~jzs/sketchground.site

8f9ad0080d6ca50151ee7cc83193b839bdd335fd — Jens Zeilund 1 year, 1 month ago a27175a mapping_part_one
Mapping article wip
1 files changed, 51 insertions(+), 0 deletions(-)

A blog/2021-02-04_Mapping_part_one.md
A blog/2021-02-04_Mapping_part_one.md => blog/2021-02-04_Mapping_part_one.md +51 -0
@@ 0,0 1,51 @@
# Maps part one

Maps has always been a big facination to me. I can spend hours browsing google
earth and looking at various places on earth. I can find myself using google 
street view "walking" random streets in Akihabara, Tokyo, Japan or random small
towns in germany, italy, etc. For korean street views i have been using Naver
maps.

It is far from only sattelite imaging and street view imaging that has 
fascinated me. Digital maps, physical maps and handdrawn stylistic maps has an
equally big fascination.

Todays biggest commercial options for digital maps seems to be Google Maps and
Apple Maps. However when it comes to open map data, nothing beats Open Street
Maps.

With this series of blog posts i will take a deep dive into Open Street Maps
and document my learnings from how to get from getting the map data to draw an
actual map based on these data. I will be document protocols and data formats
widely used as well as venturing in to writing my own implementations of some
of it. But i might also use off the shelf software to convert data between 
formats.
This will be no small feat and take whatever time it takes.

## Datasets

To render a map you need a dataset of some sort. When you hand draw a map, your
dataset is most likely another map along with the special knowledge you want to
portray in your own map. That could be special shops you want to highlight,
sights, walking routes, bus lines, elevation, etc. It all depends on the
purpose of your map.

The same applies to digital maps. In Denmark we have a bunch of open map data
available such as height data, administrative regional data, property rights 
data, city names data etc. I will use openstreetmaps which is an open source
project as my primary datasource as it collects data from the whole planet.

### Openstreetmap

The current dataset of openstreetmap(osm) is currently at ~1370gb uncompressed
and ~100gb compressed. For my experiments however my goal is to focus on 
Denmark which you can download as an extract from 
[geofabrik.de](https://download.geofabrik.de/europe.html) and that lands at
366mb for the ".osm.pbf" format. 

## Getting from data to map

- Explore the osm.pbf format.
- Explore existing map presentation libraries.
- Explore various ways to serve data for rendering.
- Explore various formats used in the wild. (mbtiles, vector tiles, image tiles) etc.