Home → Community Archive →

How do you learn map data?

• Edit

Aleksej

02 Mar, 2014 09:21 PM

How do you learn map data? How do you best make your cards for that?

1. 1 Posted by vokietis on 02 Mar, 2014 09:52 PM

   

   I want to learn the following:
   Given a region, locate it and tell its capital;
   Given a location, know its name;
   Given a capital, know the corresponding region.
   (I know that task 1 violates the minimal information principle. You may easily split it.)

   An example fact:
   Field 1 (Département): Haute-Saône
   Field 2 (Capital): Vesoul
   Field 3 (Map): [a copy of http://de.wikipedia.org/wiki/Datei:D%C3%A9partement_70_in_France.svg]

   Card 1: Question: [[Département]] /br [a copy of http://commons.wikimedia.org/wiki/File:France_location_map-Regions_...]
   Answer: [Département] /br [[Map]]

   (When confronted with the question, I tell the capital and point to the Département. If I'm right, the region below the pointer becomes red.)

   The other cards are built accordingly.

   Works very well for me, also with rivers and cities.

   • Edit

2. 2 Posted by Heather on 03 Mar, 2014 02:36 AM

   

   The 'image occlusion' add-on could be helpful if you want maps...

   • Edit

3. 3 Posted by Aleksej on 03 Mar, 2014 05:56 PM

   

   What I want to learn now is relative locations of streets (and sometimes objects like house numbers).

   Cards for those are pretty easy to generate using Image Occlusion 2, notwithstanding the bug which sometimes suddenly makes it impossible to generate them. I make a screenshot of an area on OpenStreetMap and add rectangles where the captions are (long enough to avoid hinting at the shortness).

   That's not perfect:

   1. I have to find a zoom level and/or a renderer where the captions are in convenient places and not duplicated.
      
   2. The captions can be small (streets thin), so the picture needs zooming in.
      
   3. Too much information gets into the screenshot. This is what I'd like to find a solution for.

   The extra information is:
   1. Captions from streets, bus stops (which often duplicate those I want to occlude) etc.
   2. Streets, buildings and anything else that is impossible to look through immediately, and thus slows me down or lets me answer correctly before I read the original question.

   (I've been making rectangular screenshots so far. Maybe I should try freehand ones using KSnapshot, but maybe that would be yet another bad hint.)

   I imagine this workflow:

   1. Choose an area on a map.
      
   2. Select all objects and hide them, which would be indicated by dimming.
      
   3. Select the objects I need in the card and unhide them.
      
   4. Tweak the captions.
      
   5. Render with a preview.
      
   6. Copy to clipboard.

   Maybe there is a (FLOSS) program that allows doing that quickly?

   • Edit

4. System closed this discussion on 22 Feb, 2016 11:50 PM.