Earthware Blog

For the third blog post in our series comparing Google Maps and Bing Maps we are focusing on the size of download users require to use these platforms and the speed which the core code and map imagery to delivered.

Download size

Both mapping platforms use a large amount of javascript and css that is required to get the actual base mapping working before the user even gets to the code you as a developer write for your application. Below you can see a comparison of the download size in kb (when gzipped) required for each platform, slightly complicated by the fact Google Maps currently have two current APIs ver 2 and ver 3 (as ver 3 still has features missing that are available in ver 2).

As you can see Bing Map’s javascript is considerably larger. This is, in our opinion, due to Bing Maps having a lot more of its features built into the main javascript file (e.g. client side clustering, drag-zoom, labelled markers, speed optimized marker addition) rather than in optional community based js files like Google’s GMap Utility Library.

We have deliberately not compared (or included in our calculations above) the image tiles used to show the actual maps in both api’s as both use the same formats for these and they are entirely dependent on the imagery at each location.

Delivery speeds

As the vast majority of data used in your mapping applications is the image tiles used to create the map the speed which these image files are delivered to your users worldwide is arguably more important than the initial javascript download size. For example one 256×256 pixel square road map tile is typically ~20kb so these very quickly become the biggest data download for your application.

Bing Maps have openly discussed their recent implementation of a Content Delivery Network (CDN) for Bing Maps which means all the Bing Maps files are delivered by a network of servers worldwide and will be delivered by the server closest to your users therefore increasing the delivery speeds. We cannot find any official word from Google about whether Google Maps is doing the same but we would be surprised if they are not.

Help us test delivery speeds

As we promised to try and use objective measurements when comparing mapping API’s we need your help to test delivery speeds all over the world. If you can add your location and the speed of delivery you get from both Google Maps and Bing Maps (see how to do this below) to the comments on this article we will update the article and add the results for everyone to see. Obviously the delivery speed will vary on your connection speed but we will still be able to compare the difference between loading a Google Maps tile and loading a Bing Maps tile on the same connection.

How to test delivery speeds

To take part in this experiment you will need to download Firefox and the Firebug plug-in. Follow the steps below and post your location and results in the comments below.

1. Open a new browser tab and then open firebug (tools->firebug->open firebug)
2. Click on the “Net” tab in firebug
3. Load the following url in the tab:
   http://mt1.google.com/vt/v=ap.115&x=233&y=154&z=8
4. You should see a graph like the one below showing you the filesize (16kb) and how long it took to load (191ms in our case).

   As suggested by readers in the comments we recommend you take the average time taken of at least three tests, for each test ensure you hold down the SHIFT key and click Firefox’s refresh button so that it does not read the tile from the browser cache (ensure firebug does not say 16kb from cache).
   

   

5. Now to get the download speed in kb/s simply divide the filesize (16kb) by the download time then multiple by 1000, giving in our example a download speed of: (16 / 191) * 1000 = 83.7 kb/s
   
6. Now repeat for this Bing url:
   http://ecn.t3.tiles.virtualearth.net/tiles/r0231113.png?g=392
   
7. Finally post your results and location below.

Conclusions

We will have to wait until we have a number of results in from around the world until we can draw some useful conclusions to this blog post. Until then we can say if initial filesize is really important to your users (for example those on mobile connections) then you may wish to look at the Google Maps ver 3, however initial file size is far from the whole picture so it would be best to wait to see the results of the delivery speed tests before reading much more into this.

This entry was posted on Thursday, December 31st, 2009 at 3:55 pm and is filed under Bing Maps, General Mapping, Google Maps. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.