Quantcast
Channel: maps for developers - Medium
Viewing all 2230 articles
Browse latest View live

Visualizing 3 Billion Tweets

$
0
0

This is a look at 3 billion tweets - every geotagged tweet since September 2011, mapped, showing facets of Twitter's ecosystem and userbase in incredible new detail, revealing demographic, cultural, and social patterns down to city level detail, across the entire world. We were brought in by the data team at Gnip, who have awesome APIs and raw access to the Twitter firehose, and together Tom and data artist Eric Fischer used our open source tools to visualize the data and build interfaces that let you explore the stories of space, language, and access to technology.

This is big data, and there's a significant level of geographic overlap between tweets, so Eric wrote an open-source tool that de-duplicated 2.7 billion overlapping datapoints, leaving 280 million unique locations.

Locals and Touristssee the full screen Locals and Tourists site.

Locals and Tourists analyzes behavior over time to highlight areas of cities popular with locals and places that are usually documented by tourists. It's a great expansion of Eric Fischer's previous work on the concept with Flickr data and city-level maps.

To make this map, Tweets are grouped by user and sorted into locals—who post in one city for one consecutive month—and tourists—whose tweets are centered in another city. Relatively inactive users simply don't appear on the map, since we can't confidently determine their group.

Mobile Devices

Mobile Devices is a map that reveals the information about phone brands that is stored when people use an official Twitter App and is hidden in the metadata attached to each tweet. Each brand of phone is a different color and can be independently toggled. The patterns of usage in each city often reflect economic stratification. For example iPhones, in red, are predominantly in wealthy sections of the city while Android phones, in green, have more coverage in poorer sections. On a global level, national trends reveal a complicated set of cultural preferences. (Tweets from web browsers and from other Twitter clients don't appear on this map)

domination of Blackberry in Indonesia

Mobile Devices showing the domination of Blackberry in Indonesia. see the full screen Mobile Devices site.

Languages of Twitter

Languages of Twitter is a high-definition evolution of an earlier world language map. It takes advantage of Gnip's language detection technology to identify the language of each Tweet so that the areas where each language dominates can have a distinct color. Because English is so widely used in Twitter around the world, it is colored gray to let other languages stand out.

Languages of Twittersee the full screen Languages of Twitter site.

Spanish is the second most common language on Twitter, especially in the United States, so it is also individually selectable on the map. The other languages of the world are colored so that ones that are spoken near each other will be as far apart as possible on the color spectrum. In many cases they follow national borders, but in other cases language usage shows a more complicated pattern. Tweets whose language could not be determined don't appear on this map.

Each of the maps was made from the same set of 280 million locations, rendered to map tiles in different combinations of colors and layers to reflect the map's theme, using a program related to the one that Eric presented at State of the Map US.

Built with Open Source

These maps were generated with a variation of datamaps, a custom toolchain for visualization. This generated raw tiles, which were compiled to MBTiles with mbutil and uploaded with TileMill to MapBox.com. MapBox.js made it easy to build a dynamic interface around each map that lets users toggle categories and explore in detail. Each map uses a low-level geocoder to orient users if they're in an unfamiliar part of the world, and a forward geocoder to pop to their city or favorite place. The Language map includes a legend that doubles as a visualization of worldwide language distribution, and Locals and Tourists includes an index of the most interesting places to look - tourist attractions, airports, and transit in major cities.


Using Leaflet plugins with MapBox.js: A Showcase

$
0
0

One of the reasons we built MapBox.js v1 on top of Leaflet was its robust plugin ecosystem. The Leaflet plugins page lists dozens of plugins, and more are being added every week. In fact, MapBox.js itself is structured as a Leaflet plugin, unlocking a world of interesting possibilities for map interaction and enhancement. Here are a few of our favorites.

Leaflet.draw

MapBox.js supports displaying vector layers out of the box thanks to Leaflet -- it's what drives GitHub's new GeoJSON preview. With Leaflet.draw, these layers become editable: you can draw new lines, circles, or polygons, edit existing ones, and delete what you no longer need. When you're finished, use L.LayerGroup#toGeoJSON to get the data back as GeoJSON and complete the loop. It's your own lightweight map editor.

Example: Using the Leaflet.draw plugin

Leaflet.markercluster

Sometimes you have a lot of markers to display, and plotting them all on the map would be overwhelming. Leaflet.markercluster is the solution. It automatically groups markers that are close to each other into a single group. When you zoom in, the groups are recalculated, and when there's enough room, individual markers are shown. USA Today used this plugin to great effect in a visualization of images from the Civil Air Patrol following the Moore, OK, tornado.

Example: Using the Leaflet.markercluster plugin

Leaflet-MiniMap

If you are showing a detailed, large-scale map, it can be helpful to orient the viewer with a secondary map at a smaller scale. This is exactly what Leaflet-MiniMap does. The mini-map tracks pan and zoom events in the main map, and can be configured to use a different tile set -- for example, you could show MapBox Streets in the main map and MapBox terrain in the mini-map.

Example: Using the Leaflet-MiniMap plugin

Leaflet-hash and Leaflet.fullscreen

Leaflet-hash and Leaflet.fullscreen are two ways to enhance your map UI quickly and unobtrusively. Leaflet-hash provides permalinks for easily bookmarking and sharing map locations. And our own Leaflet.fullscreen adds a button that goes fullscreen with the HTML5 API when it's available and falls back to an alternate implementation when not.

Example: Hash and Fullscreen UI elements

Using Leaflet plugins

In order to use these or other Leaflet plugins, you'll want to familiarize yourself with their documentation, which is typically provided in a README.md on GitHub. In most cases, you'll need to include a <script> tag linking to the plugin's JavaScript source file and a <link> tag for its CSS. You'll need to host these assets on your server -- unlike MapBox.js, they aren't hosted on MapBox's CDN. From there, it's usually just a couple lines of JavaScript to configure and initialize a plugin -- check out the examples above for details.

New MapBox Stickers

$
0
0

We took some time out of our busy schedules here at MapBox to churn out some new goodies. Check it out.

Inspired by the dozens of bold, iconic NASA shuttle mission patches, our new stickers aim to capture the magic of space exploration and satellite imaging with a little kick of nostalgia.

Want one of your own? Just fill out our sticker request form below. Slap 'em on your bumper, laptop, satellite, whatever. We've got you covered.

MapBox Sticker Request

* Required

MapBox Transparency Report

$
0
0

We just launched the Mapbox Transparency Report, an effort directly inspired by our commitment to protecting our users' privacy and data. Recent revelations of the scale of government surveillance on private digital communications have prompted many established services -- including Twitter, Microsoft, Google, and Yahoo -- to publicly disclose government requests for user data. At MapBox, we want to be clear and proactive, always taking the side of privacy and disclosure to the fullest extent we can under current law.

Our commitment to our users

  • We will never voluntarily share user data.
  • We will report all requests that we are legally allowed to report.
  • We will always proactively notify users of requests unless we’re prohibited from doing so by law.

"Sunlight is said to be the best of disinfectants; electric light the most efficient policeman"

U.S. Supreme Court Associate Justice Louis D. Brandeis

The MapBox Transparency Report is a proactive transparency effort reflecting our commitment to our users. The report will be updated quarterly to include all government requests for user data that we are not legally prohibited from disclosing. We will also report all copyright takedown notices filed under the Digital Millennium Copyright Act (DMCA).

Current Status

Request Number Received All Time Number received 2013
Government requests received for user information 0 0
Government requests received to withhold content 0 0
DMCA takedown notices received from copyright holders 0 0

Your Transparent Future

We believe strongly in government accountability and transparency; for this reason, MapBox has joined with more than 490,000 others in recognizing that citizen security and citizen privacy do not need to be at odds through the Stop Watching Us campaign.

Pig Roast This Friday in the MapBox Garage

$
0
0

This Friday we're throwing open the garage doors to host a pig roast at MapBox. Join us for a whole barbecued pig smoked by the folks at Rocklands, a few ice cold kegs to help you fend off the 90°+ temperatures, and some rounds of foosball and ping pong.

This Friday we're throwing open the garage doors to host a pig roast at MapBox. Join us for a whole barbecued pig smoked by the folks at Rocklands, a few ice cold kegs to help you fend off the 90°+ temperatures, and some rounds of foosball and ping pong. Our full team is currently in Washington, DC having traveled from San Francisco, Portland, Germany, Peru, England, and Washington state and are excited to celebrate together with our local friends. We'll get started at 6:00 pm this Friday, June 28 and go until late. RSVP here.

The MapBox Garage is located at 1714 14th Street NW, and our entrance is in the alley behind 14th Street (behind Peregrine). The best way to get to us is to go to the corner of 14th and R, walk west on R, and north up the alley. You'll see our sign from R Street - we are the blue building with the glass garage doors.

Christopher Jones Joins MapBox

$
0
0

Christopher Jones, a long time colleague in the open source and open data space has just joined MapBox. Christopher will be working closely with Ian, Will and myself to expand the ability of our cloud service to meet the needs of our rapidly growing subscriber base. He has a diverse experience in backend systems which have supported everything from stolen bike tracking to medical metrics and financial databases. He brings an unshakable calm to our devops team and a willingness to dive right into dark and tricky parts of infrastructure management. Welcome!

Chris Jones

Announcing MapBox.js v1.1.0 & Leaflet 0.6.1

$
0
0

Today we're announcing the release of MapBox.js v1.1.0, which integrates with today's release of Leaflet 0.6.1. This week has been the first official Leaflet code sprint, hosted at the MapBox office in DC with Leaflet's creator, Vladimir Agafonkin.

When MapBox.js Switched to Leaflet in v1.0.0, we had bigger plans in mind - to invest in the core of Leaflet to make web mapping better for all. The ecosystem of Leaflet is one of the most important parts of building on an open plaform that has wide reach and a large community of developers and mappers. We're making sure that Leaflet's core is fast and flexible, so that open maps win.

Improving & Refining

Before MapBox.js v1.1.0, we relied on a post-0.5.1 version of Leaflet that wasn't official tagged. Now that Leaflet 0.6.1 is tagged after 5 months of development, we can rely on a major version - and more importantly - distribute MapBox.js as a proper Leaflet plugin, without a bundled Leaflet.

MapBox.js v1.1.0 also includes many improvements to our thorough API Documentation, bugfixes, and improvements to the map API. This version also includes some changes that will make SSL support simpler and easier.

Leaflet gained many improvements in the push, including improvements from our original push on MapBox.js. Leaflet now uses mocha for testing, provides a wider array of layer getters and setters, and has improved behavior around map interaction. The project is also wisely staying small and simple by moving some previously-core functionality into plugins like Leaflet.draw

We also included a new control, L.mapbox.shareControl, which lets you add Facebook and Twitter share buttons to your map in one easy line.

TileMill Seamless Blur

$
0
0

Experimenting with the new ability in TileMill to seamlessly blur geometries -- coming soon in the next release of TileMill.


GeoDC Meetup Tonight on Geodata Formats

$
0
0

Join us tonight for the July GeoDC meetup on geodata formats starting at 7:00 pm at Stetson's. John Firebaugh, in town from San Francisco, will present on the GeoJSON file format, walking through when and why to use it, how to convert to it, and how to publish and edit data in the format. He'll be joined by Ben Balter of GitHub who will talk about GitHub's new support of GeoJSON, how this lets you go from data to map in seconds, and the bright future GitHub sees for open source mapping. Finally, Andrew Turner will do a quick overview of the geodata formats that exist and when you should use them. See you tonight.

SSL Maps for Everyone

$
0
0

MapBox now supports SSL on every plan and across every service - at no extra charge. SSL is a necessity for every secure website, and it's vital to have all remote resources sent via the HTTPS protocol in order to protect users' privacy and to avoid mixed-content warnings.

We've made it incredibly easy to start using SSL. MapBox.js v1.1.0 will automatically detect when you are using a secure webpage and will use purely SSL resources when necessary. Users can access the full REST API securely as well, including tiles, static images, geocoding, embeds, and JSON endpoints. These requests have no additional cost; HTTP and HTTPS traffic is billed at the same rate.

On mobile, MapBox iOS SDK v1.0.3 has the enablingSSL option that will make all requests securely for a particular map layer.

SSL is no longer limited to Plus and Premium plans and no longer requires a beta process - so many users can save money by switching to a plan that perfectly fits their usage level, and now everyone can make their sites and maps more secure.

Batticaloa Progress on OpenStreetMap

$
0
0

OpenStreetMap mapping progress in and around Batticaloa, Sri Lanka (7°43′0″N 81°42′0″E) since May 1, 2013.

The progress made in mapping was impressive, with the entire Batticaloa area traced and on the ground information being added at the time of this posting. This ongoing mapping effort is a part of the Open Cities Project, seeking to unlock OpenStreetMap's potential for urban planning and emergency preparedeness.

The checker board pattern showing in the visualization is a result of the use of the HOT Tasking manager, where contributors claim tasks as square-shaped areas.

Easy Maps with GitHub

$
0
0

Live from the GeoDC meetup -- Ben Balter of GitHub talks about making data social now that GitHub added maps and supports GeoJSON. It's as easy as committing your JSON, GeoJSON or TopoJSON file to GitHub. The data is then plotted on a map that can even be embedded! The whole meetup focused on open data formats, also check out the rest of the talks from John Firebaugh and Andrew Turner.

Real Time Imagery via MapBox Satellite Live

$
0
0

We're launching MapBox Satellite Live later this summer. This new service will provide instant access to timely satellite imagery from anywhere in the world, directly though MapBox.com. We are currently building out the direct pipeline from the satellites, optimizing the processing workflows to ensure that the imagery moves as quickly as possible from our network of satellites → ground stations → MapBox.com. Our goal is to have live imagery for publishing within 6 hours of an event, anywhere in the world.

MapBox Satellite Live data pipeline

Not only do we make it easy to publish in minutes, but having MapBox Satellite Live directly as part of your MapBox.com account means you have instant scale for millions of viewers - which is critical for our large media partners. We are also hearing a lot of interest for this imagery behind paywalls, using MapBox's default SSL.

We will be opening up a private beta testing program to selected partners in early August. If you have questions ping me (@hrwgc) or Dave (@dhcole) or just email support (at) mapbox.com.

Dashboards for Data

$
0
0

The MapBox Garage now sports five dashboard displays that visualize the growth and form of MapBox and OpenStreetMap.

MapBox's server and CDN architecture is the core of our speed and reliability. We need to keep track of response time across the world continuously, so we use a number of services, including Server Density and Librato to collect real-time statistics. This display makes that tangible, and includes the current position of Landsat 8 for fun too.

Support Leaderboard

We do a lot of support - hundreds of issues are filed and responded to by hardworking folks. We can now see this, and give extra cookies to people who help out on support.mapbox.com

Stats Leaderboard

This shows traffic flowing from websites to mapbox.com, and what people are interested in, each day of a week.

We use a simple proxy server that makes certain APIs CORS-enabled, thus making it possible to build these panels purely in JavaScript. They're each written with d3js and a few small modules like web-rrd and d3tron that tackle specific issues.

Pager Duty

The MapBox devops team uses Pager Duty to keep track of late-night server issues. This pulls from the API and shows recent incidents, how long they stayed open, and who resolved them. It's also on a 24-hour clock to highlight the 1am calls that also merit extra cookies.

DC Headlines

Most of the team is based in Washington, DC and nobody owns a car - so it matters when the 54 bus is on its way and when the local Capital Bikeshare station is full. This combines that information with Dark Sky forecasts and SunCalc dawn & dusk predictions to help everyone time their errands and commutes perfectly.

Hiring Business Development Lead at MapBox San Francisco

$
0
0

We are looking for a head of Business Development to lead partnerships and build out a sales team at MapBox in San Francisco. We are rapidly growing, and to date we have not had a dedicated sales team -- everything has been organic. Now we're looking to radically expand our partnerships in the Bay Area and across the country and build out a full sales and account team to continue our rapid growth. We're just getting started.

You'll be expected to think creatively and strategically about establishing a diverse array of partnerships. MapBox works closely with organizations like the Knight Foundation and OpenStreetMap US, fast-paced tech startups like Foursquare, Evernote, and GitHub, and with some of the leading satellite and aerial imagery infrastructure providers like DigitalGlobe and Skybox. We're also constantly pushing public-private sector partnerships through open data by leveraging the products of NASA, USGS, the US Census Bureau and more.

Our team is a tight-knit group of 35, including cartographers, data analysts, and software engineers. In the past three years, we've built our entire stack using open source software and open data. You’ll be a core part of this team, helping establish business direction for our products and making the most of our fast-paced R+D.

Qualities we're looking for

  • A great communicator with passion for learning whatever it takes to get the job done.
  • A technologist who is hungry to learn the details of how MapBox builds and runs its products.
  • Past business development leadership experience, including building a sales team from the ground up.
  • Understands the geo/location space and is ready to be part of a disruptive team.
  • Rapid-growth startup experience (a plus).

To apply

Please send a note with the following to jobs@mapbox.com:

  • Why you're interested in MapBox
  • Your past experience in business development
  • Resume

Smoking Hot Coffee Interview

$
0
0

I recently had a great interview with the guys at Smoking Hot Coffee, a show covering the startup space. We talked at length about MapBox's strategy, operation, history, and a bit about what might be next.

It was nice to be able to dive into a couple areas in depth with Amul and Jeff such as the MapBox Transparency Report in a discussion about the NSA, GitHub's recent addition of MapBox maps for all, and our work earlier this Spring to launch Cloudless Atlas, the first cloud-free satellite map available at scale.

Watch the interview now: "MapBox is battling Apple & Google for the future of location-based data with Open Source"

Portland: The Age of a City

$
0
0

Portland: The Age of a City

619,000 buildings, 544,033 of them list the year they were built in Portland. Justin Palmer of GitHub created this beautiful map of Portland, Oregon's buildings by age with TileMill. The color scheme that he chose, as well as the varying patterns of neighborhoods, some in neat post-war rows, some in more organic arrangements, makes for a really pleasing visual. Read more about Justin's work in his post Portland, Oregon: The Age of a City

Tracing Landsat 8 for OpenStreetMap

$
0
0

Want to help update OpenStreetMap with fresh, free Landsat 8 imagery? Here’s how to get started tracing whatever interests you, even if you’ve never edited OpenStreetMap before.

Start by finding a place where OSM needs attention to relatively large features – Landsat imagery’s strength is on things like highways, coastlines, and rivers. One of my favorite spots is the Amazon basin. Following the Landsat 8 downloading and processing instructions blogged earlier, I’ve loaded a relatively clear scene of an area in northern Brazil into TileMill. (One note on the processing: it can help with tracing to boost the imagery’s brightness and contrast. For example, on this scene I used -sigmoidal-contrast 100x11% instead of -sigmoidal-contrast 50x16%. Do whatever works best.)

I like to add raster-scaling: lanczos; to the image style for smoother scaling, but just as with the color adjustment, you’re free to adjust as you see fit. Now we’ll export the image through the menu in the top left right of TileMill, choosing Upload:

In the export screen, you can crop to the general area of the scene that you’re interested in. Be sure to render down to zoom level 15 – though Landsat 8 isn’t perfectly crisp at that resolution, it works really well for tracing:

After you upload – which may take a few minutes if you’ve chosen a large area or if you’re on a slower internet connection – you can visit your map on mapbox.com:

Notice the part of the URL that has your username, a dot, and the name of the map. (If you didn’t give the map a short name in TileMill, it may be randomly generated gibberish – don’t worry.) That’s the map ID, in this case char.amzn-tracing. You can also find it under the mapbox.js tab of the Publish interface on the site. Now we’ll make a URL template that tells mapping code where to fetch the image tiles for your map. To do that, simply replace my map ID with yours in this URL:

https://tiles.mapbox.com/v3/ + account.mapid + /{z}/{x}/{y}.png

We’re almost there! Open the iD editor, and if you don’t have an OSM account, sign up for one now and work through the iD demo – it’s speedy and painless. Find your way to the area you’ll be tracing in iD, open the Background menu (the top item in the left-hand sidebar), choose Custom, and paste the URL template:

By the way, notice the background brightness setting (red highlight) – if you find your imagery a little too bright or dark, adjust that. Zoom in on your new Landsat 8 layer, and you’re ready to trace:

Save your work often! With each save, OpenStreetMap will be a little bit better than it was before, and MapBox Streets will be serving the updated river (or lake, mining operation, glacier – whatever you’ve chosen) within minutes.

One big advantage of Landsat over other open imagery is that it arrives so often. Even when it isn’t the sharpest view of a particular region, it can be sushi-fresh. This makes it especially important for features like streams, glaciers, and vegetation, which can change quickly – for worse, like rainforest logging, or better, like river restoration. Interpreting the tremendous flow of open data into timely open information is a big job, but OSM shows it’s possible.

Mapping Millions of Dots

$
0
0

I've made a lot of maps out of dots before, but the ones we made from Gnip's Twitter archives, which were featured in Wired Design a few days ago, were my first experience trying to make dot maps that would work at any scale, anywhere in the world. Here are some of the lessons I learned in the process about making dot maps look good on the web.

Los Angeles

Scaling and brightness

The first thing that becomes clear when you start drawing the same dots at different scales is that it doesn't look right if you just scale the dots proportionately as you scale the area. Each time you zoom in on a web map, only a quarter of the area that was visible before is still visible, but if you match that and draw the dots four times as big as you did at the previous zoom level, the image is very crowded and fuzzy by the time you get zoomed in all the way. The Gnip maps instead double the area of the dots for each level you zoom in. Here's what it looks like to zoom in on Times Square with dots that quadruple, double, or don't change size at all with each zoom level.

Times Square

Doubling the dot size looks pretty good, but it still isn't quite right. If the brightness is right for the neighborhood level, by the time you zoom out to the continental scale it will be too bright there. To keep the density looking right you also have to increase the brightness of the dots as you zoom in and decrease it as you zoom out. By experimenting, I settled on 1.23 as the brightness scaling factor, so between that and the doubling of the dot sizes, the total brightness goes up by 2.46 with each additional zoom level. This isn't an exact science, and other numbers close to that also work pretty well. Here's the northeastern US and Canada with the large cities looking overexposed from strict 2x brightness change at each zoom level, and then moderated by the additional brightness scaling.

Northeast

The other thing that goes along with this brightness scaling is to draw fewer dots at lower zoom levels. By the time you get most of a continent on the screen, the dots are so much smaller than pixels and there are so many of them to draw, that it looks the same and is much faster if you draw half as many dots at twice the brightness apiece. I actually do this at every level where the dots should be smaller than a pixel, and keep the size of the dots at a pixel and draw correspondingly fewer of them instead. Here's that same area looking pretty much the same without dropping any dots, and then looking terrible by dropping too many of them (4x per zoom level instead of 2x).

Northeast pixel dropping

You also have to make a choice about which zoom level should be the baseline where each dot has an area of one pixel. At low zoom levels, it will look the same, but when you are zoomed in it will control how sharp or fuzzy the map looks. For most data sets, the right choice seems to be around zoom level 13, where each pixel is about 50 feet on the ground, which is pretty close to the width of a street or the size of a building lot. For sparser data sets, you might want to use a lower zoom level so that the bigger dots fill in the gaps. Here is Las Vegas Boulevard drawn with three different sizes of brushes (with base zoom levels 15, 13, and 11).

Las Vegas

When dots overlap

One thing I wish I had realized earlier is to use the full range of color saturation. The maps of tweets use fully saturated color for the most densely geotagged areas, but it actually looks better if you have the color range run from black all the way through fully saturated color in the midrange, ending almost at white in the densest areas. It gives everything a nice glowing look and provides a wider range of distinguishable densities. Compare the glowing Tokyo at the bottom with the flatter one above.

Tokyo

Paying attention to saturation also helps combine colors more naturally. The Gnip maps suffered from being rendered as separate layers that could mask each other. The way it should work instead is that if points of different colors try to occupy the same pixel, each color is a vote for a particular direction on the color wheel, and disputed pixels end up in between, or in the middle of the color wheel in gray. In the case of this mix of tourists and locals in Barcelona, the red and blue mix to purple.

Barcelona

Antialiased drawing lets you see more detail in the data than integer pixel coordinates can convey. The Gnip maps positioned dots to half-pixel increments, but it would have been better to look at even more bits of detail. Taking full advantage of subpixel precision means that even when you are zoomed out pretty far, you can still see the grain of varying density instead of having it get mashed into uniformity by the aggregation to pixels, as happens in the top view of Toronto where all the dots are aligned to pixel boundaries.

Toronto

The last thing I want to mention is that some areas of the Earth have a lot more people and a lot more going on than others do, and it is easy to end up making sparse areas too dim to see or dense areas so bright that you can't see any relative changes. Gamma adjustment seems to do a reasonable job of making areas of different density work on the same map. I had originally thought the adjustment should be logarithmic but ended up using the square root of the density instead.

Here are three versions of the same area, with the Netherlands in the center, Paris near the bottom, and Newcastle near the top. The top image is the most honest of the three, with a completely linear response curve, which means you really can't see much besides the major cities. The middle shows the square root of tweet density, which still highlights the densest areas but also lets you see some context. The bottom, I think, goes too far, with the fourth root of density, obscuring too much of the contrast in the attempt to make everything visible.

Gamma

Making Maki Even Better

$
0
0

Maki, our open-source point-of-interest icon project, has gone through some exciting changes recently.

Better for collaboration

Previously, Maki icons existed in a single source file, bunched together inside one canvas. This was a major pain point when it came to updating and adding new icons, both by designers at MapBox and the greater community.

Thanks to some code wizardry, each icon is now stored in its own SVG source file and compiled into a complete sprite down the line. This makes it incredibly easy and fast to update and contribute new icons.

Ready for your map

We also did work under the hood to turn Maki into an API. Our online map-making tool fetches Maki directly from the source – any updates to Maki are automatically available for use in your markers. The addition of an API for Maki also allows better integration with our iOS SDK. Soon, we will start bundling the Maki icons with the SDK, enabling easier offline use of markers, as well as customizable marker shapes and colors for apps that live completely offline.

We'll be adding new icons frequently and welcome your contributions.

Viewing all 2230 articles
Browse latest View live