Comments on: Making a Big OSM Map http://crschmidt.net/blog/archives/362/making-a-big-osm-map/ Ramblings of a GIS Hacker Tue, 16 Mar 2010 22:56:59 +0000 http://wordpress.org/?v=2.7.1 hourly 1 By: Valent http://crschmidt.net/blog/archives/362/making-a-big-osm-map/comment-page-1/#comment-20514 Valent Fri, 06 Nov 2009 14:49:33 +0000 http://crschmidt.net/blog/?p=362#comment-20514 Whoow that is a gorgeous map! I'm so jealous :) You wrote that it is trivial tu make big image using Mapnik but for us who have no experience using Mapnik some pointers would be a huge help so that we can also make big images for our Mapping Parties. Cheers from Croatia. Whoow that is a gorgeous map! I’m so jealous :)

You wrote that it is trivial tu make big image using Mapnik but for us who have no experience using Mapnik some pointers would be a huge help so that we can also make big images for our Mapping Parties.

Cheers from Croatia.

]]> By: Michal Migurski http://crschmidt.net/blog/archives/362/making-a-big-osm-map/comment-page-1/#comment-20207 Michal Migurski Thu, 12 Feb 2009 20:21:23 +0000 http://crschmidt.net/blog/?p=362#comment-20207 I finally compiled mapnik on Debian for the first time recently - I had to compile Boost from source in order to get it to work, I think there are some notes from this process I posted to mapnik-users when it finally worked. I finally compiled mapnik on Debian for the first time recently - I had to compile Boost from source in order to get it to work, I think there are some notes from this process I posted to mapnik-users when it finally worked.

]]> By: crschmidt http://crschmidt.net/blog/archives/362/making-a-big-osm-map/comment-page-1/#comment-20206 crschmidt Thu, 12 Feb 2009 19:22:14 +0000 http://crschmidt.net/blog/?p=362#comment-20206 1. I couldn't get Mapnik to build. I've never built Mapnik under Debian Stable; I use Debian packages, but there hasn't been a release since Cairo support was added. So I didn't have the ability to render to a Cairo surface. 2. Since I was 'printing' the image at 600dpi natively, I don't think this would have changed anything. Had I, instead, printed at 90dpi and let the printer resize it, it probably would have worked, but I didn't realize that what I had wouldn't work until after it was already done. 3. "printing" that area at 90dpi or what have you would have left out a lot of the details that I wanted to see; Since the goal was really to get a coverage map, I wanted to see everything, and 'zooming out' by a factor of 6 leaves out a lot of the lower level names and so on. In this case, I couldn't see them anyway, but with magnification of some kind I would. 4. The printer doesn't cleanly support all versions of PDF, so I wasn't sure that printing to a Cairo PDF would work. 5. Since rasterizing to png required 3GB of memory, I'm not *entirely* sure that vectorizing via Cairo would have been able to complete under the 32bit binary of Mapnik I have built. In theory, though, that's not an unreasonable question; I just don't think it would have solved all my problems in this case. (the first not being solved by magnification anyway, because I still can't compile Mapnik :)) 1. I couldn’t get Mapnik to build. I’ve never built Mapnik under Debian Stable; I use Debian packages, but there hasn’t been a release since Cairo support was added. So I didn’t have the ability to render to a Cairo surface.

2. Since I was ‘printing’ the image at 600dpi natively, I don’t think this would have changed anything. Had I, instead, printed at 90dpi and let the printer resize it, it probably would have worked, but I didn’t realize that what I had wouldn’t work until after it was already done.

3. “printing” that area at 90dpi or what have you would have left out a lot of the details that I wanted to see; Since the goal was really to get a coverage map, I wanted to see everything, and ‘zooming out’ by a factor of 6 leaves out a lot of the lower level names and so on. In this case, I couldn’t see them anyway, but with magnification of some kind I would.

4. The printer doesn’t cleanly support all versions of PDF, so I wasn’t sure that printing to a Cairo PDF would work.

5. Since rasterizing to png required 3GB of memory, I’m not *entirely* sure that vectorizing via Cairo would have been able to complete under the 32bit binary of Mapnik I have built.

In theory, though, that’s not an unreasonable question; I just don’t think it would have solved all my problems in this case. (the first not being solved by magnification anyway, because I still can’t compile Mapnik :))

]]> By: tbonfort http://crschmidt.net/blog/archives/362/making-a-big-osm-map/comment-page-1/#comment-20205 tbonfort Thu, 12 Feb 2009 19:07:52 +0000 http://crschmidt.net/blog/?p=362#comment-20205 Why didn't you use svg (or pdf, don't know if that exists by I would suspect so knowing it's cairo behind the scenes) output? this would seem the perfect output for the task, and you wouldn't have to worry about the magnification stuff. Why didn’t you use svg (or pdf, don’t know if that exists by I would suspect so knowing it’s cairo behind the scenes) output? this would seem the perfect output for the task, and you wouldn’t have to worry about the magnification stuff.

]]> By: crschmidt http://crschmidt.net/blog/archives/362/making-a-big-osm-map/comment-page-1/#comment-20204 crschmidt Thu, 12 Feb 2009 17:57:38 +0000 http://crschmidt.net/blog/?p=362#comment-20204 In the post, now. In the post, now.

]]> By: STH http://crschmidt.net/blog/archives/362/making-a-big-osm-map/comment-page-1/#comment-20203 STH Thu, 12 Feb 2009 17:43:59 +0000 http://crschmidt.net/blog/?p=362#comment-20203 But where is the picture of the printed map? But where is the picture of the printed map?

]]>