An Open-source World

Over the last few years we have seen the release of numerous mapping platforms/frameworks to allow the manipulation and representation of open-source mapping and other spatial data. This has been, and continues to be, a significant step-change in cartography, probably the most significant in decades. All of these new frameworks have enabled anyone to manipulate spatial data, present this data, style topographic maps, create custom maps, and much more. Mapbox’s Geotagger’s World Atlas from 2015 (shown below)—an interactive map that visualises 10 years’ worth of location data from photos uploaded to Flickr—is a good example of this. This article is a brief summary of the tools available.

The open-source world is very important in (online) map-making. For me, the key frameworks/platforms below are of particular interest as tools to manage and manipulate data, find patterns and correlations, create new perspectives, render eye-catching (geo)visualisations and generate file formats that can be further enhanced, manipulated and re-purposed in Adobe applications. All this without mentioning best practice in data visualisation, as championed by Edward Tufte, or a proper understanding of data collection, significance and statistics—Garbage in garbage out (GIGO) to be discussed in a later article.

OpenStreetMap (OSM)

This is worth starting with first as it is the definitive open source mapping available online, and referenced by most of the frameworks listed below. To imagine this map database is the work of thousands of volunteers across the globe, ever-expanding and ever updating, is testament to the democratisation of mapping. There is a discipline and process that ensures that the accuracy is now good enough for an ever increasing number of professional and critical users and applications. In the early days, it was interesting, but not reliable enough for professional application. Today, it’s much improved, through still really only for smaller-scale products. Personally, it’s use at scales greater than 1:5,000 is questionable. At smaller scales, it is more impressive. It’s an ever improving and developing product so can only get better. I’ve often used it as a compilation source for work. The pedestrian wayfinding mapping for Bologna, Italy, used OSM as the underlying topography.

Mapbox

Professionally, I feel this is the platform (as Mapbox describe their offer) that I will invest my time in mastering, though it can also be accessed by newbies! Essentially, it’s designed to integrate any location-based or spatially referenced data with OSM and high resolution satellite imagery, working with 2D and 3D data. It’s the platform that seems to have developed most and fastest these last few years. It’s being used by an ever-increasing number of both private and public organisations to generate applications that span numerous industries from agriculture to travel. Applications cover asset routing through to data visualisations. Interestingly, this is both a vector and raster-based platform.

As a cartographer, Mapbox provides a ‘Studio’ environment to allow the development of map products within the browser. It provides tools to import and manage data types and allows total control over styling and layout with tools grounded in existing web standards, such as CSS. This is a professional tool for designers. You can work with existing or custom data sets, creating cartographic visualisations that can be interacted with across numerous platforms including iOS, Android and the web. Excitingly, there has been integration with other visualisation tools like Unity and Tableau, that provides an extension in to 3D worlds, virtual realty (VR), data dashboards and more. In fact, Mapbox sum it perfectly themselves:

Build gorgeous maps to present data in new ways that help users discover [new] insights.

QGIS

Probably the mostly widely used open source geographical information system (GIS) available, supporting all major operating systems, including Mac OS X, Windows, Unix, Linux and Android. It is a professional application, under constant development by an army of enthusiasts that are also creating hundreds of useful plug-ins that extended the functionality of the application even further. For most data handling and spatial analysis tasks, QGIS holds up well against the two big mainstream providers in this arena, ESRI (ArcGIS) and Pitney Bowes (MapInfo).

Leaflet

Presented as the leading open-source JavaScript library for mobile-friendly interactive maps. Without apology, this is designed to provide the features you need in as simple and efficient package as possible, with performance in mind. That said, it is still a feature-rich framework that enables lots of interactivity types and the creation of specialist map types such as choropleth and heat maps. It also integrates other frameworks such as GeoJSON.

Kartograph

Another “simple and and lightweight framework for building interactive mapping applications without Google Maps or any other [served up] mapping service.” This has particular interest as it works with scalable vector graphics (SVG), a format that looks good across all screen-based devices. I’m unsure as to how Kartograph is developing, or will continue to develop; sometimes good ideas just fall away under the pressure of other competing frameworks. Kartograph provide a JavaScript and Python library to build interactive maps. There are some nice examples here, showing promising representational options. For me, as a ‘heavy’ Adobe Illustrator user, the interoperability with SVG makes this very interesting, as demonstrated by the East Coast example.

Maperitive

This is an interesting product, unfortunately not helped by the dated website. The initial perception is perhaps of an old technological approach. Originally, this was PC only—I’m a Mac-user—so whilst I felt it interesting, it was of no potential use to me. However, there does seem to be a Mac/Linux version available that runs on Mono. The website belies some serious and professional intent and applications. The discussion and examples looking at hill-shading and elevation colouring are particularly interesting and grounded in solid cartographic theory.

CARTO

Another very interesting tool, CARTO Builder is a web-based drag and drop analysis tool for analysts and business users to discover and predict key insights from location data. With a clear focus on location intelligence and business applications, the mapping backdrop is OSM. The growing trend for web-based frameworks delivered via custom-designed dashboards, removes the need for proprietary software to be installed and loaded on your desktop machine.

D3.js

Another JavaScript library for manipulating and visualising data, but there are a lot of visualisation options, map-based as well and diagrammatic. Whilst complex data handling and visualisations are possible, the development of D3.js is grounded in a simpler and efficient code base, built upon existing web standards. There’s lots of fascinating techniques here that are accessible to all.

There are, of course, more than this but to my mind, these are currently the ‘key players’, but with technology, that can be turned on it’s head overnight. It’s a fascinating time to be involved in mapping!

Postscript

There is another important point to raise in respect to our engagement with technology, and that is of the sense of ownership and freedom to shape these tools to our own ends. This relates to the norms of engagement with established and proprietary tools from the corporate giants such as Apple, Microsoft, Adobe, ESRI, et al., where we, arguably, feel like we are free to act only within their framework, delivering solutions that these tools allow or direct you to. After all, how many of us really know how the software works that we are using? In respect to mapping, ESRI, probably the largest GIS developer, is fast making it’s tools available online and developing ‘add-ons’ to other products, such as Adobe’s Creative Cloud apps… on a monthly subscription, of course! That of course, is the current trend—subscribe to our services on never-ending financial obligations. We no longer own tools, simply a right to use them.

Open source software (and data) encourages us to understand how the output is generated. For example, Paper.js, p5.js and Processing create an understanding of how a bezier curve is expressed as a mathematical equation because they encourage designers to ‘code’. We should all, as a matter of course, make time to discover and engage with new applications and tools that encourage new thinking and solutions.


May, 2017

Heightfields

There is always an aspect of work that emerges out of error, experimentation and/or serendipity.

National Geographic

Since the very first issue in 1888, the National Geographic magazine has been the benchmark for quality researched articles supported by well-crafted graphics.

Relief Shading, Pt I

As we seem to be moving in to presentational modes where 3D (mapping) is becoming ever more commonplace, enabled by software and hardware advances, it is relevant to look back

Swiss National Mapping

Swiss topography has long been regarded as the epitome of this form of mapping, often characterised by the iconic hill shading techniques of the late Eduard Imhof, professor of cartography at the Swiss Federal Institute of Technology.

A Monitor for a Cartographer

Screen estate and colour are very important to a cartographer. Maps can be any size from A6 to more than 5m×5m square.

An Open-source World

Over the last few years we have seen the release of numerous mapping platforms/frameworks to allow the manipulation and representation of open-source mapping and other spatial data.

OS Maps – the App

Recently, the Ordnance Survey (OS) updated an app for both mobile and desktop devices that utilises their 1:25,000 Explorer and 1:50,000 Landranger maps—classic OS mapping at it’s best.

Mac Pro 5,1 Resurrected

I decided to raise the bar on my workstation and go for something with a little more grunt, and that would be up to the job for the next 2–3 years.