logo separator

[mkgmap-dev] Update on missing POI Labels on some devices and some questions...

From Gerd Petermann gpetermann_muenchen at hotmail.com on Sat Dec 7 19:59:16 GMT 2024

Hi Scott,

OSM is just a data source, it doesn't create Garmin maps.
I think there must be two different programs which are able to generate Garmin maps with OSM data as input, one is mkgmap,
the other is the one used by Garmin. As far as I know this program creates maps in a newer IMG format and maybe there are standards how POI are treated in this newer format.
I assume that Garmin simply doesn't test new hard or firmware against old maps produced 10 years ago, and I am surprised that new devices still support this old and limited format.

It is possible that there is a not-yet-identified bit in the IMG format which influences the rendering.

The code in mkgmap which writes the IMG format is the same for .mp input and osm input, so I don't see how using OSM data as input could help you.

ciao
Gerd

________________________________________
Von: mkgmap-dev <mkgmap-dev-bounces at lists.mkgmap.org.uk> im Auftrag von scott taggart <taggart at taggarts.org>
Gesendet: Samstag, 7. Dezember 2024 20:17
An: mkgmap-dev at lists.mkgmap.org.uk
Betreff: Re: [mkgmap-dev] Update on missing POI Labels on some devices and some questions...

Thanks everyone for the input.  If I understand correctly, there seems to be no universal POI types across all devices (when I look at some of the referenced lists, there seems to be a small bit of consistency with the 0x01...0x11?) types.  However, as I have seen in practice, some of the newer devices such as XT2/Tread seem to not even follow these "standards" (the XT2 and Tread definitely do not display text for the well-known 0x1400 type).  In my particular case, I need (only) one POI type that will always display a big text label and ideally, no icon (I'm settled for now on 0x1E00 which seems to work across all devices I care about (until garmin releases some new device)). At the risk of repeating what I have previously asked, can anyone explain to me:


  *   Assuming there's no consistency across devices, how do the OSM generated maps work across devices? Do they not? (I'm not an OSM expert by any means). How the heck does OSM always know a text label will be displayed on every device for a given POI type?  When OSM maps are generated does the creator specify a device type so that internal OSM type maps can be used (sorry for my ignorance of how OSM works).  In thinking more about this, does OSM generate a TYP file that defines custom point types for all needed icons so they work across all devices? Seems like that is the only way any of this could work.

  *   Even more perplexing, does this mean that garmin must release different versions of maps for each device class (say a US topo)?  I never buy garmin maps so maybe I have missed this.  When I look online to buy the garmin 100k topo<https://www.garmin.com/en-US/p/127633#requirements>, it seems it works for all devices.

  *   Finally, and I asked this before, is there any mkgmap capabilities I can get to using the .osm input as opposed to the mp file format (assuming I generate TYP files)?  I ask this because I want to potentially save myself from trying to dive very deep into the non-mp file stuff if there's nothing to be gained.

  *   And finally, just to vent here, why in God's name would a manufacturer ever create such a mess???  Seems like a real headache even for them, let alone us reverse-engineers ;)


Thanks again for your patience.  As you can tell, I am a blind man stumbling around in a dark room here.

On 12/7/2024 8:07 AM, osm wrote:
I agree with Ticker; there is also something else which MAY help:

If your Garmin came with maps its worth checking which points do show up.

 A TYP file may be included into the gmapsupp which will should reveal
some of the types used - Gmaptool can export this supfile

Regards

Nick

On 07/12/2024 15:52, Ticker Berkin wrote:
Hi Scott

There is some consistency across Garmin models I've come across for a set of
standard POIs that have a (semi-)defined meaning; but I don't know if Garmin are
breaking this with devices like XT, Tread...

By semi-defined I mean they respond to appropriate 'FIND' searches and some
devices actually show what considers the POI to be. There are various lists of
these around the internet and, from a mkgmap distribution,
./examples/styles/default/points shows usage.

Sticking to these can make a reasonably well-featured map that works on many
devices.

Many POI types don't show at low resolution!

For the POI you've mentioned, I've noted from experimentation:
  0x14 No icon. Country. Big font. no subtypes    {major country}
  0x1e No icon. has name. State {province/region}. no subtypes

I don't think you get any difference in the final map and behaviour whether the
input is MP or from OSM (osm.pbf, o5m, etc format)

https://www.mkgmap.org.uk > Documentation is a starting point for help.

Ticker


On Fri, 2024-12-06 at 10:47 -0800, scott taggart wrote:
   As I posted here on 2024.12.01, I was having issues with POIs not displaying
labels for some garmin devices (specifically the XT2 and Tread) when
generating /img files using mp file input to mkgmap.  I did some exploration
and discovered this (maybe well known but not by me):
    * Each device model displays POIs differently (i.e., type 0x100 does not
show the same thing). There seems to be no consistency across models (Felix
echoed this in a follow-up post).
  * Each model displays labels for each POI type differently (some show no
label, others show small vs big text).  There seems to be no consistency
across models.
  * I attempted to use the custom "[_point]" feature of the mp files and mkgmap
but the custom point bitmaps only work for some garmins.  Even then, it didn't
help with my missing [poi] labels.
  * Prior to the labels not working on the XT2 and Tread units, I always used
the 0x1400 POI code type for my labels.  With a lot of cross-model
experimentation I discovered a single POI code (0x1E00) will display large
text on all garmin models I was able to test with (Montana 6XX and 7XX, XT,
XT2, Tread).  I have no idea if this POI code will work with all garmins that
support custom maps.

  Questions:
   * Are these issues with each model behaving differently with respect to POI
types well-known?  If so, how are they gotten around by (OSM) map builders?
How can a single map be built that has POIs and labels that are consistent
across more than one device.  What am I missing?
  * How does OSM handle this?  I presume that an OSM map generated for an area
works on all garmin devices?  I will admit that I don't know what the OSM map
limitations are across garmin models.  Does the JOSN model allow the devices
POI maps to be loaded on a per-map basis?  If I were to switch to JOSN model
for mkgmap input, could I get around all the device limitations I am running
into with the mp file format?  Can someone recommend a good tutorial on
getting up to speed on generating JOSN for simple map input to mkgmap?
  * OSM uses the JOSM model to feed mkgmap.  Does that model allow for more
flexibility and control than the "mp" input file model?  I presume the MP file
format is obsolete.
  * Is there any better documentation for the MPO format than the CGPSMAPPER
pdf file floating around on the internet?
  * Can anyone recommend either a different website or people whom I may
contact for further help with any of this?

  Any and all help is appreciated.


_______________________________________________
mkgmap-dev mailing list
mkgmap-dev at lists.mkgmap.org.uk<mailto:mkgmap-dev at lists.mkgmap.org.uk>
https://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
_______________________________________________
mkgmap-dev mailing list
mkgmap-dev at lists.mkgmap.org.uk<mailto:mkgmap-dev at lists.mkgmap.org.uk>
https://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
_______________________________________________
mkgmap-dev mailing list
mkgmap-dev at lists.mkgmap.org.uk<mailto:mkgmap-dev at lists.mkgmap.org.uk>
https://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev



More information about the mkgmap-dev mailing list