TileMill 0.9.0 Released with Full Windows Support (and plugins) | Hacker News Reader

TileMill 0.9.0 Released with Full Windows Support (and plugins)

mapbox.com

70 pointsAdrianRossouw21 comments

I am a huge, huge fan of TileMill. I've been making a mobile HTML5-based app that had used the Google Maps API, which is slow on even the most recent phones, and the Custom Tiles, which are limited and now have the potential to be very expensive.

So I ripped the whole lot out and replaced it with some custom TileMill-built map tiles and CloudMade's leaflet.js mapping library. It looks ten times better, works like a native map, and I can even control things like local caching of map tiles, which the Google licenses restrict you from doing. I have the tiles served from an EC2 instance running TileStream, and a CloudFront CDN to cut down on costs- it works great.

That said, it isn't for everyone- I'm lucky in that my app currently NYC-specific, so I only needed NYC map tiles. Those alone took ~9 hours to generate (x2 for separate 'retina display' tiles). I dread to think how long generating a whole world would take.

There's a preview of the tiles here, if anyone is interested: http://www.taxono.my/ . The process isn't quite seamless yet- there were some bugs related to street labels that I couldn't fix (and appear to be related to the underlying Mapnik library). This post has reminded me that I really need to write a full rundown of my experience, but in case you couldn't tell, I'm impressed.

We're aware of the labeling issues in recent Mapnik and will land a fix soon: https://github.com/mapnik/mapnik/issues/1017

That's awesome. Thanks!

I'm curious about your experiences with what sounds like a JS maps frontend in an iOS app, versus native Objective-C. Could you elaborate a bit on your decision there?

Decision in terms of why I chose JS over Obj-C?

I had intended to make as much of the app cross-platform as possible, so I was predisposed to use HTML5/JS in the first place- 95% of the app will be directly transferable to Android. I'm also not sure if you can use custom map tiles with the native map control in iOS- I've certainly never seen anyone do so. Although I'm not yet taking advantage of it in the app, the ability to cache map tiles locally is also a great plus.

I'm extremely happy with the end result- it really does work as well as the native map control. In fact, I found a demo here:

http://leaflet.cloudmade.com/examples/mobile-example.html

It's a little more work to get it looking as good, but nothing some CSS3 can't fix (box shadows, rounded corners and the like). The story is a little different on Android as pinch-to-zoom doesn't work, but I think that's a WebKit flaw, rather than a leaflet.js flaw.

You can't use the native map control (MapKit) but you can use an alternative:

https://github.com/mapbox/route-me https://github.com/mapbox/mbtiles-ios-example https://github.com/mapbox/tilestream-ios-example

The main advantage over JS is the ability to go offline, as well as performance. But that's largely negated if you need to go cross-platform as you do, where I think JS does make a lot of sense.

I've played around with extending the T.TileLayer to save remote tiles then try to load then using file:// in the future. It mostly worked, but I didn't have time to totally tweak it- my app is being entered into a competition with a deadline. I fully intend to investigate offline JS maps in the future, though.

Rendering a full planet isn't bad, I did up to zoom level 15 in 48 hours. 16 would take probably 48 hours by itself, and 17 would push into another week is my guess. But, mapnik2 is fast enough to render those tiles on the fly without a huge performance hit, about 1 second per metatile (enough to cover a couple small map pans on my big monitor). I am running it on 16gb of ram and a raid10 with 6 2tb drives. My next upgrade will be going full SSD and 196gb of ram, but that will also be a nominatim and pgrouting install, so the ram and SSD's are very important for importing the DB and indexing it inside of 1 month :) Just a tileserver isn't too bad at all.

> Those alone took ~9 hours to generate (x2 for separate 'retina display' tiles). I dread to think how long generating a whole world would take.

Map tiles should almost be a canonical example of a ridiculously parallelizable problem. Even if the data access itself is your bottleneck, you can fix that by simply bringing up multiple servers and replicating it.

Unless your map data is so dense that it takes over nine hours to generate a single tile, you can always scale the work out. It's only a matter of justifying the organizational time and whatever expenses are incurred for the extra equipment.

    > Leveraging the architecture of Bones, our server side
    > Backbone.js framework, and the package management API 
    > of npm, the node package manager, the TileMill UI can 
    > now be extended and modified using standalone, 
    > JavaScript plugins.

I'd love to hear more about how exactly this works, if you don't mind sharing. Installing JavaScript plugins into a desktop app via npm sounds like a very neat trick.

We use Bones as our code loader in this case though depending on your use case a straight node `require()` might be simple enough.

The npm JS api is quite usable, so it was simply a matter of establishing a directory for plugins (in this case `~/.tilemill/node_modules`) and using the npm API to install, uninstall, etc. plugin packages to there. We identify plugin packages by the `tilemill` tag in the npm repo and each plugin is required to declare their compatibility with a particular tilemill version at the `engines` key.

I always enjoy anything that can help me avoid using external services such as Google Maps. I'm not one who hates these services, I think they are great. But what if they get taken down? All the maps I have out there on sites are now down. It's a lot more reassuring when I can make my own.

That's one reason I like using something like Openlayers for all my web mapping needs. This way I can use Google Maps for as long as it's practical, but if something changes I can change map provider to, for example, Bing or something OSM based by changing only a couple of lines of code.

That's the beauty of TileMill. They offer the Mapbox.com service but you aren't tied to their platform. You can use something like http://tilestache.org/ to serve titles made in TileMill. I'm excited to see this update as I'm using TileMill for a project with a big dataset and I hope this will speed it up. Kudos TileMill team! :)

In fact, there are quite a few other implementations, for various python/node/whathaveyou enthusiasts: https://github.com/mapbox/mbtiles-spec/wiki/Implementations

(of course, it'd be nice if people hosted on MapBox, as it's the main driver of TileMill development :)

Do you guys have a donation link somewhere? I'm using TileMill for a small project and The smallest Mapbox hosting is way overkill. I wouldn't mind donating some money as it's an awesome product and TileMill is wonderful!

Edit: The initial dataset is huge but the implemented map is only a small subset.

I think this will be a huge advantage for Mapbox. Most GIS professionals use Windows (due to ArcGIS's requirement), and now it's easy to install TileMill on their Windows machines.

Can I just mention Maperitive

http://maperitive.net/

You fire it up, draw a bounding box around the area you want to work with, it downloads that area from open street map as raw data which you can then theme using rule files.

Plus all its commands have command-line equivalents which means you can automate much of it, so far it's been a joy to work with.

Very interested in hearing about your experiences developing a desktop app with node.js.

Anyone else having trouble installing/running the .exe?

I've installed okay, but when i click "Start TileMill" nothing happens :(

scratch that, uninstall and reinstall fixed it.

Reply on news.ycombinator.com