Facebook Graph API tweaking: fields

Want to improve the performance of your Facebook Graph API calls? Try trimming your requests down to just the information you need. During recent optimization and testing of some Facebook Graph API client code my colleague Bob determined just how much time could be saved by tuning the API requests. Typical requests result in the return of a default object, for example this URL

https://graph.facebook.com/6547384565867889

returns this album object:

{
  "id": "6547384565867889",
  "from": {
    "name": "John Doe",
    "id": "9834439837"
  }, 
  "name": "Mobile Uploads", 
  "link": "https://www.facebook.com/album.php?fbid=6547384565867889&id=9834439837&aid=743987489", 
  "cover_photo": "748397943639", 
  "privacy": "custom", 
  "count": 1, 
  "type": "album", 
  "created_time": "2011-11-24T16:04:54+0000", 
  "updated_time": "2011-11-24T16:04:55+0000", 
  "can_upload": false
}

The client application doesn't need most of this information, all it needs are the id, name, photo count, and when it was created. By adding the optional fields parameter with a comma separated list of field names to the request:

https://graph.facebook.com/6547384565867889?fields=id,name,count,created_time

an object with only the requested information is returned (apparently Facebook gives us "type" for free):

{
  "id": "6547384565867889", 
  "name": "Mobile Uploads", 
  "count": 1, 
  "created_time": "2011-11-24T16:04:54+0000", 
  "type": "album"
}

Not only does this result in less information being transmitted to the client, more importantly it results in considerably shorter response times from Facebook. It seems that retrieving this information requires significant lookup effort on Facebook's servers and asking for less information means less rummaging through the datastore for all the bits. Bob found that getting 57 albums containing 2,434 photos from his account using the default request took 90 seconds. After adding the fields parameter with only the fields required it took only 40 seconds, less than half the original time! Of course YMMV based on the network, we also found that eliminating likes and comments had the largest effect in reducing response time. If you are working on an application that gets large amounts of data from Facebook it may be worth the effort to consider what information is being provided and only get what the client needs.

Facebook client-side authorization

"OAuth"

This one word can make software developers tremble in fear. Add "Facebook" to it and those same developers start sliding the monitors on their desks together to hide behind, like pioneers circling the wagons at night.

Granted things are better now than they used to be; OAuth 2.0 and the Facebook Graph API are much easier to work with than the old REST API with OAuth 1.0 (Remember signatures? No? You're lucky). However authorizing an app or logging in with Facebook can still be a challenge. I really wanted to be able authenticate as a Facebook app in a single HTML file with javascript interspersed just to see if it could be done in one file. What I've found is, that can't be done, but something almost like it can be.

Originally I had one page that would navigate based on the current Facebook login state which was determined by the parameters passed to the page. Roughly the Facebook authorization workflow was:

1. App page initially displayed in a Facebook iframe: no access_token parameter.
2. Navigate the browser window (not the iframe) to the OAuth dialog. Pass the page URL as the redirect parameter.
3. User authorizes / logs in.
4. App page is now displayed in the browser window with the access_token (remember we left the Facebook iframe when navigating to the oauth dialog), navigate to the app's Facebook canvas page and include the access_token in the canvas page URL.
5. App page is displayed in the Facebook iframe and now it has the access_token.

This could work accept the Chrome browser only allows two redirects and this scheme requires three. After flailing with various fruitless searches ("facebook" + "javascript" + "login" does not yield a lot of helpful hits) I found a cached copy of this page that had the missing piece I needed, it used the Facebook Javascript SDK to determine if a user was logged in or not. I changed my code to use this method and then navigate to the oauth dialog or launch the app based on the result. This is the final workflow:

1. App page initially displayed in a Facebook iframe: app not authorized / user not logged in.
2. Navigate the browser window (not the iframe) to the OAuth dialog. Pass the app canvas page URL (not the URL of the HTML file) as the redirect parameter.
3. User authorizes / logs in.
4. App page is displayed in the Facebook iframe: app authorized / user logged in. Launch the app.

So there you go, authorizing an app and logging into Facebook done entirely client-side with one HTML file that includes the Facebook Javacript SDK (full disclosure: I also included jQuery for convenience though the code could be rewritten to avoid using it). You can try this code or use it by getting the source from this github gist (easy) or copy and paste it from the horribly clipped embedded view below (hard).

Animated menu for a web page

I'm working on revamping Smashingline and wanted to make the menu at the top a little more interesting. I decided to create a menu with two "pointers." One to point at the item your cursor is currently over (highlighted) and another to indicate the current item (selected). The menu was made with HTML5, Javascript, and CSS. I also used jQuery to animate various elements. I was inspired by the page eCSspert to use only HTML and CSS for the graphic elements.

Main

Search

Contact

Essentially the menu consists of a "menu bar" div with "menu item" divs above it. When the cursor passes over the bar its mouseover event is fired and the highlight pointer is shown. When the cursor moves from one menu item to the next the animation is triggered causing the pointer to move. One of the problems with this approach is that when transitioning from one item div to another the bar div fires a mouseout event even though the location of the cursor is still "inside" the bar div. Event propagation is probably one of the most common problems when working with mouseenter and mouseout events. Fortunately jQuery has solved this problem by implementing its own mouseleave event which was introduced by Microsoft and is supported natively in IE. The mouseleave event is only fired when the cursor actually exits the region of an element, no matter the number of child elements. Very handy.

The pointers are constructed from HTML div elements and CSS. Generally each of the two pointers has a base div that is the same width as the menu items. The div's height and setting its overflow property to hidden dictates how much of child div which is rotated 45 degrees is visible.

Finally since you may want the web page to take some action once the user makes a choice a callback can be set on the MenuMarker class. This function is invoked once the animation is finished. The demo here displays an annoying alert dialog with the the text from the menu item displayed.

The source code (in three separate files) is available for download from a Github Gist. These can also be combined into a single HTML file and loaded dynamically as needed using the jQuery load method.

Creating a Nexus S Wallpaper

So you've got a snazzy Nexus S Android phone and you want to use one of your own photos as wallpaper? Well it certainly can be easy enough, open a picture using the Gallery application then from the menu choose More - Save As - Wallpaper. Drag the crop box around and resize it by touching a side if necessary and your picture becomes the wallpaper. But since only a vertical slice from the wallpaper is shown on screen maybe you couldn't select as precisely as you wanted and an important part of the picture isn't visible. Or maybe you want more control over the final image, like resizing it to the actual wallpaper dimensions.

The first thing you need to understand is that only a small region of the wallpaper is shown at any time. This is because there are five different "desktops" that can be scrolled side-to-side. As the user swipes their finger across the screen the other desktops are shown and the wallpaper image scrolls slightly along with the icons and widgets. Because of this the wallpaper image needs to be wide enough to cover all five of the desktops. It turns out that the final wallpaper size is 960 by 800 pixels. The screen size is 480 by 800, so with each finger swipe the image moves by 1/5.

The starting picture is a 3:2 aspect ratio image. It has been resized so that it is 1200 pixels wide by 800 pixels high.

Original 3:2 aspect ratio photo

The first thing is to determine what the center screen should look like.

Selection indicating the region that will be visible in the center desktop. Note that the selection is not centered in the picture.

Once that's done the final wall paper region can be defined by expanding the selection so that it's 960 pixels wide.

The final wallpaper 960 pixel wide selection

Finally crop the image and save it to your Nexus S. Use the Gallery application to view the image and choose the Wallpaper menu item; for some reason Google decided that the crop box will never select the whole image, so you'll need to move and resize it to select the whole picture. There you go, your own custom wallpaper.

The final wallpaper image.
This photo is licensed under a Creative Commons Attribution-ShareAlike 3.0 Unported License.