The last time Hackerfall tried to access this page, it returned a not found error. A cached version of the page is below, or clickhereto continue anyway

The use of web maps has rocketed over the last decade. Back in 2005, Google Maps was released, facilitating the use of online maps by the masses for everyday life. Initially, web mapping was the domain of a few large companies like Google, ESRI and NASA. However, with the explosion of open source mapping projects under the banner of FOSS4G (short for free open source software for geo-applications) and affiliated with the umbrella body of the Open Source Geospatial Foundation (OSGeo), web mapping can now be undertaken by anyone with a computer and an internet connection.

Inspired by a web mapping course at the University of Gironas SIGTE service - much of which can now be done online via UNIGIS - I decided to test out three of the major options, Google Maps, OpenLayers and Leaflet. Thanks to such software, it has never been easier to make and communicate a map. To illustrate the empowering potential of this, take a look at the recently implemented City Connect consultation map:

This map is frankly amazing: it is democratic, lowering the barrier to entrance to spatial planning by making it accessible to anyone who has a computer and (Ive not tested it) perhaps even just a smartphone. The rating of comments by other people mitigates against abuse, and the option to show exactly where your issue lies enables planners to make better use of local knowledge. The map, interestingly, was custom built using Leaflet, the new kid on the block in web mapping.

For me this in many ways represents the ideal web map towards which we should strive. It is:

With the myriad of free options available to to public, it can seem very hard to get involved in web mapping. Thats why the rest of this post provides a brief overview of three of the main options for building and online web map. In order of completion date, these are:

  1. Google Maps (version 1, with basic API deployed in 2005)
  2. OpenLayers (large JavaScript library initiated in 2005)
  3. Leaflet (small JavaScript library initiated in September 2009 by Vladamir Agafonkin)

These are the three main client side web mapping tools, but there are many more. More that 44 according to one article dating from 2012! You may also want to consider map servers such as GeoServer if you are dealing with large applications but, for small application, an updated .geojson file hosted for free can work work just fine. This is amazing. It means that you, me or anyone can take control of our geographic information and communicate it to the world how we want.

Google maps

Google maps is the oldest and easiest to use web mapping tool of the three. After you have set up a unique API key, putting the below map on your site is as simple as placing the following code into the header of you web page (see this pages source to see what I mean):

<script type="text/javascript" src=""></script>
<script type="text/javascript">
  function initialize() {
    var latlng = new google.maps.LatLng(54, -2);
    var myOptions = {
      zoom: 8,
      center: latlng,
      mapTypeId: google.maps.MapTypeId.ROADMAP
    var map = new google.maps.Map(document.getElementById("map_canvas"), 

After that, you just add a single line of html in the pages body:

<div id="map_canvas" style="width: 600px; height: 200px"></div> 

All the variables are quite intuitive and quick to implement. Also, with the brute force of Googles vast server infrastructure, Google maps are guaranteed to load quickly. The same applies to the code base: with some of the best payed junior programmers in the world, it is unlikely that bugs in Google maps survive more than a couple of days. See below an image of Googles relatively flat pay structure compared with the quite unequal one of Oracle.

Google maps also has a surprising range of functionalities, although you are pretty much restricted to Googles way of doing things (you cannot, for example change the background map to be based on Open Street Map). This includes:

These are useful, but for Googles way of doing things may not be the way you want to do it: it lacks the flexibility of open source options, where literally anything can be modified (not that this is for the faint hearted). If you want complete control over your data and like the idea of supporting open source mapping then one of the open source options may be better for you.


OpenLayers is the most mature open source client side web mapping tool on the internet and probably the most widely used. It is a JavaScript library, meaning it will come naturally to anyone used to building general purpose application in that language, which is ubiquitous on the internet. OpenLayers has a very wide range of functions, and it pays the price in terms of size, occupying almost 1 Mb of digital space: not ideal for mobile applications.

Partly because of this bulkiness (less of a problem as devices become faster) the OpenLayers developers are gearing up to release a major new version: OpenLayers3 which offers major improvements in performance with less code than OpenLayers2 (113 vs 126 thousand lines). For this reason the example below uses OpenLayers3. It is not the pretiest of maps, but this is not the codes fault it is the users (thats me!). Take a look at the list of examples for more beautiful and advanced maps written in OpenLayers3.

For the record, the below map was created with even fewer lines than the Google Maps example. The following code, taken from the OpenLayers3 Quickstart goes in the html head:

<link rel="stylesheet" href="" type="text/css">
<script src="" type="text/javascript"></script>

And then the following goes in the body:

<div id="mapOL3" style="width: 100%, height: 400px"></div>
  new ol.Map({
    layers: [
      new ol.layer.Tile({source: new ol.source.OSM()})
    view: new ol.View2D({
      center: [0, 0],
      zoom: 2
    target: 'mapOL3'

One improvement over OpenLayers2 is that when adding an OpenLayers map, you no longer need to initialise it in the body tag, as is for the Google Maps version. Hence, to load the Google Map, one needs to add this to the pages body: <body onload="initialize()">. This is not needed for Leaflet maps either.


Leaflet is in fashion at the moment as a fast moving, lightweight and relatively intuitive JavaScript library that aims to be light on computer and programmer resouces. It allows you to get maps up quickly.

Dont let that make you suppose that the library lacks functionality, however. The cycle path consultation map mentioned in the introduction was built in Leaflet, and its functionality is ever growing, not least via a growing number of plugins.

This modular way of doing things intuitively makes sense to me, meaning that the core Leaflet developers can focus on getting the basics as good as possible, keeping the Leaflet.js library small enough to load rapidly on a mobile device. This is why Leaflet is probably the API of choice for open source mobile application developers and why it is used by GitHub, a central hub in the open source software movement.

Interestingly, Leaflet does not require initialisation in the body. It still requires a couple of scripts, shown below, in the head.

<link rel="stylesheet" href="" />
<script src=""></script>

For the main map description, instead asking for the script to be added in the html, Leaflet asks for it to be called after the map object is called, like so:

<div id="map" style="width: 600px; height: 400px"></div>
		var map ='map').setView([51.505, -0.09], 13);

		L.tileLayer('http://{s}{z}/{x}/{y}.png', {
			maxZoom: 18,
			attribution: 'Map data &copy; <a href="">OpenStreetMap</a> contributors, <a href="">CC-BY-SA</a>, Imagery  <a href="">CloudMade</a>'

In fact the actual code used for the above map has a few more lines of code, to add the point, the blog and the triangle. But the snippet shown still works as a stand-alone map. The example can be found here.


We have seen three common ways to build maps into html pages and these offer amazing opportunities for displaying our geographical data in new ways. The example of cycle path planning shows the potential for democratic empowerment that such technologies offer. However, for such potential to be realised, and not just gobbled by massive corporations like Google, we need diversity. The importance of this for our online future is well described in an article by Serge Wroclawski. Google, for example, can easily insert adverts into its maps without us knowing. Leaflet and OpenLayers, being open source, cannot. That said, if your aim is just to get attractive off the shelf maps up quickly, Google Maps is a good option.

OpenLayers is mature and big and works well with servers. Leaflet, as the new kid on the block is the most exciting for me and encapsulates many of the benefits of open source software in general: speed of development, flexibility, efficiency. All three can probably do 80% of the web mapping applications that are commonly required so the choice should, in many cases, depend more on what you enjoy than on what it can and cant do. Lightness considered, for me that would mean Leaflet for many applications, but watch this space for the final version of OpenLayers3 (due very soon) and perhaps even better web mapping options.

Continue reading on