area_size() function to be used in points file
Hello Gerd A while a go (30-9-2017) you made a patch to make the usage of the area_size() function available in the points-file through the mkgmap:cache_area_size value. I used it to only display names and symbols on bigger area’s to avoid cluttering. In the latest releases up to 4165 you mentioned fixes with ‘area_size()’ being not yet branched and some others which are branched. I lost the overview a little bit. I tested the special build of 30-9-2017 which does make the mkgmap:cache_area_size value available in the points file. (Example 1) and compared it to the mkgmap build 4165 (Example 2) which does not. Does this mean this feature is not branched or can I achieve a similar result else way? Thanx in advance Kind regards Joris mkgmap 30-9-2017 Example 1 Node generated from 29044917 [building=yes, jbprocessedpoints=true, leisure=sports_centre, mkgmap:admin_level2=NOR, mkgmap:admin_level4=Oslo, mkgmap:admin_level7=Oslo, mkgmap:admin_level9=Nordstrand, mkgmap:area2poi=true, mkgmap:cache_area_size=2658.2913818359375, mkgmap:country=NOR, name=Ekeberghallen] all Way 29044917 [building=yes, jbprocessedlines=true, leisure=sports_centre, mkgmap:admin_level2=NOR, mkgmap:admin_level4=Oslo, mkgmap:admin_level7=Oslo, mkgmap:admin_level9=Nordstrand, mkgmap:cache_area_size=2658.2913818359375, name=Ekeberghallen] all Way 29044917 [building=yes, jbprocessedlines=true, jbprocessedpolygons=true, leisure=sports_centre, mkgmap:admin_level2=NOR, mkgmap:admin_level4=Oslo, mkgmap:admin_level7=Oslo, mkgmap:admin_level9=Nordstrand, mkgmap:cache_area_size=2658.2913818359375, mkgmap:city=Nordstrand, mkgmap:country=NOR, mkgmap:region=Oslo, name=Ekeberghallen] all mkgmap-4165 Example 2 Node generated from 29044917 [building=yes, jbprocessedpoints=true, leisure=sports_centre, mkgmap:admin_level2=NOR, mkgmap:admin_level4=Oslo, mkgmap:admin_level7=Oslo, mkgmap:admin_level9=Nordstrand, mkgmap:area2poi=true, mkgmap:country=NOR, name=Ekeberghallen] all Way 29044917 [building=yes, jbprocessedlines=true, leisure=sports_centre, mkgmap:admin_level2=NOR, mkgmap:admin_level4=Oslo, mkgmap:admin_level7=Oslo, mkgmap:admin_level9=Nordstrand, mkgmap:cache_area_size=2658.2913818359375, name=Ekeberghallen] all Way 29044917 [building=yes, jbprocessedlines=true, jbprocessedpolygons=true, leisure=sports_centre, mkgmap:admin_level2=NOR, mkgmap:admin_level4=Oslo, mkgmap:admin_level7=Oslo, mkgmap:admin_level9=Nordstrand, mkgmap:cache_area_size=2658.2913818359375, mkgmap:city=Nordstrand, mkgmap:country=NOR, mkgmap:region=Oslo, name=Ekeberghallen] all
Hi Joris, the patch was for the use in the lines file, and I think I've committed it unchanged: http://www.mkgmap.org.uk/websvn/revision.php?repname=mkgmap&rev=4159 I have no idea why the generated POI for way 29044917 has different attributes, the patch was for way members of multipolygon relations. I don't even understand how the POI can have the mkgmap:cache_area_size tag. The tag is added by the code that handles multipolygons and by the style function area_size, which only works with ways. My current understanding is that the POI is generated before the evaluation of the style, so I don't see a way that the tag is assigned by mkgmap. I assume that you used the same input files and options for the above results? Gerd ________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Joris Bo <jorisbo@hotmail.com> Gesendet: Mittwoch, 18. April 2018 22:39:40 An: Development list for mkgmap Betreff: [mkgmap-dev] area_size() function to be used in points file Hello Gerd A while a go (30-9-2017) you made a patch to make the usage of the area_size() function available in the points-file through the mkgmap:cache_area_size value. I used it to only display names and symbols on bigger area’s to avoid cluttering. In the latest releases up to 4165 you mentioned fixes with ‘area_size()’ being not yet branched and some others which are branched. I lost the overview a little bit. I tested the special build of 30-9-2017 which does make the mkgmap:cache_area_size value available in the points file. (Example 1) and compared it to the mkgmap build 4165 (Example 2) which does not. Does this mean this feature is not branched or can I achieve a similar result else way? Thanx in advance Kind regards Joris mkgmap 30-9-2017 Example 1 Node generated from 29044917 [building=yes, jbprocessedpoints=true, leisure=sports_centre, mkgmap:admin_level2=NOR, mkgmap:admin_level4=Oslo, mkgmap:admin_level7=Oslo, mkgmap:admin_level9=Nordstrand, mkgmap:area2poi=true, mkgmap:cache_area_size=2658.2913818359375, mkgmap:country=NOR, name=Ekeberghallen] all Way 29044917 [building=yes, jbprocessedlines=true, leisure=sports_centre, mkgmap:admin_level2=NOR, mkgmap:admin_level4=Oslo, mkgmap:admin_level7=Oslo, mkgmap:admin_level9=Nordstrand, mkgmap:cache_area_size=2658.2913818359375, name=Ekeberghallen] all Way 29044917 [building=yes, jbprocessedlines=true, jbprocessedpolygons=true, leisure=sports_centre, mkgmap:admin_level2=NOR, mkgmap:admin_level4=Oslo, mkgmap:admin_level7=Oslo, mkgmap:admin_level9=Nordstrand, mkgmap:cache_area_size=2658.2913818359375, mkgmap:city=Nordstrand, mkgmap:country=NOR, mkgmap:region=Oslo, name=Ekeberghallen] all mkgmap-4165 Example 2 Node generated from 29044917 [building=yes, jbprocessedpoints=true, leisure=sports_centre, mkgmap:admin_level2=NOR, mkgmap:admin_level4=Oslo, mkgmap:admin_level7=Oslo, mkgmap:admin_level9=Nordstrand, mkgmap:area2poi=true, mkgmap:country=NOR, name=Ekeberghallen] all Way 29044917 [building=yes, jbprocessedlines=true, leisure=sports_centre, mkgmap:admin_level2=NOR, mkgmap:admin_level4=Oslo, mkgmap:admin_level7=Oslo, mkgmap:admin_level9=Nordstrand, mkgmap:cache_area_size=2658.2913818359375, name=Ekeberghallen] all Way 29044917 [building=yes, jbprocessedlines=true, jbprocessedpolygons=true, leisure=sports_centre, mkgmap:admin_level2=NOR, mkgmap:admin_level4=Oslo, mkgmap:admin_level7=Oslo, mkgmap:admin_level9=Nordstrand, mkgmap:cache_area_size=2658.2913818359375, mkgmap:city=Nordstrand, mkgmap:country=NOR, mkgmap:region=Oslo, name=Ekeberghallen] all
Hi Gerd, Yes, the only change in between was the mkmap.jar file. In both situations it looks like its processing the points file first, then the lines and then de polygons. I don't; know if that's the common order. In your post form 30-9 you wrote that you use a dirty trick which causes possible extra memory consumption. Probably for that reason it never ended up in the main branch? Gr Joris Your post from 30-9-2017 says ================================================================================================= [mkgmap-dev] mkgmap:area2poi in combination with area_size()
From Gerd Petermann GPetermann_muenchen at hotmail.com on Sat Sep 30 06:10:01 BST 2017 Hi Joris,
I see no easy way to calculate the value on demand, but the function area_size() sets a tag mkgmap:cache_area_size when called for the first time. The attached patch uses this feature for a rather dirty trick. It calls the functionn before the node for the POI is generated, therefore the new tag is also copied to the node. So, you can use something like mkgmap:area2poi=true & mkgmap:cache_area_size > 1000 {do something ...} Disadvantage: The area size is calculated for all polygons, and additional memory is needed to store the value as the tag is added to the polygon way and the POI. BTW: The current code calculates and stores the value like this: mkgmap:cache_area_size="45.771240234375" I assume it would be good enough to store the rounded value as mkgmap:cache_area_size="46" A compiled binary can be found here: http://files.mkgmap.org.uk/download/358/mkgmap.jar Please let me know if this works for you. Gerd ================================================================================================= -----Oorspronkelijk bericht----- Van: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> Namens Gerd Petermann Verzonden: woensdag 18 april 2018 23:13 Aan: Development list for mkgmap <mkgmap-dev@lists.mkgmap.org.uk> Onderwerp: Re: [mkgmap-dev] area_size() function to be used in points file Hi Joris, the patch was for the use in the lines file, and I think I've committed it unchanged: http://www.mkgmap.org.uk/websvn/revision.php?repname=mkgmap&rev=4159 I have no idea why the generated POI for way 29044917 has different attributes, the patch was for way members of multipolygon relations. I don't even understand how the POI can have the mkgmap:cache_area_size tag. The tag is added by the code that handles multipolygons and by the style function area_size, which only works with ways. My current understanding is that the POI is generated before the evaluation of the style, so I don't see a way that the tag is assigned by mkgmap. I assume that you used the same input files and options for the above results? Gerd ________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Joris Bo <jorisbo@hotmail.com> Gesendet: Mittwoch, 18. April 2018 22:39:40 An: Development list for mkgmap Betreff: [mkgmap-dev] area_size() function to be used in points file Hello Gerd A while a go (30-9-2017) you made a patch to make the usage of the area_size() function available in the points-file through the mkgmap:cache_area_size value. I used it to only display names and symbols on bigger area's to avoid cluttering. In the latest releases up to 4165 you mentioned fixes with 'area_size()' being not yet branched and some others which are branched. I lost the overview a little bit. I tested the special build of 30-9-2017 which does make the mkgmap:cache_area_size value available in the points file. (Example 1) and compared it to the mkgmap build 4165 (Example 2) which does not. Does this mean this feature is not branched or can I achieve a similar result else way? Thanx in advance Kind regards Joris mkgmap 30-9-2017 Example 1 (ok - mkgmap:cache_area_size available in poi) ----------------------------------------------------------------------------------------------------------------------------------- Node generated from 29044917 [building=yes, jbprocessedpoints=true, leisure=sports_centre, mkgmap:admin_level2=NOR, mkgmap:admin_level4=Oslo, mkgmap:admin_level7=Oslo, mkgmap:admin_level9=Nordstrand, mkgmap:area2poi=true, mkgmap:cache_area_size=2658.2913818359375, mkgmap:country=NOR, name=Ekeberghallen] all Way 29044917 [building=yes, jbprocessedlines=true, leisure=sports_centre, mkgmap:admin_level2=NOR, mkgmap:admin_level4=Oslo, mkgmap:admin_level7=Oslo, mkgmap:admin_level9=Nordstrand, mkgmap:cache_area_size=2658.2913818359375, name=Ekeberghallen] all Way 29044917 [building=yes, jbprocessedlines=true, jbprocessedpolygons=true, leisure=sports_centre, mkgmap:admin_level2=NOR, mkgmap:admin_level4=Oslo, mkgmap:admin_level7=Oslo, mkgmap:admin_level9=Nordstrand, mkgmap:cache_area_size=2658.2913818359375, mkgmap:city=Nordstrand, mkgmap:country=NOR, mkgmap:region=Oslo, name=Ekeberghallen] all mkgmap-4165 Example 2 ('not ok' - mkgmap:cache_area_size not available in poi anymore) ----------------------------------------------------------------------------------------------------------------------------------- Node generated from 29044917 [building=yes, jbprocessedpoints=true, leisure=sports_centre, mkgmap:admin_level2=NOR, mkgmap:admin_level4=Oslo, mkgmap:admin_level7=Oslo, mkgmap:admin_level9=Nordstrand, mkgmap:area2poi=true, mkgmap:country=NOR, name=Ekeberghallen] all Way 29044917 [building=yes, jbprocessedlines=true, leisure=sports_centre, mkgmap:admin_level2=NOR, mkgmap:admin_level4=Oslo, mkgmap:admin_level7=Oslo, mkgmap:admin_level9=Nordstrand, mkgmap:cache_area_size=2658.2913818359375, name=Ekeberghallen] all Way 29044917 [building=yes, jbprocessedlines=true, jbprocessedpolygons=true, leisure=sports_centre, mkgmap:admin_level2=NOR, mkgmap:admin_level4=Oslo, mkgmap:admin_level7=Oslo, mkgmap:admin_level9=Nordstrand, mkgmap:cache_area_size=2658.2913818359375, mkgmap:city=Nordstrand, mkgmap:country=NOR, mkgmap:region=Oslo, name=Ekeberghallen] all _______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
Hi Joris, ah, sorry, I did not notice the date, so I referred to the latest patch reg. mkgmap:cache_area_size, and I totally forgot about this older one. I'll have a closer look reg. the side effects of the patch today. Gerd ________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Joris Bo <jorisbo@hotmail.com> Gesendet: Donnerstag, 19. April 2018 07:20:44 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] area_size() function to be used in points file Hi Gerd, Yes, the only change in between was the mkmap.jar file. In both situations it looks like its processing the points file first, then the lines and then de polygons. I don't; know if that's the common order. In your post form 30-9 you wrote that you use a dirty trick which causes possible extra memory consumption. Probably for that reason it never ended up in the main branch? Gr Joris Your post from 30-9-2017 says ================================================================================================= [mkgmap-dev] mkgmap:area2poi in combination with area_size()
From Gerd Petermann GPetermann_muenchen at hotmail.com on Sat Sep 30 06:10:01 BST 2017 Hi Joris,
I see no easy way to calculate the value on demand, but the function area_size() sets a tag mkgmap:cache_area_size when called for the first time. The attached patch uses this feature for a rather dirty trick. It calls the functionn before the node for the POI is generated, therefore the new tag is also copied to the node. So, you can use something like mkgmap:area2poi=true & mkgmap:cache_area_size > 1000 {do something ...} Disadvantage: The area size is calculated for all polygons, and additional memory is needed to store the value as the tag is added to the polygon way and the POI. BTW: The current code calculates and stores the value like this: mkgmap:cache_area_size="45.771240234375" I assume it would be good enough to store the rounded value as mkgmap:cache_area_size="46" A compiled binary can be found here: http://files.mkgmap.org.uk/download/358/mkgmap.jar Please let me know if this works for you. Gerd ================================================================================================= -----Oorspronkelijk bericht----- Van: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> Namens Gerd Petermann Verzonden: woensdag 18 april 2018 23:13 Aan: Development list for mkgmap <mkgmap-dev@lists.mkgmap.org.uk> Onderwerp: Re: [mkgmap-dev] area_size() function to be used in points file Hi Joris, the patch was for the use in the lines file, and I think I've committed it unchanged: http://www.mkgmap.org.uk/websvn/revision.php?repname=mkgmap&rev=4159 I have no idea why the generated POI for way 29044917 has different attributes, the patch was for way members of multipolygon relations. I don't even understand how the POI can have the mkgmap:cache_area_size tag. The tag is added by the code that handles multipolygons and by the style function area_size, which only works with ways. My current understanding is that the POI is generated before the evaluation of the style, so I don't see a way that the tag is assigned by mkgmap. I assume that you used the same input files and options for the above results? Gerd ________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Joris Bo <jorisbo@hotmail.com> Gesendet: Mittwoch, 18. April 2018 22:39:40 An: Development list for mkgmap Betreff: [mkgmap-dev] area_size() function to be used in points file Hello Gerd A while a go (30-9-2017) you made a patch to make the usage of the area_size() function available in the points-file through the mkgmap:cache_area_size value. I used it to only display names and symbols on bigger area's to avoid cluttering. In the latest releases up to 4165 you mentioned fixes with 'area_size()' being not yet branched and some others which are branched. I lost the overview a little bit. I tested the special build of 30-9-2017 which does make the mkgmap:cache_area_size value available in the points file. (Example 1) and compared it to the mkgmap build 4165 (Example 2) which does not. Does this mean this feature is not branched or can I achieve a similar result else way? Thanx in advance Kind regards Joris mkgmap 30-9-2017 Example 1 (ok - mkgmap:cache_area_size available in poi) ----------------------------------------------------------------------------------------------------------------------------------- Node generated from 29044917 [building=yes, jbprocessedpoints=true, leisure=sports_centre, mkgmap:admin_level2=NOR, mkgmap:admin_level4=Oslo, mkgmap:admin_level7=Oslo, mkgmap:admin_level9=Nordstrand, mkgmap:area2poi=true, mkgmap:cache_area_size=2658.2913818359375, mkgmap:country=NOR, name=Ekeberghallen] all Way 29044917 [building=yes, jbprocessedlines=true, leisure=sports_centre, mkgmap:admin_level2=NOR, mkgmap:admin_level4=Oslo, mkgmap:admin_level7=Oslo, mkgmap:admin_level9=Nordstrand, mkgmap:cache_area_size=2658.2913818359375, name=Ekeberghallen] all Way 29044917 [building=yes, jbprocessedlines=true, jbprocessedpolygons=true, leisure=sports_centre, mkgmap:admin_level2=NOR, mkgmap:admin_level4=Oslo, mkgmap:admin_level7=Oslo, mkgmap:admin_level9=Nordstrand, mkgmap:cache_area_size=2658.2913818359375, mkgmap:city=Nordstrand, mkgmap:country=NOR, mkgmap:region=Oslo, name=Ekeberghallen] all mkgmap-4165 Example 2 ('not ok' - mkgmap:cache_area_size not available in poi anymore) ----------------------------------------------------------------------------------------------------------------------------------- Node generated from 29044917 [building=yes, jbprocessedpoints=true, leisure=sports_centre, mkgmap:admin_level2=NOR, mkgmap:admin_level4=Oslo, mkgmap:admin_level7=Oslo, mkgmap:admin_level9=Nordstrand, mkgmap:area2poi=true, mkgmap:country=NOR, name=Ekeberghallen] all Way 29044917 [building=yes, jbprocessedlines=true, leisure=sports_centre, mkgmap:admin_level2=NOR, mkgmap:admin_level4=Oslo, mkgmap:admin_level7=Oslo, mkgmap:admin_level9=Nordstrand, mkgmap:cache_area_size=2658.2913818359375, name=Ekeberghallen] all Way 29044917 [building=yes, jbprocessedlines=true, jbprocessedpolygons=true, leisure=sports_centre, mkgmap:admin_level2=NOR, mkgmap:admin_level4=Oslo, mkgmap:admin_level7=Oslo, mkgmap:admin_level9=Nordstrand, mkgmap:cache_area_size=2658.2913818359375, mkgmap:city=Nordstrand, mkgmap:country=NOR, mkgmap:region=Oslo, name=Ekeberghallen] all _______________________________________________ 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 all, the attached patch is an improved version of the poi_aera.patch from 2017-09-30. Like the first version it adds the tag mkgmap:cache_area_size to the POI generated for polygons, but now also for POI generated for multipolygons. It also changes the value in the tag so that it is rounded to 1/1000 , so instead of e.g. "1211.34554234542" you'll get "1211.346" when you evaluate style function area_size() or this special tag. I guess nobody is interested in the precise value, but I did not want to remove all decimals. For example small buildings can have values below 1. I think the additional computations are no problem and the rounding reduces the additional memory needed. A compiled binary is here: http://files.mkgmap.org.uk/download/431/mkgmap.jar If I hear no complains I'll commit this patch on sunday. Gerd ________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Gerd Petermann <gpetermann_muenchen@hotmail.com> Gesendet: Donnerstag, 19. April 2018 08:01:58 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] area_size() function to be used in points file Hi Joris, ah, sorry, I did not notice the date, so I referred to the latest patch reg. mkgmap:cache_area_size, and I totally forgot about this older one. I'll have a closer look reg. the side effects of the patch today. Gerd ________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Joris Bo <jorisbo@hotmail.com> Gesendet: Donnerstag, 19. April 2018 07:20:44 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] area_size() function to be used in points file Hi Gerd, Yes, the only change in between was the mkmap.jar file. In both situations it looks like its processing the points file first, then the lines and then de polygons. I don't; know if that's the common order. In your post form 30-9 you wrote that you use a dirty trick which causes possible extra memory consumption. Probably for that reason it never ended up in the main branch? Gr Joris Your post from 30-9-2017 says ================================================================================================= [mkgmap-dev] mkgmap:area2poi in combination with area_size()
From Gerd Petermann GPetermann_muenchen at hotmail.com on Sat Sep 30 06:10:01 BST 2017 Hi Joris,
I see no easy way to calculate the value on demand, but the function area_size() sets a tag mkgmap:cache_area_size when called for the first time. The attached patch uses this feature for a rather dirty trick. It calls the functionn before the node for the POI is generated, therefore the new tag is also copied to the node. So, you can use something like mkgmap:area2poi=true & mkgmap:cache_area_size > 1000 {do something ...} Disadvantage: The area size is calculated for all polygons, and additional memory is needed to store the value as the tag is added to the polygon way and the POI. BTW: The current code calculates and stores the value like this: mkgmap:cache_area_size="45.771240234375" I assume it would be good enough to store the rounded value as mkgmap:cache_area_size="46" A compiled binary can be found here: http://files.mkgmap.org.uk/download/358/mkgmap.jar Please let me know if this works for you. Gerd ================================================================================================= -----Oorspronkelijk bericht----- Van: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> Namens Gerd Petermann Verzonden: woensdag 18 april 2018 23:13 Aan: Development list for mkgmap <mkgmap-dev@lists.mkgmap.org.uk> Onderwerp: Re: [mkgmap-dev] area_size() function to be used in points file Hi Joris, the patch was for the use in the lines file, and I think I've committed it unchanged: http://www.mkgmap.org.uk/websvn/revision.php?repname=mkgmap&rev=4159 I have no idea why the generated POI for way 29044917 has different attributes, the patch was for way members of multipolygon relations. I don't even understand how the POI can have the mkgmap:cache_area_size tag. The tag is added by the code that handles multipolygons and by the style function area_size, which only works with ways. My current understanding is that the POI is generated before the evaluation of the style, so I don't see a way that the tag is assigned by mkgmap. I assume that you used the same input files and options for the above results? Gerd ________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Joris Bo <jorisbo@hotmail.com> Gesendet: Mittwoch, 18. April 2018 22:39:40 An: Development list for mkgmap Betreff: [mkgmap-dev] area_size() function to be used in points file Hello Gerd A while a go (30-9-2017) you made a patch to make the usage of the area_size() function available in the points-file through the mkgmap:cache_area_size value. I used it to only display names and symbols on bigger area's to avoid cluttering. In the latest releases up to 4165 you mentioned fixes with 'area_size()' being not yet branched and some others which are branched. I lost the overview a little bit. I tested the special build of 30-9-2017 which does make the mkgmap:cache_area_size value available in the points file. (Example 1) and compared it to the mkgmap build 4165 (Example 2) which does not. Does this mean this feature is not branched or can I achieve a similar result else way? Thanx in advance Kind regards Joris mkgmap 30-9-2017 Example 1 (ok - mkgmap:cache_area_size available in poi) ----------------------------------------------------------------------------------------------------------------------------------- Node generated from 29044917 [building=yes, jbprocessedpoints=true, leisure=sports_centre, mkgmap:admin_level2=NOR, mkgmap:admin_level4=Oslo, mkgmap:admin_level7=Oslo, mkgmap:admin_level9=Nordstrand, mkgmap:area2poi=true, mkgmap:cache_area_size=2658.2913818359375, mkgmap:country=NOR, name=Ekeberghallen] all Way 29044917 [building=yes, jbprocessedlines=true, leisure=sports_centre, mkgmap:admin_level2=NOR, mkgmap:admin_level4=Oslo, mkgmap:admin_level7=Oslo, mkgmap:admin_level9=Nordstrand, mkgmap:cache_area_size=2658.2913818359375, name=Ekeberghallen] all Way 29044917 [building=yes, jbprocessedlines=true, jbprocessedpolygons=true, leisure=sports_centre, mkgmap:admin_level2=NOR, mkgmap:admin_level4=Oslo, mkgmap:admin_level7=Oslo, mkgmap:admin_level9=Nordstrand, mkgmap:cache_area_size=2658.2913818359375, mkgmap:city=Nordstrand, mkgmap:country=NOR, mkgmap:region=Oslo, name=Ekeberghallen] all mkgmap-4165 Example 2 ('not ok' - mkgmap:cache_area_size not available in poi anymore) ----------------------------------------------------------------------------------------------------------------------------------- Node generated from 29044917 [building=yes, jbprocessedpoints=true, leisure=sports_centre, mkgmap:admin_level2=NOR, mkgmap:admin_level4=Oslo, mkgmap:admin_level7=Oslo, mkgmap:admin_level9=Nordstrand, mkgmap:area2poi=true, mkgmap:country=NOR, name=Ekeberghallen] all Way 29044917 [building=yes, jbprocessedlines=true, leisure=sports_centre, mkgmap:admin_level2=NOR, mkgmap:admin_level4=Oslo, mkgmap:admin_level7=Oslo, mkgmap:admin_level9=Nordstrand, mkgmap:cache_area_size=2658.2913818359375, name=Ekeberghallen] all Way 29044917 [building=yes, jbprocessedlines=true, jbprocessedpolygons=true, leisure=sports_centre, mkgmap:admin_level2=NOR, mkgmap:admin_level4=Oslo, mkgmap:admin_level7=Oslo, mkgmap:admin_level9=Nordstrand, mkgmap:cache_area_size=2658.2913818359375, mkgmap:city=Nordstrand, mkgmap:country=NOR, mkgmap:region=Oslo, name=Ekeberghallen] all _______________________________________________ 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
Hi Gerd, I tested this version and I do see a 'severe (subdivision) error', but the map is created successfully and area_size is available at poi-level. Does it make sense to already test larger maps or does this first needs finetuning? =========== Node generated from 29044917 [building=yes, jbprocessedpoints=true, leisure=sports_centre, mkgmap:admin_level2=NOR, mkgmap:admin_level4=Oslo, mkgmap:admin_level7=Oslo, mkgmap:admin_level9=Nordstrand, mkgmap:area2poi=true, mkgmap:cache_area_size=2658.291, mkgmap:country=NOR, name=Ekeberghallen] all Way 29044917 [building=yes, jbprocessedlines=true, leisure=sports_centre, mkgmap:admin_level2=NOR, mkgmap:admin_level4=Oslo, mkgmap:admin_level7=Oslo, mkgmap:admin_level9=Nordstrand, mkgmap:cache_area_size=2658.291, name=Ekeberghallen] all Way 29044917 [building=yes, jbprocessedlines=true, jbprocessedpolygons=true, leisure=sports_centre, mkgmap:admin_level2=NOR, mkgmap:admin_level4=Oslo, mkgmap:admin_level7=Oslo, mkgmap:admin_level9=Nordstrand, mkgmap:cache_area_size=2658.291, mkgmap:city=Nordstrand, mkgmap:country=NOR, mkgmap:region=Oslo, name=Ekeberghallen] all =========== SEVERE (Subdivision): C:\Temp\Garmin\8030 - JBM - TypTest\Output Splitter\80300001.osm.pbf: Subdivision neg width 0 min= 495438 mid= 495438 max= 495439 shift= 0 mask= 0 @ 2789482/495438 SEVERE (Subdivision): C:\Temp\Garmin\8030 - JBM - TypTest\Output Splitter\80300001.osm.pbf: Subdivision neg height 0 min= 2789482 mid= 2789482 max= 2789483 shift= 0 mask= 0 @ 2789482/495438 =========== -----Oorspronkelijk bericht----- Van: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> Namens Gerd Petermann Verzonden: donderdag 19 april 2018 09:09 Aan: Development list for mkgmap <mkgmap-dev@lists.mkgmap.org.uk> Onderwerp: Re: [mkgmap-dev] area_size() function to be used in points file Hi all, the attached patch is an improved version of the poi_aera.patch from 2017-09-30. Like the first version it adds the tag mkgmap:cache_area_size to the POI generated for polygons, but now also for POI generated for multipolygons. It also changes the value in the tag so that it is rounded to 1/1000 , so instead of e.g. "1211.34554234542" you'll get "1211.346" when you evaluate style function area_size() or this special tag. I guess nobody is interested in the precise value, but I did not want to remove all decimals. For example small buildings can have values below 1. I think the additional computations are no problem and the rounding reduces the additional memory needed. A compiled binary is here: http://files.mkgmap.org.uk/download/431/mkgmap.jar If I hear no complains I'll commit this patch on sunday. Gerd ________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Gerd Petermann <gpetermann_muenchen@hotmail.com> Gesendet: Donnerstag, 19. April 2018 08:01:58 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] area_size() function to be used in points file Hi Joris, ah, sorry, I did not notice the date, so I referred to the latest patch reg. mkgmap:cache_area_size, and I totally forgot about this older one. I'll have a closer look reg. the side effects of the patch today. Gerd ________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Joris Bo <jorisbo@hotmail.com> Gesendet: Donnerstag, 19. April 2018 07:20:44 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] area_size() function to be used in points file Hi Gerd, Yes, the only change in between was the mkmap.jar file. In both situations it looks like its processing the points file first, then the lines and then de polygons. I don't; know if that's the common order. In your post form 30-9 you wrote that you use a dirty trick which causes possible extra memory consumption. Probably for that reason it never ended up in the main branch? Gr Joris Your post from 30-9-2017 says ================================================================================================= [mkgmap-dev] mkgmap:area2poi in combination with area_size() From Gerd Petermann GPetermann_muenchen at hotmail.com on Sat Sep 30 06:10:01 BST 2017 Hi Joris, I see no easy way to calculate the value on demand, but the function area_size() sets a tag mkgmap:cache_area_size when called for the first time. The attached patch uses this feature for a rather dirty trick. It calls the functionn before the node for the POI is generated, therefore the new tag is also copied to the node. So, you can use something like mkgmap:area2poi=true & mkgmap:cache_area_size > 1000 {do something ...} Disadvantage: The area size is calculated for all polygons, and additional memory is needed to store the value as the tag is added to the polygon way and the POI. BTW: The current code calculates and stores the value like this: mkgmap:cache_area_size="45.771240234375" I assume it would be good enough to store the rounded value as mkgmap:cache_area_size="46" A compiled binary can be found here: http://files.mkgmap.org.uk/download/358/mkgmap.jar Please let me know if this works for you. Gerd ================================================================================================= -----Oorspronkelijk bericht----- Van: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> Namens Gerd Petermann Verzonden: woensdag 18 april 2018 23:13 Aan: Development list for mkgmap <mkgmap-dev@lists.mkgmap.org.uk> Onderwerp: Re: [mkgmap-dev] area_size() function to be used in points file Hi Joris, the patch was for the use in the lines file, and I think I've committed it unchanged: http://www.mkgmap.org.uk/websvn/revision.php?repname=mkgmap&rev=4159 I have no idea why the generated POI for way 29044917 has different attributes, the patch was for way members of multipolygon relations. I don't even understand how the POI can have the mkgmap:cache_area_size tag. The tag is added by the code that handles multipolygons and by the style function area_size, which only works with ways. My current understanding is that the POI is generated before the evaluation of the style, so I don't see a way that the tag is assigned by mkgmap. I assume that you used the same input files and options for the above results? Gerd ________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Joris Bo <jorisbo@hotmail.com> Gesendet: Mittwoch, 18. April 2018 22:39:40 An: Development list for mkgmap Betreff: [mkgmap-dev] area_size() function to be used in points file Hello Gerd A while a go (30-9-2017) you made a patch to make the usage of the area_size() function available in the points-file through the mkgmap:cache_area_size value. I used it to only display names and symbols on bigger area's to avoid cluttering. In the latest releases up to 4165 you mentioned fixes with 'area_size()' being not yet branched and some others which are branched. I lost the overview a little bit. I tested the special build of 30-9-2017 which does make the mkgmap:cache_area_size value available in the points file. (Example 1) and compared it to the mkgmap build 4165 (Example 2) which does not. Does this mean this feature is not branched or can I achieve a similar result else way? Thanx in advance Kind regards Joris mkgmap 30-9-2017 Example 1 (ok - mkgmap:cache_area_size available in poi) ----------------------------------------------------------------------------------------------------------------------------------- Node generated from 29044917 [building=yes, jbprocessedpoints=true, leisure=sports_centre, mkgmap:admin_level2=NOR, mkgmap:admin_level4=Oslo, mkgmap:admin_level7=Oslo, mkgmap:admin_level9=Nordstrand, mkgmap:area2poi=true, mkgmap:cache_area_size=2658.2913818359375, mkgmap:country=NOR, name=Ekeberghallen] all Way 29044917 [building=yes, jbprocessedlines=true, leisure=sports_centre, mkgmap:admin_level2=NOR, mkgmap:admin_level4=Oslo, mkgmap:admin_level7=Oslo, mkgmap:admin_level9=Nordstrand, mkgmap:cache_area_size=2658.2913818359375, name=Ekeberghallen] all Way 29044917 [building=yes, jbprocessedlines=true, jbprocessedpolygons=true, leisure=sports_centre, mkgmap:admin_level2=NOR, mkgmap:admin_level4=Oslo, mkgmap:admin_level7=Oslo, mkgmap:admin_level9=Nordstrand, mkgmap:cache_area_size=2658.2913818359375, mkgmap:city=Nordstrand, mkgmap:country=NOR, mkgmap:region=Oslo, name=Ekeberghallen] all mkgmap-4165 Example 2 ('not ok' - mkgmap:cache_area_size not available in poi anymore) ----------------------------------------------------------------------------------------------------------------------------------- Node generated from 29044917 [building=yes, jbprocessedpoints=true, leisure=sports_centre, mkgmap:admin_level2=NOR, mkgmap:admin_level4=Oslo, mkgmap:admin_level7=Oslo, mkgmap:admin_level9=Nordstrand, mkgmap:area2poi=true, mkgmap:country=NOR, name=Ekeberghallen] all Way 29044917 [building=yes, jbprocessedlines=true, leisure=sports_centre, mkgmap:admin_level2=NOR, mkgmap:admin_level4=Oslo, mkgmap:admin_level7=Oslo, mkgmap:admin_level9=Nordstrand, mkgmap:cache_area_size=2658.2913818359375, name=Ekeberghallen] all Way 29044917 [building=yes, jbprocessedlines=true, jbprocessedpolygons=true, leisure=sports_centre, mkgmap:admin_level2=NOR, mkgmap:admin_level4=Oslo, mkgmap:admin_level7=Oslo, mkgmap:admin_level9=Nordstrand, mkgmap:cache_area_size=2658.2913818359375, mkgmap:city=Nordstrand, mkgmap:country=NOR, mkgmap:region=Oslo, name=Ekeberghallen] all _______________________________________________ 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
Hi Joris, thanks for testing. The severe messages were introduced with the latest patches from Ticker: http://www.mkgmap.org.uk/websvn/revision.php?repname=mkgmap&rev=4167 The added source in Subdivision.java: if (w <= 0) { // think can get this from empty/unbounded maps log.error("Subdivision neg width ", w, "min=", area.getMinLong(), "mid=", longitude, "max=", area.getMaxLong(), "shift=", shift, "mask=", mask, "@", getCenter()); w = 0x7fff; } @Ticker: I don't know why we have those new messages? Gerd ________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Joris Bo <jorisbo@hotmail.com> Gesendet: Donnerstag, 19. April 2018 09:31:40 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] area_size() function to be used in points file Hi Gerd, I tested this version and I do see a 'severe (subdivision) error', but the map is created successfully and area_size is available at poi-level. Does it make sense to already test larger maps or does this first needs finetuning? =========== Node generated from 29044917 [building=yes, jbprocessedpoints=true, leisure=sports_centre, mkgmap:admin_level2=NOR, mkgmap:admin_level4=Oslo, mkgmap:admin_level7=Oslo, mkgmap:admin_level9=Nordstrand, mkgmap:area2poi=true, mkgmap:cache_area_size=2658.291, mkgmap:country=NOR, name=Ekeberghallen] all Way 29044917 [building=yes, jbprocessedlines=true, leisure=sports_centre, mkgmap:admin_level2=NOR, mkgmap:admin_level4=Oslo, mkgmap:admin_level7=Oslo, mkgmap:admin_level9=Nordstrand, mkgmap:cache_area_size=2658.291, name=Ekeberghallen] all Way 29044917 [building=yes, jbprocessedlines=true, jbprocessedpolygons=true, leisure=sports_centre, mkgmap:admin_level2=NOR, mkgmap:admin_level4=Oslo, mkgmap:admin_level7=Oslo, mkgmap:admin_level9=Nordstrand, mkgmap:cache_area_size=2658.291, mkgmap:city=Nordstrand, mkgmap:country=NOR, mkgmap:region=Oslo, name=Ekeberghallen] all =========== SEVERE (Subdivision): C:\Temp\Garmin\8030 - JBM - TypTest\Output Splitter\80300001.osm.pbf: Subdivision neg width 0 min= 495438 mid= 495438 max= 495439 shift= 0 mask= 0 @ 2789482/495438 SEVERE (Subdivision): C:\Temp\Garmin\8030 - JBM - TypTest\Output Splitter\80300001.osm.pbf: Subdivision neg height 0 min= 2789482 mid= 2789482 max= 2789483 shift= 0 mask= 0 @ 2789482/495438 =========== -----Oorspronkelijk bericht----- Van: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> Namens Gerd Petermann Verzonden: donderdag 19 april 2018 09:09 Aan: Development list for mkgmap <mkgmap-dev@lists.mkgmap.org.uk> Onderwerp: Re: [mkgmap-dev] area_size() function to be used in points file Hi all, the attached patch is an improved version of the poi_aera.patch from 2017-09-30. Like the first version it adds the tag mkgmap:cache_area_size to the POI generated for polygons, but now also for POI generated for multipolygons. It also changes the value in the tag so that it is rounded to 1/1000 , so instead of e.g. "1211.34554234542" you'll get "1211.346" when you evaluate style function area_size() or this special tag. I guess nobody is interested in the precise value, but I did not want to remove all decimals. For example small buildings can have values below 1. I think the additional computations are no problem and the rounding reduces the additional memory needed. A compiled binary is here: http://files.mkgmap.org.uk/download/431/mkgmap.jar If I hear no complains I'll commit this patch on sunday. Gerd ________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Gerd Petermann <gpetermann_muenchen@hotmail.com> Gesendet: Donnerstag, 19. April 2018 08:01:58 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] area_size() function to be used in points file Hi Joris, ah, sorry, I did not notice the date, so I referred to the latest patch reg. mkgmap:cache_area_size, and I totally forgot about this older one. I'll have a closer look reg. the side effects of the patch today. Gerd ________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Joris Bo <jorisbo@hotmail.com> Gesendet: Donnerstag, 19. April 2018 07:20:44 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] area_size() function to be used in points file Hi Gerd, Yes, the only change in between was the mkmap.jar file. In both situations it looks like its processing the points file first, then the lines and then de polygons. I don't; know if that's the common order. In your post form 30-9 you wrote that you use a dirty trick which causes possible extra memory consumption. Probably for that reason it never ended up in the main branch? Gr Joris Your post from 30-9-2017 says ================================================================================================= [mkgmap-dev] mkgmap:area2poi in combination with area_size() From Gerd Petermann GPetermann_muenchen at hotmail.com on Sat Sep 30 06:10:01 BST 2017 Hi Joris, I see no easy way to calculate the value on demand, but the function area_size() sets a tag mkgmap:cache_area_size when called for the first time. The attached patch uses this feature for a rather dirty trick. It calls the functionn before the node for the POI is generated, therefore the new tag is also copied to the node. So, you can use something like mkgmap:area2poi=true & mkgmap:cache_area_size > 1000 {do something ...} Disadvantage: The area size is calculated for all polygons, and additional memory is needed to store the value as the tag is added to the polygon way and the POI. BTW: The current code calculates and stores the value like this: mkgmap:cache_area_size="45.771240234375" I assume it would be good enough to store the rounded value as mkgmap:cache_area_size="46" A compiled binary can be found here: http://files.mkgmap.org.uk/download/358/mkgmap.jar Please let me know if this works for you. Gerd ================================================================================================= -----Oorspronkelijk bericht----- Van: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> Namens Gerd Petermann Verzonden: woensdag 18 april 2018 23:13 Aan: Development list for mkgmap <mkgmap-dev@lists.mkgmap.org.uk> Onderwerp: Re: [mkgmap-dev] area_size() function to be used in points file Hi Joris, the patch was for the use in the lines file, and I think I've committed it unchanged: http://www.mkgmap.org.uk/websvn/revision.php?repname=mkgmap&rev=4159 I have no idea why the generated POI for way 29044917 has different attributes, the patch was for way members of multipolygon relations. I don't even understand how the POI can have the mkgmap:cache_area_size tag. The tag is added by the code that handles multipolygons and by the style function area_size, which only works with ways. My current understanding is that the POI is generated before the evaluation of the style, so I don't see a way that the tag is assigned by mkgmap. I assume that you used the same input files and options for the above results? Gerd ________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Joris Bo <jorisbo@hotmail.com> Gesendet: Mittwoch, 18. April 2018 22:39:40 An: Development list for mkgmap Betreff: [mkgmap-dev] area_size() function to be used in points file Hello Gerd A while a go (30-9-2017) you made a patch to make the usage of the area_size() function available in the points-file through the mkgmap:cache_area_size value. I used it to only display names and symbols on bigger area's to avoid cluttering. In the latest releases up to 4165 you mentioned fixes with 'area_size()' being not yet branched and some others which are branched. I lost the overview a little bit. I tested the special build of 30-9-2017 which does make the mkgmap:cache_area_size value available in the points file. (Example 1) and compared it to the mkgmap build 4165 (Example 2) which does not. Does this mean this feature is not branched or can I achieve a similar result else way? Thanx in advance Kind regards Joris mkgmap 30-9-2017 Example 1 (ok - mkgmap:cache_area_size available in poi) ----------------------------------------------------------------------------------------------------------------------------------- Node generated from 29044917 [building=yes, jbprocessedpoints=true, leisure=sports_centre, mkgmap:admin_level2=NOR, mkgmap:admin_level4=Oslo, mkgmap:admin_level7=Oslo, mkgmap:admin_level9=Nordstrand, mkgmap:area2poi=true, mkgmap:cache_area_size=2658.2913818359375, mkgmap:country=NOR, name=Ekeberghallen] all Way 29044917 [building=yes, jbprocessedlines=true, leisure=sports_centre, mkgmap:admin_level2=NOR, mkgmap:admin_level4=Oslo, mkgmap:admin_level7=Oslo, mkgmap:admin_level9=Nordstrand, mkgmap:cache_area_size=2658.2913818359375, name=Ekeberghallen] all Way 29044917 [building=yes, jbprocessedlines=true, jbprocessedpolygons=true, leisure=sports_centre, mkgmap:admin_level2=NOR, mkgmap:admin_level4=Oslo, mkgmap:admin_level7=Oslo, mkgmap:admin_level9=Nordstrand, mkgmap:cache_area_size=2658.2913818359375, mkgmap:city=Nordstrand, mkgmap:country=NOR, mkgmap:region=Oslo, name=Ekeberghallen] all mkgmap-4165 Example 2 ('not ok' - mkgmap:cache_area_size not available in poi anymore) ----------------------------------------------------------------------------------------------------------------------------------- Node generated from 29044917 [building=yes, jbprocessedpoints=true, leisure=sports_centre, mkgmap:admin_level2=NOR, mkgmap:admin_level4=Oslo, mkgmap:admin_level7=Oslo, mkgmap:admin_level9=Nordstrand, mkgmap:area2poi=true, mkgmap:country=NOR, name=Ekeberghallen] all Way 29044917 [building=yes, jbprocessedlines=true, leisure=sports_centre, mkgmap:admin_level2=NOR, mkgmap:admin_level4=Oslo, mkgmap:admin_level7=Oslo, mkgmap:admin_level9=Nordstrand, mkgmap:cache_area_size=2658.2913818359375, name=Ekeberghallen] all Way 29044917 [building=yes, jbprocessedlines=true, jbprocessedpolygons=true, leisure=sports_centre, mkgmap:admin_level2=NOR, mkgmap:admin_level4=Oslo, mkgmap:admin_level7=Oslo, mkgmap:admin_level9=Nordstrand, mkgmap:cache_area_size=2658.2913818359375, mkgmap:city=Nordstrand, mkgmap:country=NOR, mkgmap:region=Oslo, name=Ekeberghallen] all _______________________________________________ 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
I was being a bit too severe. Zero and negative width/height subdivisions previously were allowed through without being noticed, with the negative quietly overflowing looking positive. The attached patch downgrades the message for zero size to a warning but flags an error for neg. Regards Ticker On Thu, 2018-04-19 at 09:49 +0000, Gerd Petermann wrote:
Hi Joris,
thanks for testing. The severe messages were introduced with the latest patches from Ticker: http://www.mkgmap.org.uk/websvn/revision.php?repname=mkgmap&rev=4167
The added source in Subdivision.java: if (w <= 0) { // think can get this from empty/unbounded maps log.error("Subdivision neg width ", w, "min=", area.getMinLong(), "mid=", longitude, "max=", area.getMaxLong(), "shift=", shift, "mask=", mask, "@", getCenter()); w = 0x7fff; }
@Ticker: I don't know why we have those new messages?
Gerd
Hi Ticker, thanks for the quick patch. I don't know if a height or widht value of 0 is OK or not. If it is I see no reason for a warning, if not it should be handled somewhere else. @Steve: Do you know more? Gerd ________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Ticker Berkin <rwb-mkgmap@jagit.co.uk> Gesendet: Donnerstag, 19. April 2018 12:46:43 An: mkgmap-dev@lists.mkgmap.org.uk Betreff: Re: [mkgmap-dev] area_size() function to be used in points file I was being a bit too severe. Zero and negative width/height subdivisions previously were allowed through without being noticed, with the negative quietly overflowing looking positive. The attached patch downgrades the message for zero size to a warning but flags an error for neg. Regards Ticker On Thu, 2018-04-19 at 09:49 +0000, Gerd Petermann wrote:
Hi Joris,
thanks for testing. The severe messages were introduced with the latest patches from Ticker: http://www.mkgmap.org.uk/websvn/revision.php?repname=mkgmap&rev=4167
The added source in Subdivision.java: if (w <= 0) { // think can get this from empty/unbounded maps log.error("Subdivision neg width ", w, "min=", area.getMinLong(), "mid=", longitude, "max=", area.getMaxLong(), "shift=", shift, "mask=", mask, "@", getCenter()); w = 0x7fff; }
@Ticker: I don't know why we have those new messages?
Gerd
Hi I tested some bigger areas like Benelux and Scandinavia and besides the severe warnings it works fine. Thanks again Gr Joris -----Oorspronkelijk bericht----- Van: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> Namens Gerd Petermann Verzonden: donderdag 19 april 2018 11:50 Aan: Development list for mkgmap <mkgmap-dev@lists.mkgmap.org.uk> Onderwerp: Re: [mkgmap-dev] area_size() function to be used in points file Hi Joris, thanks for testing. The severe messages were introduced with the latest patches from Ticker: http://www.mkgmap.org.uk/websvn/revision.php?repname=mkgmap&rev=4167 The added source in Subdivision.java: if (w <= 0) { // think can get this from empty/unbounded maps log.error("Subdivision neg width ", w, "min=", area.getMinLong(), "mid=", longitude, "max=", area.getMaxLong(), "shift=", shift, "mask=", mask, "@", getCenter()); w = 0x7fff; } @Ticker: I don't know why we have those new messages? Gerd ________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Joris Bo <jorisbo@hotmail.com> Gesendet: Donnerstag, 19. April 2018 09:31:40 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] area_size() function to be used in points file Hi Gerd, I tested this version and I do see a 'severe (subdivision) error', but the map is created successfully and area_size is available at poi-level. Does it make sense to already test larger maps or does this first needs finetuning? =========== Node generated from 29044917 [building=yes, jbprocessedpoints=true, leisure=sports_centre, mkgmap:admin_level2=NOR, mkgmap:admin_level4=Oslo, mkgmap:admin_level7=Oslo, mkgmap:admin_level9=Nordstrand, mkgmap:area2poi=true, mkgmap:cache_area_size=2658.291, mkgmap:country=NOR, name=Ekeberghallen] all Way 29044917 [building=yes, jbprocessedlines=true, leisure=sports_centre, mkgmap:admin_level2=NOR, mkgmap:admin_level4=Oslo, mkgmap:admin_level7=Oslo, mkgmap:admin_level9=Nordstrand, mkgmap:cache_area_size=2658.291, name=Ekeberghallen] all Way 29044917 [building=yes, jbprocessedlines=true, jbprocessedpolygons=true, leisure=sports_centre, mkgmap:admin_level2=NOR, mkgmap:admin_level4=Oslo, mkgmap:admin_level7=Oslo, mkgmap:admin_level9=Nordstrand, mkgmap:cache_area_size=2658.291, mkgmap:city=Nordstrand, mkgmap:country=NOR, mkgmap:region=Oslo, name=Ekeberghallen] all =========== SEVERE (Subdivision): C:\Temp\Garmin\8030 - JBM - TypTest\Output Splitter\80300001.osm.pbf: Subdivision neg width 0 min= 495438 mid= 495438 max= 495439 shift= 0 mask= 0 @ 2789482/495438 SEVERE (Subdivision): C:\Temp\Garmin\8030 - JBM - TypTest\Output Splitter\80300001.osm.pbf: Subdivision neg height 0 min= 2789482 mid= 2789482 max= 2789483 shift= 0 mask= 0 @ 2789482/495438 =========== -----Oorspronkelijk bericht----- Van: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> Namens Gerd Petermann Verzonden: donderdag 19 april 2018 09:09 Aan: Development list for mkgmap <mkgmap-dev@lists.mkgmap.org.uk> Onderwerp: Re: [mkgmap-dev] area_size() function to be used in points file Hi all, the attached patch is an improved version of the poi_aera.patch from 2017-09-30. Like the first version it adds the tag mkgmap:cache_area_size to the POI generated for polygons, but now also for POI generated for multipolygons. It also changes the value in the tag so that it is rounded to 1/1000 , so instead of e.g. "1211.34554234542" you'll get "1211.346" when you evaluate style function area_size() or this special tag. I guess nobody is interested in the precise value, but I did not want to remove all decimals. For example small buildings can have values below 1. I think the additional computations are no problem and the rounding reduces the additional memory needed. A compiled binary is here: http://files.mkgmap.org.uk/download/431/mkgmap.jar If I hear no complains I'll commit this patch on sunday. Gerd ________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Gerd Petermann <gpetermann_muenchen@hotmail.com> Gesendet: Donnerstag, 19. April 2018 08:01:58 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] area_size() function to be used in points file Hi Joris, ah, sorry, I did not notice the date, so I referred to the latest patch reg. mkgmap:cache_area_size, and I totally forgot about this older one. I'll have a closer look reg. the side effects of the patch today. Gerd ________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Joris Bo <jorisbo@hotmail.com> Gesendet: Donnerstag, 19. April 2018 07:20:44 An: Development list for mkgmap Betreff: Re: [mkgmap-dev] area_size() function to be used in points file Hi Gerd, Yes, the only change in between was the mkmap.jar file. In both situations it looks like its processing the points file first, then the lines and then de polygons. I don't; know if that's the common order. In your post form 30-9 you wrote that you use a dirty trick which causes possible extra memory consumption. Probably for that reason it never ended up in the main branch? Gr Joris Your post from 30-9-2017 says ================================================================================================= [mkgmap-dev] mkgmap:area2poi in combination with area_size() From Gerd Petermann GPetermann_muenchen at hotmail.com on Sat Sep 30 06:10:01 BST 2017 Hi Joris, I see no easy way to calculate the value on demand, but the function area_size() sets a tag mkgmap:cache_area_size when called for the first time. The attached patch uses this feature for a rather dirty trick. It calls the functionn before the node for the POI is generated, therefore the new tag is also copied to the node. So, you can use something like mkgmap:area2poi=true & mkgmap:cache_area_size > 1000 {do something ...} Disadvantage: The area size is calculated for all polygons, and additional memory is needed to store the value as the tag is added to the polygon way and the POI. BTW: The current code calculates and stores the value like this: mkgmap:cache_area_size="45.771240234375" I assume it would be good enough to store the rounded value as mkgmap:cache_area_size="46" A compiled binary can be found here: http://files.mkgmap.org.uk/download/358/mkgmap.jar Please let me know if this works for you. Gerd ================================================================================================= -----Oorspronkelijk bericht----- Van: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> Namens Gerd Petermann Verzonden: woensdag 18 april 2018 23:13 Aan: Development list for mkgmap <mkgmap-dev@lists.mkgmap.org.uk> Onderwerp: Re: [mkgmap-dev] area_size() function to be used in points file Hi Joris, the patch was for the use in the lines file, and I think I've committed it unchanged: http://www.mkgmap.org.uk/websvn/revision.php?repname=mkgmap&rev=4159 I have no idea why the generated POI for way 29044917 has different attributes, the patch was for way members of multipolygon relations. I don't even understand how the POI can have the mkgmap:cache_area_size tag. The tag is added by the code that handles multipolygons and by the style function area_size, which only works with ways. My current understanding is that the POI is generated before the evaluation of the style, so I don't see a way that the tag is assigned by mkgmap. I assume that you used the same input files and options for the above results? Gerd ________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Joris Bo <jorisbo@hotmail.com> Gesendet: Mittwoch, 18. April 2018 22:39:40 An: Development list for mkgmap Betreff: [mkgmap-dev] area_size() function to be used in points file Hello Gerd A while a go (30-9-2017) you made a patch to make the usage of the area_size() function available in the points-file through the mkgmap:cache_area_size value. I used it to only display names and symbols on bigger area's to avoid cluttering. In the latest releases up to 4165 you mentioned fixes with 'area_size()' being not yet branched and some others which are branched. I lost the overview a little bit. I tested the special build of 30-9-2017 which does make the mkgmap:cache_area_size value available in the points file. (Example 1) and compared it to the mkgmap build 4165 (Example 2) which does not. Does this mean this feature is not branched or can I achieve a similar result else way? Thanx in advance Kind regards Joris mkgmap 30-9-2017 Example 1 (ok - mkgmap:cache_area_size available in poi) ----------------------------------------------------------------------------------------------------------------------------------- Node generated from 29044917 [building=yes, jbprocessedpoints=true, leisure=sports_centre, mkgmap:admin_level2=NOR, mkgmap:admin_level4=Oslo, mkgmap:admin_level7=Oslo, mkgmap:admin_level9=Nordstrand, mkgmap:area2poi=true, mkgmap:cache_area_size=2658.2913818359375, mkgmap:country=NOR, name=Ekeberghallen] all Way 29044917 [building=yes, jbprocessedlines=true, leisure=sports_centre, mkgmap:admin_level2=NOR, mkgmap:admin_level4=Oslo, mkgmap:admin_level7=Oslo, mkgmap:admin_level9=Nordstrand, mkgmap:cache_area_size=2658.2913818359375, name=Ekeberghallen] all Way 29044917 [building=yes, jbprocessedlines=true, jbprocessedpolygons=true, leisure=sports_centre, mkgmap:admin_level2=NOR, mkgmap:admin_level4=Oslo, mkgmap:admin_level7=Oslo, mkgmap:admin_level9=Nordstrand, mkgmap:cache_area_size=2658.2913818359375, mkgmap:city=Nordstrand, mkgmap:country=NOR, mkgmap:region=Oslo, name=Ekeberghallen] all mkgmap-4165 Example 2 ('not ok' - mkgmap:cache_area_size not available in poi anymore) ----------------------------------------------------------------------------------------------------------------------------------- Node generated from 29044917 [building=yes, jbprocessedpoints=true, leisure=sports_centre, mkgmap:admin_level2=NOR, mkgmap:admin_level4=Oslo, mkgmap:admin_level7=Oslo, mkgmap:admin_level9=Nordstrand, mkgmap:area2poi=true, mkgmap:country=NOR, name=Ekeberghallen] all Way 29044917 [building=yes, jbprocessedlines=true, leisure=sports_centre, mkgmap:admin_level2=NOR, mkgmap:admin_level4=Oslo, mkgmap:admin_level7=Oslo, mkgmap:admin_level9=Nordstrand, mkgmap:cache_area_size=2658.2913818359375, name=Ekeberghallen] all Way 29044917 [building=yes, jbprocessedlines=true, jbprocessedpolygons=true, leisure=sports_centre, mkgmap:admin_level2=NOR, mkgmap:admin_level4=Oslo, mkgmap:admin_level7=Oslo, mkgmap:admin_level9=Nordstrand, mkgmap:cache_area_size=2658.2913818359375, mkgmap:city=Nordstrand, mkgmap:country=NOR, mkgmap:region=Oslo, name=Ekeberghallen] all _______________________________________________ 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 mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
participants (3)
-
Gerd Petermann -
Joris Bo -
Ticker Berkin