemap DEM will be used instead. But at 5km where it's most important 
no DEM shown.

I kinda guess there is some pointer we are missing here in mkgmap that is not 
about the 0m DEM at all, but about the thing as a whole. It's really hard to 
find any kind of pattern that shows with resolution DEM is taken at which zoom 
level - it seems to me e.g.  39768,53024,106048 - the first 39768 will be taken 
for all zoom levels - while producing e.g.  53024,106048 without x-dem the dem 
will never be used, however dem is shown if I a some x-dem value.
Also sometimes I manged to brick my Oregon 600 to such a state that even 
without the mSD with maps inside it would not boot anymore - it got stuck on 
the first  message "Oregon 600 - Copyright 2017" and not make it to the next 
steps of loading photos, then loading maps.. Had to start it via USB, switch 
off, then  it booted now with the maps on mSD. Actually that happened to me 
also without the x-dem option when I tried different dem-dists. Maybe it's only 
safe to first remove the old dem gmapssupp.img, boot, shut down, then insert a 
new dem only gmapsupp.img - or not use the same name if you change it?

On 18 April 2018 at 21:46, Felix Hartmann 
> wrote:
Thanks a lot Gerd,
I hope I find soon some time to try it out on my devices (etrex 30, oregon 600)

On 18 April 2018 at 09:33, Gerd Petermann 
> wrote:
Hi Felix,

attached is a patch that implements option --x-dem-empty-dist as discussed.
I tried with
--dem-dists=19884,39768,53024,106048  --x-dem-empty-dist=212096
and found no visual effect, and with
--dem-dists=19884,39768,53024  --x-dem-empty-dist= 106048
the hill shading didn't work at all, and my Oregon froze when I tried to show 
the elevation data for a route.

So, I've also tried with a different patch that writes height 0 instead of 
"undefined". This seems to work,
but I am not sure if it shows the wanted effect. I've attached this 2nd version.
Please try on your own.

Gerd


Von: mkgmap-dev 
>
 im Auftrag von Felix Hartmann 
>
Gesendet: Dienstag, 17. April 2018 12:59:41
An: Development list for mkgmap
Betreff: Re: [mkgmap-dev] Could mkgmap be extended to create an empty dem   
dist layer?

I prefer to have Basemap - but I know many people delete it by accident or 
don't activate it - and I just tried a gmapsupp DEM map will also give shading 
to overview map.
And yeah - maybe call it
--dem-empty-dist

so you would have e.g.
--dem-dists=19884,39768,53024,106048  --dem-empty-dist=212096
or
--dem-dists=19884,39768,53024  --dem-empty-dist= 106048


not sure what values will work best. And yes I would expect only uknown or 
undefined heights should do the trick - but of course I do not know. With 
garmins usual approach i do guess it's possible to create it in a way that it 
switches off at some zoom level by inserting empty data.

On 17 April 2018 at 12:40, Gerd Petermann 
>>
 wrote:
Hi Felix,

it should be easy to add an empty layer, but I'd prefer a special option for 
that.
Any ideas for a good option name?
I guess an empty layer would be one that contains only "unknown" or "undefined" 
heights.
No idea what the device will do with such a layer, I'll try it.
If I got you right you prefer to have no basemap on your device? Or maybe some 
devices don't have
one by default?

Gerd


Von: mkgmap-dev 
>>
 im Auftrag von Felix Hartmann 

[mkgmap-dev] Commit r4175: poi_area-v2.patch : set special tag mkgmap:cache_area_size in generated POI for polygons and mulltipolygons

[svn commit]

Version mkgmap-r4175 was committed by gerd on Sun, 22 Apr 2018

poi_area-v2.patch : set special tag mkgmap:cache_area_size in generated POI for 
polygons and mulltipolygons

possible usage:
mkgmap:area2poi=true & mkgmap:cache_area_size > 1000 {do something ...}

Also changes the result of the area_size() style function to round the value to 
1/1000 map units.


http://www.mkgmap.org.uk/websvn/revision.php?repname=mkgmap=4175
___
mkgmap-dev mailing list
mkgmap-dev@lists.mkgmap.org.uk
http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev

[mkgmap-dev] Commit r4174: improve WrongAngleFixer: Don't remove points of ways which are end points of other ways in Douglas-Peucker.

[svn commit]

Version mkgmap-r4174 was committed by gerd on Sun, 22 Apr 2018

improve WrongAngleFixer: Don't remove points of ways which are end points of 
other ways in Douglas-Peucker.

The corresponding flag was not yet set when DP was executed.

http://www.mkgmap.org.uk/websvn/revision.php?repname=mkgmap=4174
___
mkgmap-dev mailing list
mkgmap-dev@lists.mkgmap.org.uk
http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev

Re: [mkgmap-dev] Commit r4172: improve WrongAngleFixer performance and logging

[Gerd Petermann]

Hi Bernd,

oops, that message was not meant to arrive in the final code. Fixed with r4173.

Gerd


Von: mkgmap-dev  im Auftrag von Bernd 
Weigelt 
Gesendet: Sonntag, 22. April 2018 09:50:02
An: Development list for mkgmap
Betreff: Re: [mkgmap-dev] Commit r4172: improve WrongAngleFixer performance 
and logging

Hi Gerd

SCHWERWIEGEND (WrongAngleFixer): /home/bernd/map_build/tiles/65030248.o5m:
Douglas Peucker preparation removed many points in way Way  219879749 20 -> 2

I think this messages are a little bit to much ;-)
more then 1000 for one O5M tile created with 120 maxnodes



Bernd


Am Sonntag, 22. April 2018, 08:55:26 CEST schrieb svn commit:
> Version mkgmap-r4172 was committed by gerd on Sun, 22 Apr 2018
>
> improve WrongAngleFixer performance and logging
>
> - use DouglasPeuckerFilter with error distance 0.05 m before checking
> angles. This filters oversampled ways to a reasonable amoumt of nodes and
> doesn't remove too much needed information while it is faster than the
> complex angles algo. - change level from severe to warning for "Removing
> wrong angles - didn't finish in 20 passes, giving up!" message and report
> the places that were changed in the last pass.
>
> http://www.mkgmap.org.uk/websvn/revision.php?repname=mkgmap=4172
> ___
> mkgmap-dev mailing list
> mkgmap-dev@lists.mkgmap.org.uk
> http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev

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

[mkgmap-dev] Commit r4173: remove debug code

[svn commit]

Version mkgmap-r4173 was committed by gerd on Sun, 22 Apr 2018

remove debug code

http://www.mkgmap.org.uk/websvn/revision.php?repname=mkgmap=4173
___
mkgmap-dev mailing list
mkgmap-dev@lists.mkgmap.org.uk
http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev

Re: [mkgmap-dev] Commit r4172: improve WrongAngleFixer performance and logging

[Bernd Weigelt]

Hi Gerd

SCHWERWIEGEND (WrongAngleFixer): /home/bernd/map_build/tiles/65030248.o5m: 
Douglas Peucker preparation removed many points in way Way  219879749 20 -> 2

I think this messages are a little bit to much ;-)
more then 1000 for one O5M tile created with 120 maxnodes



Bernd


Am Sonntag, 22. April 2018, 08:55:26 CEST schrieb svn commit:
> Version mkgmap-r4172 was committed by gerd on Sun, 22 Apr 2018
> 
> improve WrongAngleFixer performance and logging
> 
> - use DouglasPeuckerFilter with error distance 0.05 m before checking
> angles. This filters oversampled ways to a reasonable amoumt of nodes and
> doesn't remove too much needed information while it is faster than the
> complex angles algo. - change level from severe to warning for "Removing
> wrong angles - didn't finish in 20 passes, giving up!" message and report
> the places that were changed in the last pass.
> 
> http://www.mkgmap.org.uk/websvn/revision.php?repname=mkgmap=4172
> ___
> mkgmap-dev mailing list
> mkgmap-dev@lists.mkgmap.org.uk
> http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev



signature.asc
Description: This is a digitally signed message part.
___
mkgmap-dev mailing list
mkgmap-dev@lists.mkgmap.org.uk
http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev

Re: [mkgmap-dev] Error message WrongAngleFixer

[Gerd Petermann]

Hi Arndt,

this was fixed with r4172:
http://www.mkgmap.org.uk/websvn/revision.php?repname=mkgmap=4172

Gerd


Von: mkgmap-dev  im Auftrag von Gerd 
Petermann 
Gesendet: Samstag, 21. April 2018 08:01:59
An: Development list for mkgmap
Betreff: Re: [mkgmap-dev] Error message WrongAngleFixer

Hi Arndt,

I am working on a fix for this problem. I think I'll use the Douglas-Peucker 
filter with a very small error value to remove those duplicate points.

Something you might want to change:
I've noticed that your files contain a lot of elevation lines which also have 
many redundant points. I don't know which phyghtmap parameter would
change that but there is really no nead to write all those points, pass them 
through splitter and finally remove them again in mkgmap.

Gerd


Von: mkgmap-dev  im Auftrag von Arndt 
Röhrig 
Gesendet: Freitag, 20. April 2018 16:18:18
An: Development list for mkgmap
Betreff: Re: [mkgmap-dev] Error message WrongAngleFixer

Hi Gerd,

thank you!

Arndt

Gerd Petermann hat am 20. April 2018 um 16:07 geschrieben:


Hi Arndt,

it seems that these ways are causing trouble:
https://www.openstreetmap.org/way/203282008
https://www.openstreetmap.org/way/450539752

In both cases the mapper has added a huge amount of nodes to map a circular 
part of the way and the 20 passes to reduce wrong angles are not enough.
Please ignore the error message for now, I'll need more time to find out why 
this is causing stress. It is probably related to the latest changes in the 
angles branch.

Gerd


Von: mkgmap-dev im Auftrag von Gerd Petermann
Gesendet: Freitag, 20. April 2018 13:57:53
An: Development list for mkgmap
Betreff: Re: [mkgmap-dev] Error message WrongAngleFixer

Hi Arndt,

there is nothing that you can do. I'll try to reproduce why 20 passes are not 
enough.

Gerd


Von: mkgmap-dev im Auftrag von Arndt Röhrig
Gesendet: Freitag, 20. April 2018 13:20:35
An: Development list for mkgmap
Betreff: [mkgmap-dev] Error message WrongAngleFixer

Hi @all,

mkgmap r4165 tell me this:

SCHWERWIEGEND (WrongAngleFixer): 
.\Baustelle\Speiche_Fabrik_Splitter\88014475.osm.pbf: Removing wrong angles - 
didn't finish in 20 passes, giving up!
SCHWERWIEGEND (WrongAngleFixer): 
.\Baustelle\Speiche_Fabrik_Splitter\88013939.osm.pbf: Removing wrong angles - 
didn't finish in 20 passes, giving up!

What does that mean and what can i do to correct this error.


Greets

Arndt


www.speichenkarte.de/88014475.osm.pbf

www.speichenkarte.de/88013939.osm.pbf
___
mkgmap-dev mailing list
mkgmap-dev@lists.mkgmap.org.uk
http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
___
mkgmap-dev mailing list
mkgmap-dev@lists.mkgmap.org.uk
http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
___
mkgmap-dev mailing list
mkgmap-dev@lists.mkgmap.org.uk
http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
___
mkgmap-dev mailing list
mkgmap-dev@lists.mkgmap.org.uk
http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev

[mkgmap-dev] Commit r4172: improve WrongAngleFixer performance and logging

[svn commit]

Version mkgmap-r4172 was committed by gerd on Sun, 22 Apr 2018

improve WrongAngleFixer performance and logging

- use DouglasPeuckerFilter with error distance 0.05 m before checking angles. 
This filters oversampled ways to a reasonable amoumt of nodes and doesn't 
remove too much needed information while it is faster than the complex angles 
algo.
- change level from severe to warning for "Removing wrong angles - didn't 
finish in 20 passes, giving up!" message and report the places that were 
changed in the last pass.

http://www.mkgmap.org.uk/websvn/revision.php?repname=mkgmap=4172
___
mkgmap-dev mailing list
mkgmap-dev@lists.mkgmap.org.uk
http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev

Re: [mkgmap-dev] Commit r4171: fix possible java.lang.AssertionError introduced with r4169

[Gerd Petermann]

Hi all,

I've just noticed a typo in the commit message:
This problem was introduced with r4167, not 4169. Just in case someone 
wonders...

Gerd


Von: mkgmap-dev  im Auftrag von Gerd 
Petermann 
Gesendet: Samstag, 21. April 2018 19:35:10
An: Development list for mkgmap
Betreff: Re: [mkgmap-dev] Commit r4171: fix possible
java.lang.AssertionError introduced with r4169

Hi Bernd,

since a few days Microsoft puts all my mails which are returned from the list 
into the junk-mail folder.
No idea why, maybe you find it there?

Gerd


Von: mkgmap-dev  im Auftrag von Bernd 
Weigelt 
Gesendet: Samstag, 21. April 2018 19:00:49
An: Development list for mkgmap
Betreff: Re: [mkgmap-dev] Commit r4171: fix possible
java.lang.AssertionError introduced with r4169

Am Samstag, 21. April 2018, 18:31:41 CEST schrieb svn commit:
> Version mkgmap-r4171 was committed by gerd on Sat, 21 Apr 2018
>
> fix possible  java.lang.AssertionError introduced with r4169
>
> nodecount has more than 8 bits
>
> http://www.mkgmap.org.uk/websvn/revision.php?repname=mkgmap=4171
> ___
> mkgmap-dev mailing list
> mkgmap-dev@lists.mkgmap.org.uk
> http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev


problem is solved

@Gerd:
I can see your answer in the mailing list archive, but i didn't received it
here

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