One thing I think I missed, is that each tile on a garmin device needs a unique mapid. A different family-id and mapname for each gmap is not sufficient. Does this sound correct? I'd been downloading US states from Geofabrik, creating maps with unique family-id, mapname, every name & id that seemed important, (with mkgmap) but when I put them on the Garmin they all wouldn't show up. Only when I went back to the splitter step & used a different mapid did they all show up. Sometimes I'm not too good at reading directions, did I overlook this someplace? Brad
Hi all, I've never tried it, but my understanding was that you can have the same tile (mapid) in two different gmapsupp files. E.g. you might create a tile containing an area around Hamburg and put that tile into a gmapsupp for Germany as well as another one for Lower Saxony. If that is right it would explain why you have to make sure that the content of two tiles with the same id doesn't differ. @Brad: I think the current documentation about the meaning of all those names and ids is not perfect because the authors of mkgmap were still learning when the docu was created, and it's an ongoing process ;-) Feel free to improve the docu (post a patch or maybe a new version of one or more files). Gerd ________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von brad <bradhaack@fastmail.com> Gesendet: Montag, 17. Dezember 2018 01:47 An: mkgmap-dev@lists.mkgmap.org.uk Betreff: [mkgmap-dev] splitter, mapid One thing I think I missed, is that each tile on a garmin device needs a unique mapid. A different family-id and mapname for each gmap is not sufficient. Does this sound correct? I'd been downloading US states from Geofabrik, creating maps with unique family-id, mapname, every name & id that seemed important, (with mkgmap) but when I put them on the Garmin they all wouldn't show up. Only when I went back to the splitter step & used a different mapid did they all show up. Sometimes I'm not too good at reading directions, did I overlook this someplace? Brad _______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
Hi Gerd I've learnt this lesson the hard way, walking through france with 2 maps sharing the same map id :( None of them worked! Similarly,if you want your Recent Findings to work, make sure you use the same map ids else you get the following message: 'This route does not match the available maps' I would also add that more than likely the FID has to be the same . It appears Garmin attaches the tile number (and FID) to each POI search result. r Nick On 17/12/2018 06:57, Gerd Petermann wrote:
Hi all,
I've never tried it, but my understanding was that you can have the same tile (mapid) in two different gmapsupp files. E.g. you might create a tile containing an area around Hamburg and put that tile into a gmapsupp for Germany as well as another one for Lower Saxony. If that is right it would explain why you have to make sure that the content of two tiles with the same id doesn't differ.
@Brad: I think the current documentation about the meaning of all those names and ids is not perfect because the authors of mkgmap were still learning when the docu was created, and it's an ongoing process ;-) Feel free to improve the docu (post a patch or maybe a new version of one or more files).
Gerd
________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von brad <bradhaack@fastmail.com> Gesendet: Montag, 17. Dezember 2018 01:47 An: mkgmap-dev@lists.mkgmap.org.uk Betreff: [mkgmap-dev] splitter, mapid
One thing I think I missed, is that each tile on a garmin device needs a unique mapid. A different family-id and mapname for each gmap is not sufficient. Does this sound correct?
I'd been downloading US states from Geofabrik, creating maps with unique family-id, mapname, every name & id that seemed important, (with mkgmap) but when I put them on the Garmin they all wouldn't show up. Only when I went back to the splitter step & used a different mapid did they all show up.
Sometimes I'm not too good at reading directions, did I overlook this someplace?
Brad _______________________________________________ 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
Hi, On Mon, Dec 17, Gerd Petermann wrote:
Hi all,
I've never tried it, but my understanding was that you can have the same tile (mapid) in two different gmapsupp files. E.g. you might create a tile containing an area around Hamburg and put that tile into a gmapsupp for Germany as well as another one for Lower Saxony. If that is right it would explain why you have to make sure that the content of two tiles with the same id doesn't differ.
Correct, this works, but only if the tile is 100% identical for both maps. Or in other words: create a big map and select tiles from this map to build smaller maps. So the same as if you select tiles in mapsource from a bigger map and transfer them to the GPS device. Thorsten
@Brad: I think the current documentation about the meaning of all those names and ids is not perfect because the authors of mkgmap were still learning when the docu was created, and it's an ongoing process ;-) Feel free to improve the docu (post a patch or maybe a new version of one or more files).
Gerd
________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von brad <bradhaack@fastmail.com> Gesendet: Montag, 17. Dezember 2018 01:47 An: mkgmap-dev@lists.mkgmap.org.uk Betreff: [mkgmap-dev] splitter, mapid
One thing I think I missed, is that each tile on a garmin device needs a unique mapid. A different family-id and mapname for each gmap is not sufficient. Does this sound correct?
I'd been downloading US states from Geofabrik, creating maps with unique family-id, mapname, every name & id that seemed important, (with mkgmap) but when I put them on the Garmin they all wouldn't show up. Only when I went back to the splitter step & used a different mapid did they all show up.
Sometimes I'm not too good at reading directions, did I overlook this someplace?
Brad _______________________________________________ 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
-- Thorsten Kukuk, Distinguished Engineer, Senior Architect SLES & MicroOS SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nuernberg, Germany GF: Felix Imendoerffer, Jane Smithard, Graham Norton, HRB 21284 (AG Nuernberg)
I'd be happy to take a stab at updating some docs. It looks like the docs are the same on mkgmap.org and wiki.openstreetmap.org. Do the changes on mkgmap.org flow to OSM, or the other way? Should I post updates here? http://www.mkgmap.org.uk/doc/splitter.html On 12/16/18 11:57 PM, Gerd Petermann wrote:
Hi all,
I've never tried it, but my understanding was that you can have the same tile (mapid) in two different gmapsupp files. E.g. you might create a tile containing an area around Hamburg and put that tile into a gmapsupp for Germany as well as another one for Lower Saxony. If that is right it would explain why you have to make sure that the content of two tiles with the same id doesn't differ.
@Brad: I think the current documentation about the meaning of all those names and ids is not perfect because the authors of mkgmap were still learning when the docu was created, and it's an ongoing process ;-) Feel free to improve the docu (post a patch or maybe a new version of one or more files).
Gerd
________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von brad <bradhaack@fastmail.com> Gesendet: Montag, 17. Dezember 2018 01:47 An: mkgmap-dev@lists.mkgmap.org.uk Betreff: [mkgmap-dev] splitter, mapid
One thing I think I missed, is that each tile on a garmin device needs a unique mapid. A different family-id and mapname for each gmap is not sufficient. Does this sound correct?
I'd been downloading US states from Geofabrik, creating maps with unique family-id, mapname, every name & id that seemed important, (with mkgmap) but when I put them on the Garmin they all wouldn't show up. Only when I went back to the splitter step & used a different mapid did they all show up.
Sometimes I'm not too good at reading directions, did I overlook this someplace?
Brad _______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
Hi Brad, I think you should start with the sources in splitter\doc You can get the source package from http://www.mkgmap.org.uk/download/splitter.html or via svn with svn co https://svn.mkgmap.org.uk/svn/splitter Gerd ________________________________________ Von: brad <bradhaack@fastmail.com> Gesendet: Montag, 17. Dezember 2018 17:59 An: Gerd Petermann; Development list for mkgmap Betreff: Re: AW: [mkgmap-dev] splitter, mapid I'd be happy to take a stab at updating some docs. It looks like the docs are the same on mkgmap.org and wiki.openstreetmap.org. Do the changes on mkgmap.org flow to OSM, or the other way? Should I post updates here? http://www.mkgmap.org.uk/doc/splitter.html On 12/16/18 11:57 PM, Gerd Petermann wrote:
Hi all,
I've never tried it, but my understanding was that you can have the same tile (mapid) in two different gmapsupp files. E.g. you might create a tile containing an area around Hamburg and put that tile into a gmapsupp for Germany as well as another one for Lower Saxony. If that is right it would explain why you have to make sure that the content of two tiles with the same id doesn't differ.
@Brad: I think the current documentation about the meaning of all those names and ids is not perfect because the authors of mkgmap were still learning when the docu was created, and it's an ongoing process ;-) Feel free to improve the docu (post a patch or maybe a new version of one or more files).
Gerd
________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von brad <bradhaack@fastmail.com> Gesendet: Montag, 17. Dezember 2018 01:47 An: mkgmap-dev@lists.mkgmap.org.uk Betreff: [mkgmap-dev] splitter, mapid
One thing I think I missed, is that each tile on a garmin device needs a unique mapid. A different family-id and mapname for each gmap is not sufficient. Does this sound correct?
I'd been downloading US states from Geofabrik, creating maps with unique family-id, mapname, every name & id that seemed important, (with mkgmap) but when I put them on the Garmin they all wouldn't show up. Only when I went back to the splitter step & used a different mapid did they all show up.
Sometimes I'm not too good at reading directions, did I overlook this someplace?
Brad _______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
Hi, basically map ID should be unique for all tiles, regardless of FID. If you load 2 tiles with the same map ID to a GPS, it will silently ignore one of these tiles. Most probably GPS assume, that these are the same tiles and there is no reason to process both. Search index contains map ID. If you use 2 maps with repeated map ID, then most probably search won't work correctly. There is no problem with map ID in Basecamp or Mapsource, these programs don't support multiple maps. I suggest to link map ID with FID using following formula: First map ID in a mapset = FID*2048 + 6316032 Or in hex: First map ID in a mapset = FID*800h + 606000h This formula assigns 2048 IDs for each FID. It should be true also for original Garmin's maps, except for City Navigator series. Gerd, maybe you could include these formula into splitter? Something like: ;--fid=27795 : Calculate mapid basing on a FID value. Can be used instead of --mapid. Splitter sets first mapid value as 2048*fid + 6316032 -- Best regards, Andrzej
What is the max famid? the docs say 9999, but I know there are mapid's larger. maybe 99999? On 12/17/18 7:17 AM, Andrzej Popowski wrote:
Hi,
basically map ID should be unique for all tiles, regardless of FID. If you load 2 tiles with the same map ID to a GPS, it will silently ignore one of these tiles. Most probably GPS assume, that these are the same tiles and there is no reason to process both.
Search index contains map ID. If you use 2 maps with repeated map ID, then most probably search won't work correctly.
There is no problem with map ID in Basecamp or Mapsource, these programs don't support multiple maps.
I suggest to link map ID with FID using following formula:
First map ID in a mapset = FID*2048 + 6316032
Or in hex: First map ID in a mapset = FID*800h + 606000h
This formula assigns 2048 IDs for each FID. It should be true also for original Garmin's maps, except for City Navigator series.
Gerd, maybe you could include these formula into splitter? Something like: ;--fid=27795 : Calculate mapid basing on a FID value. Can be used instead of --mapid. Splitter sets first mapid value as 2048*fid + 6316032
65535, 0xFFFF On 12/17/18 4:41 PM, brad wrote:
What is the max famid? the docs say 9999, but I know there are mapid's larger. maybe 99999?
On 12/17/18 7:17 AM, Andrzej Popowski wrote:
Hi,
basically map ID should be unique for all tiles, regardless of FID. If you load 2 tiles with the same map ID to a GPS, it will silently ignore one of these tiles. Most probably GPS assume, that these are the same tiles and there is no reason to process both.
Search index contains map ID. If you use 2 maps with repeated map ID, then most probably search won't work correctly.
There is no problem with map ID in Basecamp or Mapsource, these programs don't support multiple maps.
I suggest to link map ID with FID using following formula:
First map ID in a mapset = FID*2048 + 6316032
Or in hex: First map ID in a mapset = FID*800h + 606000h
This formula assigns 2048 IDs for each FID. It should be true also for original Garmin's maps, except for City Navigator series.
Gerd, maybe you could include these formula into splitter? Something like: ;--fid=27795 : Calculate mapid basing on a FID value. Can be used instead of --mapid. Splitter sets first mapid value as 2048*fid + 6316032
_______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
Hi Andrzej, thanks for that formula. I like the idea to use the fid as option for splitter. I'm not yet sure if I understand. Up to now we use values for mapid that are up to 99.999.999 so that the tile names can use the number. Presuming that fid can be any value up to 65535 we get 2048 * 65535 + 6316032 = 140.531.712. Should splitter change to hex in this case, e.g. 08605800.osm? Can you explain the odd number 6316032? Did you find it by looking at the numbers used in Garmin maps? Gerd ________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Andrzej Popowski <popej@poczta.onet.pl> Gesendet: Montag, 17. Dezember 2018 15:17 An: mkgmap-dev@lists.mkgmap.org.uk Betreff: Re: [mkgmap-dev] splitter, mapid Hi, basically map ID should be unique for all tiles, regardless of FID. If you load 2 tiles with the same map ID to a GPS, it will silently ignore one of these tiles. Most probably GPS assume, that these are the same tiles and there is no reason to process both. Search index contains map ID. If you use 2 maps with repeated map ID, then most probably search won't work correctly. There is no problem with map ID in Basecamp or Mapsource, these programs don't support multiple maps. I suggest to link map ID with FID using following formula: First map ID in a mapset = FID*2048 + 6316032 Or in hex: First map ID in a mapset = FID*800h + 606000h This formula assigns 2048 IDs for each FID. It should be true also for original Garmin's maps, except for City Navigator series. Gerd, maybe you could include these formula into splitter? Something like: ;--fid=27795 : Calculate mapid basing on a FID value. Can be used instead of --mapid. Splitter sets first mapid value as 2048*fid + 6316032 -- Best regards, Andrzej _______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
Hi Gerd, yes, I got the formula after examining numerous Garmin's maps. I know, that City Navigator is different, my guess is that next CN map gets next free mapid. I'm not sure about marine maps and other new maps from Garmin. It looks like the formula is easy to support up to FID 45744. For bigger values you could take different strategy. The rules I know are following: Mapid inside img data is 4-byte value, no problem here. File names are independent of mapid, could be any arbitrary 8-digit number (number from file name could be even written in TRE header, at offset 0xCF). Names of subfiles like 12345678.TRE can be written in HEX, in this case you should use prefix 'I'. So 12345678.TRE is equivalent of I0BC614E.TRE. Mapid 140.531.712 could correspond to I8605800.TRE. My first suggestion is to limit --fid to max 45744. For bigger FID option --mapid should be used. Second proposition is to overlap mapid values for big FIDs, but with an offset of 1024. So for FID > 45744 formula could be like this: mapid = (fid - 19792)*2048 + 6316032 + 1024 -- Best regards, Andrzej
Hi Andrzej, sorry, you totally lost me here. I fight with a cold since 8 days ;-) I'll look at this again once I feel good. again. Gerd ________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Andrzej Popowski <popej@poczta.onet.pl> Gesendet: Dienstag, 18. Dezember 2018 14:14 An: mkgmap-dev@lists.mkgmap.org.uk Betreff: Re: [mkgmap-dev] splitter, mapid Hi Gerd, yes, I got the formula after examining numerous Garmin's maps. I know, that City Navigator is different, my guess is that next CN map gets next free mapid. I'm not sure about marine maps and other new maps from Garmin. It looks like the formula is easy to support up to FID 45744. For bigger values you could take different strategy. The rules I know are following: Mapid inside img data is 4-byte value, no problem here. File names are independent of mapid, could be any arbitrary 8-digit number (number from file name could be even written in TRE header, at offset 0xCF). Names of subfiles like 12345678.TRE can be written in HEX, in this case you should use prefix 'I'. So 12345678.TRE is equivalent of I0BC614E.TRE. Mapid 140.531.712 could correspond to I8605800.TRE. My first suggestion is to limit --fid to max 45744. For bigger FID option --mapid should be used. Second proposition is to overlap mapid values for big FIDs, but with an offset of 1024. So for FID > 45744 formula could be like this: mapid = (fid - 19792)*2048 + 6316032 + 1024 -- Best regards, Andrzej _______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
Hi Andrzej, looked at this again now. What I don't understand is why you want to use exactly the same formular as Garmin. I think my approach would be to make sure that we DON'T use their numbers to avoid conflicts. Gerd popej wrote
Hi Gerd,
yes, I got the formula after examining numerous Garmin's maps. I know, that City Navigator is different, my guess is that next CN map gets next free mapid. I'm not sure about marine maps and other new maps from Garmin.
It looks like the formula is easy to support up to FID 45744. For bigger values you could take different strategy. The rules I know are following:
Mapid inside img data is 4-byte value, no problem here.
File names are independent of mapid, could be any arbitrary 8-digit number (number from file name could be even written in TRE header, at offset 0xCF).
Names of subfiles like 12345678.TRE can be written in HEX, in this case you should use prefix 'I'. So 12345678.TRE is equivalent of I0BC614E.TRE. Mapid 140.531.712 could correspond to I8605800.TRE.
My first suggestion is to limit --fid to max 45744. For bigger FID option --mapid should be used.
Second proposition is to overlap mapid values for big FIDs, but with an offset of 1024. So for FID > 45744 formula could be like this:
mapid = (fid - 19792)*2048 + 6316032 + 1024
-- Best regards, Andrzej _______________________________________________ mkgmap-dev mailing list
mkgmap-dev@.org
-- Sent from: http://gis.19327.n8.nabble.com/Mkgmap-Development-f5324443.html
Hi Gerd, we can use any formula to assign mapid. If this formula would be widely accepted, there wouldn't be much conflicts between maps created by different developers. If we use the original Garmin's formula, then we would avoid conflicts with many existing Garmin's maps too. The only requirement would be to avoid FIDs already used by Garmin. As for my other divagation, it was about how to implement Garmin's formula. For full implementation, we would need some changes to mkgmap too. Without changes to mkgmap, we can implement forumla only for FIDs up to some limit. I have suggested, how this partial implementation could look like. -- Best regards, Andrzej
participants (5)
-
Andrzej Popowski -
brad -
Gerd Petermann -
osm＠pinns -
Thorsten Kukuk