Category Archives: iOS

GIDS 2014

A couple of weeks ago, I had the privilege of speaking at GIDS (the Great Indian Developer Summit) 2014 in Bangalore. GIDS is a multi-day event, with each day dedicated to a particular theme or topic. I spoke on the Wednesday (the Web/Mobile day) and had three sessions covering mobile app development using iBeacons, tips and tricks using AngularJS, and automated Web testing using JavaScript.

Thanks to everyone who came to each of the sessions. I really enjoyed the feedback and audience participation, and look forward to similar sessions in the future. I’ve included links to the three decks below on Slideshare, and have posted all of the source code here.

iBeacon Demo at Gartner AADI

Last week, I joined my colleagues from Neudesic at the Gartner AADI (Application Architecture Development & Integration) Summit in Las Vegas. One of the neat things that we were showing was a demo of iBeacon technology, in order to help locate and pinpoint individuals within indoor environments where GPS is not an option. 

To achieve this, we placed a selection of iBeacons around the expo floor (for the iBeacons we used RedBear’s BLE Mini boards, flashed with their iBeacon firmware).

Red Bear BLE Mini

Using Bluetooth LE (a.k.a. Bluetooth Smart), each beacon transmits a unique ID (made up of a major and minor value) on a particular “frequency” (known as a proximity UUID). iBeacon supported mobile applications are then able to “tune in” to this proximity UUID, and based on the power signal for each beacon, determine which area of the floor a user is located closest to.


Gartner AADI screenshot

Using this information we created a mobile app that reports on all of the beacons within range on the expo floor. As you can see in the above screenshot, we have 5 iBeacons in range, listed in order of proximity, with an estimated distance calculated by the power signal. As you can likely gather from the data, we were located in the Neudesic booth at the time this screenshot was taken, with the theater, and a selection of other booths in range.

For the show, we developed two versions of the application – one for iOS and one for Android. Both are native mobile applications written using Xamarin, using CLLocationManager for iOS, and Radius Network’s binding for supporting iBeacon on Android. The Radius implementation is especially interesting in that the beacon detection runs as a background service, polling for new iBeacons and raising intents to the UI as necessary (even though Bluetooth LE is a lot more responsible with power vs. regular Bluetooth, we still need to be careful not to kill the battery when using this however).

While this is neat to show just as a mobile application, we wanted to take it one step further and demonstrate how this could be applied in the real world with Neuron, an ESB product from Neudesic. For our expo demo, Neuron provided a backbone infrastructure to allow the mobile app to resolve a name for a beacon ID (think of his like DNS for iBeacons!) and also give a scalable way for users to “check in” to an iBeacon the same way that they would check in via Foursquare or Facebook.

Neuron process flow

As shown above, we developed a process flow using Neuron to accept incoming messages from the mobile application, and then provided logic to determine whether the device was trying to acquire a name for a beacon, or whether the user had walked in to the area of a beacon and wanted to “check in”. The benefit for using Neuron in this situation vs. just propping up a single web service is that our application can now scale to hundreds of thousands of concurrent mobile clients without needing to make any major adjustments on the server-side.

If you were able to stop by the booth at the Gartner summit, I hope you enjoyed the demo and the conversation. If you weren’t able to attend, but would like to know more about how Neudesic is working with various organizations on iBeacon technology and/or Neuron, feel free to drop me a line.

Designing a Web API for Mobile Apps

At Neudesic, almost every mobile project that we’ve been involved in has had some dependency on a Web API.  This has included APIs that we’ve had to create from scratch, as well as ones that already exist that the app has to consume.

As you can imagine, over the past few years, we’ve seen a fair share of good and bad API design.  In this blog post, I wanted to share some of my observations, thoughts, and questions that I ask of a well designed Web API, especially one that will be called from a mobile app.

1.  Are the basics being met?

First, the basics.  Without question, every Web API should be stateless.  There should be no session state, cookies, or server-side values used to hold any state of any kind.  Adding state adds complexity, and limits the ability for the API to scale, which for a mobile application that could reach millions of users is something that we want to avoid.

Also, endpoints of the API should be exposed through SSL by default.  SSL is easy to setup, performant, and should be enabled for any API that we either consume or create.  As you may have observed, using SSL as default seems to be a direction in which many other APIs are heading.

2.  How do we authenticate with the API?

Authentication is critical to a Web API, and (in my opinion) is one of the common pitfalls.  As a rule, user credentials should never be passed as part an API call.  Take the following example:

GET /accounts?username=simon&password=simon

Doing this is bad for three reasons:

Firstly, although the URL gets encapsulated as part of the HTTPS session, it is likely still visible in any logs on the Web server.  This often also applies to user credentials passed as HTTP headers.

Secondly, it makes debugging with users and other developers awkward (because you often need to ask for their username and password).  This is especially bad if the credentials are corporate accounts used for other systems.

Finally, and most importantly, these types of credentials typically have a long shelf life.  If the call is ever compromised, there’s a good chance that it can be replayed back to the service up until the password is changed, which could be many months (if ever at all).

To overcome this, some APIs use an application key or some other token derived from a HMAC algorithm.  This may work for some scenarios, but unfortunately if the key is exposed, it can be difficult to revoke.  This is especially true if the key has been embedded in a mobile app running on thousands of devices.

Fortunately, to overcome both of these issues, there is OAuth 2.0.  OAuth 2.0 works by having the user pass a set of credentials (typically a username and password) to an authentication service.  Assuming the credentials are valid, the user/application/consumer receives back an access token.  This access token is then passed as a HTTP Authorization Header to the Web API to verify the authenticity of the request.  Moreover, this access token has an expiry (I find an hour to be a good time frame) so that if someone were to get hold of the token, their usage of the API is limited to this timeframe.  (You can read up much more about OAuth 2.0 here)

There’s no doubt that implementing OAuth 2.0 involves more work, including setting up an authentication API and handing the lifetime of the token, but the end result is an API that is more secure and will also reflect the security model used by many others (e.g. Facebook, Google) – which means that you can even use these third parties as identity providers if you so choose.

3.  Is the API really using REST?

I’ve seen many examples of people thinking that they have a REST API when really they don’t.  Correct use of REST is about nouns, not verbs.  Let’s take this URL for example:

GET /GetAccountBalance?account_id=1234

Although the above URL is accessed over HTTP, it’s really not a REST API.

Using REST to it’s true intention means combining HTTP VERBS together with nouns or entities in the URL that represent the data you are exposing.  Instead of the previous URL, a correct REST syntax would be the following:

GET /accounts/1234

To create a new account we would use a HTTP PUT (together with a payload with the new account information)

PUT /accounts

To delete an account, we would use:

DELETE /accounts/1234

This noun-based approach can also work with hierarchical data.  For example:

GET /accounts/1234/transactions/50

Will return the a transaction (with id of 50) for Account 1234.

Overall, a good understanding of REST, together with a focus on exposing nouns instead of functional methods will go a long way to create a more more usable and accepted API.

4. How should we consume this API?

If you are dealing with an API that exposes a lot of entities, in addition to exposing generic REST endpoints, there are typically six things also worth considering:  Sorting, Searching, Filtering, Pagination, Helpers, and Partial Responses.

Sorting.  Should the API return a sorted list of data to the consumer/application?  If so, a sort parameter on the noun can be useful:

GET /accounts?sort=id

GET /accounts?sort=-id

As shown above, a leading hyphen to indicate ascending or descending sort order can be a great timesaver (and often negates another query string parameter for sort order).

Searching.  Similar to sorting, providing a way for consumers to search entities can be useful.  The use of the “q” parameter for search is somewhat standard:

GET /api/accounts?q=acme

Filtering.  Another useful pivot for any REST based API is the ability to filter on a particular noun.

GET /accounts?filter=balance%3E500

(You can choose to use the URL encoded character for > as shown above, or I’ve seen many other APIs use gt, lt query parameters).

Pagination.  A type of filtering, especially useful for large datasets.

GET /accounts?limit=50&offset=25

This above call will get the next 50 accounts, starting at the 25th entry.

Helpers.  With many APIs there are a number of common requests.  Maybe it’s the list of top ten customers, or the best sales people of the month. Putting these common requests as “helpers” into the API can be very useful for consumers, and can also help reduce the “chattiness” of the API by reducing the number of repeat requests.

GET /accounts/top-ten

Partial responses.  Finally, many consumers of the API (especially mobile applications) will want only a summary set of data.  This can be useful to build a list of items (in a master/detail view), without having to send the entire details for each item.

GET /accounts?fields=id,name,balance

Of course all of the above parameters can be combined as required.

GET /accounts?fields=id,name&sort=id&limit=100&offset=50

5.   What will the API return?

For the majority of APIs, especially those that will be consumed from a mobile application over a potentially slow connection, returning JSON is always good practice.  Compared to XML, data returned in JSON format will likely be more lightweight, and will require less parsing and processor overhead on the device.

With that said, there are cases where other formats might be required – for example, a legacy system that is already expecting data in XML.  In which case, you might want to consider allowing the consumer to specify what type of data to return either through the HTTP Accept header or through a URL action (useful if you anticipate doing a lot of debugging).

GET /accounts?format=xml

There has also been a lot of talk recently about HATEOAS (Hypermedia As The Engine Of Application State), a term coined by Roy Fielding.  While there are many articles and presentations that explain Roy’s initial intentions, for the purpose of this blog post (and my own sanity), HATEOAS in a Web API referring to the use of links that instruct the consumer where to go for related information.

For example, let’s imagine we made the following API call:

GET /accounts/1234

We might receive the following response:

{ "account_id" : "1234", "balance" : "100.90" }

With a “HATEOAS-compliant” Web API, we may also receive embedded links.  For example:

{ "account_id" : "1234", "balance" : "100.90", { "_links" : { "transactions" : { "href" : "/accounts/1234/transactions" } } } }

As you can see above, the API returns the data for the account, but also returns a link to the API call that will return all of the transactions for that account.  Think of these links as helping the consumer navigate to other related API calls.  (Incidentally there are a number of JSON formats for doing this, although I would recommend JSON HAL

6.  Are the methods of the API consistent? 

While it’s difficult to recommend what you should name your methods and other parts of your API, the key to success is often consistency.  For example, if you have the endpoints for your accounts here:

GET /accounts

For your invoices, it would be silly to have them here:

GET /order_entry/ledger/invoices_full

In an ideal world (and even one without HATEOAS!), a user should be able to guess what the API should be based on previous usage.  Keeping the paths and names consistent are key to making this happen.

Related to this, choosing the right case for APIs can be very important.  Having these two apis:

GET /accounts

GET /Invoices

will likely lead to issues because of the case mismatch on the entity name.  My recommendation is to use lowercase throughout (then there is no ambiguity) and to use snake case to conjoin words.  For example:

GET /customer_details

Spinal case (using hyphens) is also acceptable, but if you are doing a lot of descending sorting, you may want to be careful.

Finally, in terms of consistency, it’s always nice to be consistent with pluralization:

GET /accounts/1234/invoice

Assuming there are more than one invoice per account, this could also run people into trouble.  I would recommend deferring everything to plural to ensure consistency.

GET /accounts/1234/invoices

7.  How is the API versioned?

Versioning is important, especially if there are breaking changes in production environments.  There are a couple of ways to achieve this:

For large volume APIs where version consistency is critical, I would recommend placing the version information as part of the API call.

GET /v1.0.0/accounts

Versioning by using the URL makes it very explicit as to the version that the consumer is using.

For less critical systems, or for APIs where breaking changes are going to be rare, I would recommend that consumers pass an optional version number as part of the HTTP header.  If the version number is passed as part of the post, the consumer gets a specific versioned response, otherwise they’ll be receiving the latest version.

In addition to version numbers, I always like to see specific environments affiliated with the URL.  This is most easily done as part of the host subdomain, as it will likely correspond with the physical or virtual machine that the API is hosted from:




The above makes it very clear whether I’m hitting the development, UAT, or production version of the APIs when I make my calls.

8.  How is the API documented?

If you have a well designed API, you do not need to spend hours of time documenting the API in a Word document.  If anything you are going to end up with a Word document that will become quickly out of date.  In terms of documentation, there are two things that I find invaluable:

Firstly, mandatory and optional methods and parameters should be documented.  It’s important that consumers understand what they can and cannot pass to the API.  What’s nice is that this documentation can typically be auto generated from the method signatures or comment blocks (which will keep your documentation in sync with your code).

Secondly, sample code to show how to call the API.  A few sample calls for each method can be a life saver and much more valuable than reams of documents.  In these samples, show how the request should be formatted and what the response looks like.

9.  What does the API return when things go wrong?

Returning useful error messages to consumers of your API is really important.  You should think about two types of error messages – ones that can be expressed with HTTP result codes, and ones that cannot.

For the ones that can be expressed through a result code, simply return the result code with an optional body of information (in JSON format).  For example, if the access token has expired, return a 401 HTTP error code, and maybe some JSON payload to help debugging.  Also, if any part of the system is down (e.g. the database connection can’t be established), I would recommend returning a 500 for clarity.  With any HTTP result code, remember to pass the right one.  A good rule of thumb is that result codes in the 400′s typically indicate an error with the client, whereas codes in the 500′s means that something has gone wrong on the server.

For errors that can’t be expressed through a HTTP result code, you should be returning a JSON payload that contains a minimum of two pieces of data – a unique error code for the issue, and a message aimed for the consumer/developer of the application to help them debug.  For example, if the consumer tried to create a new account without including all of the mandatory fields, this would be a useful error to return:

{ "error" : 16, "debug" : "The mandatory field for balance was missing in the request" }

Some recommend returning a user message also, which can be useful.  Others use the error code to create their own message to pass to the user.

10.  Finally, what else should we be thinking about for using the API in production?

There are many considerations for using APIs in production – here are just a few:

How are you going to make money from your API?  Are you thinking about a transaction cost per call, freemium, capped model, or something else?  Many of these systems are going to require some kind of API metering – which isn’t necessarily hard, but is definitely something else to consider.

Are you going to rate limit your API?  How are you going to prevent a rogue customer, application, or process, who wishes to call your API hundreds of thousands of times?  Fortunately, there are answers to this – including RFC6585 which specifically deals with rate limiting – but again, something that you should be considering.

Should your API provide cached results? Is this something that can improve the performance for your consumers, and also prevent unnecessary calls to back end databases and other systems?

How is your API going to work over a low bandwidth connection?  Your API might work great on your FIOS line here in the US, but do consumers get the same experience when calling the API from a J2ME application from a cell phone in the middle of Africa?  There are many ways to simulate throttled connections and this should be something that is definitely worth testing for.

Finally, and arguably most importantly, how can you get everything to go through your API?  Instead of thinking of an API as a companion to your existing Web-based applications, what would it take to push everything through this API – treating the API as a single source of truth?  It might sound scary, and undoubtedly it’s additional work to have everything using the API – but a single API that every application uses has the potential to offer a lot of re-use and sharing as you develop your API over time.

jQuery Mobile and AngularJS Working Together

Both jQuery Mobile (jQM) and AngularJS are awesome at what they do, but getting them to play nicely together can be tricky.  As you may have discovered, both want to manipulate the URL/routes and DOM such that it’s very easily to get them in conflict.  Having been through this recently, I wanted to share some recommendations to get them working together:

Load jQM libs before AngularJS

Because both frameworks heavily manipulate the DOM, it’s important to get the load order right.  I found that loading AngularJS first led to some interesting (and annoying!) UI functionality.  The correct order should be jQuery first, followed by jQM, and then AngularJS:

<script src=""></script>
<script src=""></script>
<script src=""></script>
<script src=""></script>

Let jQM handle the URL routing

I’m likely to get flamed for this by MV* purists, but I recommend letting jQM handle all of the URL routing – and not using AngularJS for any routing.  Firstly, I spent a lot of time trying it the other way (disabling routing for jQM, and configuring various routes, templates, partial files, etc.).  Even when it did work, it was just a mess – it looked like someone had taken a shotgun to my jQM app and blown it into several pieces.  Secondly, I would argue that URL routing really shouldn’t be a primary consideration for a mobile Web app.  The app is more likely launched by an icon on the home screen vs. a search or link with any type of query string.  Even if it does, a simple check for a null scope is all that’s required.

Create a single Angular controller for a group of jQM pages

Conforming to #2 means that you can create a single controller that spans a number of individual jQM pages.  This usage results in very elegant single HTML page together with a single controller – yet has the advantage of offering multiple pages to the user.

To demonstrate this in more detail, and because you can’t have enough Todo list apps, I’ve put together a sample using jQM, AngularJS, speaking to a service using Node, Mongoose, and MongoDB.  (To run, you’ll need a local Mongo DB called “todo” with a collection called “tasks”).  It definitely shows the power of both frameworks running together.  In just 75 lines of HTML and 29 lines of JavaScript for the controller, I have a mobile app with full CRUD support.  Hope you find it useful.


On-Demand VPN using OpenVPN for iOS

Many large organizations have existing VPNs in place, and the rise in development of enterprise mobile applications often requires that apps running outside the firewall need to “VPN in” before they are allowed to access any corporate resources.

While enabling a VPN is fairly easy on most mobile devices, I’ve found that many IT organizations would like to automate this such that the device automatically connects to the VPN when the Enterprise App or an internal Website is launched. Having gone through this with a couple of customers, I’ve put this post together to highlight how this can be done.

Which VPN types support On Demand for iOS?

iOS 4.2 introduced features designed for the enterprise. These features included several VPN clients (L2TP, PPTP, Cisco IPSEC), many of which can be configured to automatically initiate the connection based upon DNS requests for certain domains.

While this on demand feature has been around for a while, it does have a couple of caveats. Firstly, it’s only available for SSL based VPNs. An SSL-based VPN is a connection that relies on a client/server side certificate for authentication. Having an SSL-based VPN means that the connection can be established without asking the user for credentials (e.g. a password or RSA token). Because an on-demand VPN could be connecting and disconnecting every few minutes, this makes sense as prompting the user could cause a jarring user experience. Secondly, the on demand VPN can only be setup using the iPhone Configuration utility. While you can create a VPN connection on iOS devices (under general/settings), it’s not possible to create an SSL-based VPN or specify any of the on demand domains.

With this in mind, I’ll be showing how to establish an On Demand VPN using OpenVPN and the OpenVPN client for iOS that was released earlier this year.

Server-Side Setup

The first step is of course to setup an OpenVPN server. If you don’t have access to one already, there are plenty of guides to help you set one up. If you don’t have access to a spare machine on the network, you can also use a Linux VM instance hosted on EC2, Azure, or any other hosting provider. This article is a pretty good guide for setting this up on EC2 and the instructions should be able to be tailored to other hosting environments.

Generating Certficates and Keys

Once you have your OpenVPN server running, you’ll need to either import or generate certificates and keys required to establish the connection. If you don’t have access to certificates already, you can find some great documentation here on how to create your own using the EasyRSA project on Github. (Scroll down and look for instructions on using easy-rsa)

These are the certificates you’ll need to generate using EasyRSA:

ca.crt (CA certificate required for both server and client)
dh2048pem (A 2048bit Diffie Hellman key that is required on the server)
server.crt (The server-side certificate)
server.key (The key for the server-side certificate)
client1.crt (The client-side certificate)
client1.key (The key for the client-side certificate)

Configuring OpenVPN 

Place the certificates in a folder called keys (under /etc/openvpn/keys) and modify your /etc/openvpn.conf file to look similar to the following:

port 443
proto tcp-server
dev tun
keepalive 10 120
verb 3
push "ifconfig"
push "redirect-gateway"
dh keys/dh2048.pem
ca keys/ca.crt
cert keys/server.crt
key keys/server.key

This should be fairly straightforward to understand from the documentation, especially if you’ve used OpenVPN before. As you can see, we are creating a peer to peer connection using two private IP addresses ( and Don’t worry if these don’t match your internal network on the OpenVPN box – they don’t need to in order to get this up and running.

For the purposes of this article, I have the VPN running on a tcp:443 connect, but feel free to adjust the protocol and port to match your own environment (assuming that you have the necessary ports open on your Firewall).

Once you have the server configured, you can start OpenVPN:

openvpn --config /etc/openvpn/openvpn.conf

Assuming that the server starts OK, you can move to the next step.

Installing and Configuring the Client

The OpenVPN client can be found on the AppStore. At the time of writing it’s version 1.0 build 47, which has a few bugs here and here, but still seems to work well.

After you have this installed, go ahead and install the iPhone Configuration Utility on your Mac. This can be found on the Enterprise iPhone Support page ( Don’t worry – although it’s called the iPhone Configuration Utility, this will also work for setting up a VPN on an iPad device also.

After launching the utility, create a new configuration profile. In the general tab, enter the mandatory fields for name and identifier. Then, click on the credentials tab. We need to import the client certificate and key that was generated earlier using EasyRSA.

Unfortunately, the iPhone Configuration Utility (which we’ll call IPCU from now on) doesn’t support importing .CRT and .KEY files directly, so we’ll need to generate a PKCS#12 file for use here. To do this, in the Terminal locate your client certificate and key files, and run the following command:

openssl pkcs12 -export -in client1.crt -inkey client1.key -out client1.p12

The result of this should be a .p12 file, that you can now import into the credentials section of IPCU.

Note:  When you generate the .p12 file, you will be asked for an export password for the file. Enter something, and use the same password in IPCU (there is a field just under the certificate picture). If you don’t enter a password, you will likely get a profile error when you try to deploy this to the device.

With this done, now navigate to the VPN tab and create a new VPN connection. IPCU doesn’t support including an OpenVPN config file, so we’ll have to create the majority of the settings in this tab – this is where things get fun :-)

Give the connection a name, and select “Custom SSL” for the connection type. For the identifier, use the following:


This is telling the VPN client that a specific bundle ID (the OpenVPN app) should be used for this connection.

Next, in the Server field, enter the DNS name or IP address for your OpenVPN server.

The Custom Data contains keys and strings that replicate what would have normally gone into a config.ovpn file. These are the entries that you will need:

ca – This is a tricky one to get right. We can’t point it to a ca.crt file, because there is no way of bundling a file using the IPCU tool. To overcome this, open the ca.crt file in TextEdit and replace all of the newline/carriage returns with n. What you should end up with is a single (long) line of text that starts with —–BEGIN CERTIFICATE—– and has several lines delimitered by n ending with —–END CERTIFICATE—–. Once you have this, paste this entire line into the value for the ca key.

comp-lzo - enter the key, but you don’t have to give it a value.

dev – set this to tun

port – set to 443

proto – set to tcp

(Again if you are using something other than TCP port 443, feel free to change)

remote – set to [your server name] 443

(replace your server name with the DNS name or IP address of your OpenVPN server)

verb – set to 3

Here’s a quick screenshot of my profile:


That should wrap up the custom data piece.

Under User Authentication, select Certificate – and then under Credential select the client certificate from the drop down.

Finally, check the “Enable VPN On Demand” box, and in the below table enter the domain names or IP addresses for the hosts that VPN is required for. For this tutorial, I have a single entry for with “Always Establish”.

That’s it!

This configuration should be enough to get VPN On Demand working with OpenVPN on the iOS client. To test, deploy the configuration profile to the device. Assuming that this works, open a browser and browse to – this should invoke the VPN connection and forward the requests to the OpenVPN server. If you have a test page running on the OpenVPN server then this should be displayed in the browser.

If it didn’t work…

There’s a good chance that things won’t work first time, just due the complexity of the setup.  Here are some of the common tips/tricks that I found during the process.

- Firstly, the console window in IPCU will be your best friend. The most common error I saw was related to the CA certificate. If you get these, it likely means that your CA entry in IPCU is wrong. Go back and make sure you’ve correctly replaced the linebreaks with n’s and retest.

- There are a few “bugettes” in the iOS client. As of build 47 these include:

The tcp-client param is not supported in the iOS client. Use tcp instead.

The client expects ifconfig information to be pushed to it (using the push lines in the server config). It doesn’t seem to be possible to configure ifconfig lines locally.

The “redirect-gateway” seems to be required for the iOS client – whereas Tunnelblick connects without it.

- Workflow. Getting VPN On Demand working right off the bat can be hard, especially if multiple parts of the configuration are wrong. If you are new to OpenVPN, you might want to setup a VPN with a the following configurations (in terms of complexity just to get working):

OpenVPN w/ Secret Key – using Desktop machine (use Tunnelblick for the Mac)
OpenVPN w/ Certificate – using Desktop machine
OpenVPN for iOS w/ Certificate – using the actual OpenVPN for iOS Client (not IPCU)
OpenVPN for iOS w/ Certificate – using IPCU 

- Browser Refresh.  If it looks like everything is working (VPN established, etc.) but you still don’t get your Web page, try refreshing the browser/making a 2nd request.  I’m not sure whether this is something in the config, or a “feature” of the OpenVPN client for iOS – but I’ve found that refreshing often helps brings the page to life after the VPN connection has been established.

Uploading Photos from Mobile Web Applications

Here at Neudesic, we’re fortunate to be involved in many exciting HTML5/Mobile Web applications for different organizations. For many of these projects, one of the common requests, especially for field-facing mobile applications, is the ability to upload a photo from within a web page in a mobile browser.

If you are not familiar with the space, you may think this should be default behaviour, but surprisingly uploading media is one of those areas that is still going through the standardization process. Eric Bidelman has a great overview of the three “rounds” of standardization that have taken place so far as part of the Device API working group.

What can you do today?

While Eric’s article gives a great overview of what’s likely to come, the purpose of this post is to explore what’s possible today, outlining the relative pros and cons of each approach.

Firstly, let’s take a look at what’s supported out of the box:


If you are using Android 3.x+ (i.e. either Android tablets running Honeycomb or later, and Android phones running ICS) – or users are running Chrome or Firefox Mobile for Android, the browser will support an input element that can invoke the camera:

<input type="file" accept="image/*" capture="camera">

Simply add this element to your web page, and together with some server-side processing for the upload, users can upload the image from their device.

Android Image Upload

As Eric mentions in his article, it does look like things are heading towards getUserMedia() instead. Although this new API works with latest versions of the desktop version of Chrome, we haven’t found anything that works on mobile browsers as of yet.


Unfortunately, the input element above does not work on Mobile Safari on iOS5 today – the experience for the user is just a disabled button. This is true for the recently released version of Chrome for iOS also (as it’s just a wrapper over UIWebView).

Things are changing however.  At WWDC this year, Apple publicly announced support for photo and video uploads within Mobile Safari shipping with iOS 6. It’s difficult to say much more about the implementation here (as the developer/beta programs are under NDA), but if you have access to Apple’s developer program, it’s definitely worth checking out.

Apache Cordova (PhoneGap)

If you want to support earlier Android versions, and/or can’t wait for iOS 6, arguably the most popular choice is Apache Cordova (previously known as PhoneGap). Apache Cordova provides a native wrapper around HTML-based content, and supports several platforms today – including iOS and Android. The Cordova API supports media capture, and with a couple of lines of JavaScript, it’s possible to instantiate the camera or invoke the camera rolls within your Mobile Web application.

Camera Roll on iOS

Pros: Works with the most popular versions of mobile browsers, and as of now, Cordova provides a good user experience for the user. The user presses a “add photo” button in the HTML application, and the native wrapper invokes the camera control without the user needing to know whats going on. If using the camera roll, the API also supports the option of uploading multiple photos.

Cons: Using Cordova does however change the deployment model for the application. No longer can you just visit a Web site to use your application – you now need to think about distributing your application – either via the AppStore for public-facing applications or through another channel for enterprise apps.

In addition to the deployment, Cordova also introduces several options that need to be considered for actually transfering the image to the site. The default is to use Base64 encoding, which is good for small images, but we’ve experienced performance problems on large images from 5MP+ cameras. The Cordova API does support a File Transfer API, which works well except it doesn’t yet support authentication, so you’ll need to create an anonymous area to post your photos to. If you do need authentication, you’ll want to send the photo directly after capture, which will likely mean writing a custom Cordova plug in and using NSUrlConnection to send your photo to the server.

Companion Application

An alternative approach to wrapping your HTML content with native code is to ship a “companion application” responsible for uploading the photo.

Both iOS and Android support custom URL schemes, which means that applications can respond to different URL requests from the browser. For example, I can create a native application responsible for taking and uploading pictures, using a custom URL scheme called simonphoto:// which I can then pass various parameters – for example, the ID of a project that I’m uploading photos for. Once the user clicks on the simonphoto:// link within the mobile browser, the application launches and I can take and upload as many photos as I want.

Pros: The biggest draw to this approach is that the HTML application doesn’t need to be wrapped with any native code.

Cons: While I don’t have to wrap my HTML, the user still needs to obtain the companion application, which will likely involve a trip to the AppStore or other link for download. In addition, even when the companion application is downloaded, the user experience isn’t quite as slick as the the PhoneGap approach. For example, in iOS there is no natural way of getting back to the Web app once the native application is ready to hand back control (unless you pass a return or app id in the parameters). Also, there is no way of checking whether the companion application is installed or not from the Web page.


Picup is an example of a free companion app that works in a similar way as outlined above. It uses the fileupload:// URL scheme to invoke the Picup app, which much be installed.

Picup Application

Pros: As above.

Cons: As above. No version for Android (although I’m sure there are equivalent apps available).


Finally, a very simple approach, but one that I’ve seen a couple of people use is sending the photo via email. Your application has a “mailto” link which includes a subject (maybe the ID corresponding to the application) and body that reads “Please remember to attach your photo”. The user attaches the photo manually and it’s sent to a server-monitored email store for processing.

Pros: Nothing native here, and deployment model stays the same.

Cons: Funky experience for the user. You’ll also need to setup a server side environment to handle the incoming emails, strip the attachments, and correlate with your application.


A lengthy post, but hopefully it gives you an overview of how to do photo upload from mobile devices today. To summarize:

If you are writing an application that won’t be released for a few months, and will primarily target Android ICS and iOS6, you should be able to use the <input/> element outlined above to invoke the camera and upload a picture. If you do this, make sure you understand the transition to getUserMedia as the specifications mature.

If you are writing an application that needs to target today’s platforms – namely Android 2.x and iOS5, you’ll either need to use Cordova to create a wrapper or a companion application (either hand crafted, or something like Picup). Choosing between Cordova and a companion application is a balance of user experience vs. deployment. A Cordova application will keep the user experience seamless, but you will be responsible for the deployment of the application as a result.

If you don’t want to do any device development, and are just looking to provide a shortcut for users to send photos, you could consider the basic mailto: / email link approach.

Announcing Neudesic Slingshot!

Getting SharePoint working on a mobile device can be hard. The “out of the box” experience, even with SharePoint 2010, is very basic and doesn’t take advantage of the device. Solutions on the AppStore are a step in the right direction, but many don’t do anything over providing the ability to browse a SharePoint site.

At Neudesic, we are hoping to change all this through a project we call Slingshot.

Slingshot is an open source mobile client and library for SharePoint. Built upon the jQuery Mobile and Apache Cordova frameworks, Slingshot makes it ridiculously easy to expose tasks, announcements, documents, and virtually any other object directly from SharePoint to any mobile device.

myWPEdit Image

We demonstrated Slingshot for the first time in public at the SharePointFest in Denver this week, and the response has been tremendous. For those that didn’t make it to the event, I wanted to use this blog post to expand on what the framework can do.

Slingshot is very lightweight, and can be deployed in one of two ways: The code can run on the SharePoint server, and users access it using their mobile browser. Alternatively the same code can run on the mobile device in a native application, with no changes required on the SharePoint server.

Here’s a quick run through of what Slingshot supports today:

SharePoint Lists and Items

Slingshot uses the ODATA support in SharePoint 2010 to expose virtually any list or item to the mobile device. The out of the box demo shows announcements, tasks, and organizational details, and it’s easy to extend to workflows and other lists. Each of these items can be exposed through a form with touch native controls.  For example, task items have a slider control to indicate percentage complete.

Bi-Directional Updates

The sample app also demonstrates bi-directional updating. Update an item from SharePoint UI, and it automatically gets updated on the mobile device. Alternatively, update on the mobile device and the item is automatically updated in SharePoint. This update is seamless, so the user doesn’t have to hit “save” after making the change on the mobile device.

Document Library Support

One of the core scenarios for using Slingshot is to browse document libraries, which Slingshot handles really well. Supported file types (such as PDFs) can also be opened directly from the mobile app.

Integration with Photo Capture

Browsing and opening files from a mobile device is useful, but Slingshot also supports uploading data from the device. The majority of phones now come with a camera. The sample application shows how to take a photo on the device and upload it to a document library or attach it to an item. We find this functionality well suited for field employees who need to interact with workflows that involve taking pictures and uploading them to SharePoint.

Support for Offline Scenarios

Because Slingshot works locally on the device it enables a number of offline scenarios. Currently the application has the ability to work offline, and we are working on synchronization of SharePoint lists and items to the device, which will provide a true offline experience if no connection is available.


Authentication works in one of two ways. If you are accessing Slingshot via a mobile browser, the browser will prompt for credentials (the same way as if you’d just navigated to any other SharePoint page). If you are running Slingshot locally on the device, we use a form-based mechanism that can be customized and extended as needed.

Multi Platform Support

Because we’ve developed all of this using jQuery Mobile and Cordova (PhoneGap), it is supported on multiple platforms by default. Both these frameworks support up to six platforms today, including iOS, Android, Windows Phone 7, and Blackberry. We provide default templates for iPhone and Android, and creating new templates for other devices is as simple as creating new CSS files.

Easy to Extend and Update

Apart from the device integration, everything in Slingshot is based on HTML5 and JavaScript. For those coming from a web background, this makes it very simple to extend and update.

Best of all, we’ve licensed Slingshot as an open source framework, under the MIT license. You can get all the bits for free, and we even have other developers that are signing up to contribute.

Ready to see more? If you didn’t have an opportunity to stop by the booth at SharePointFest in Denver, check out the repo on GitHub – or drop me a line using the contact form if you’d like more information. Neudesic is actively extending this framework for many other scenarios and customers, and we’d be happy to help you extend this for your own needs also.

An Objective View of MEAPs

I recently had the opportunity to put together some research for a customer who has been interested in the MEAP (Mobile Enterprise Application Platform) space. My premise is that the market has become flooded with MEAPs (Mobile Enterprise Application Platforms, as coined by Gartner), yet most of them are fundamentally taking organizations and developers down the wrong path when it comes to developing mobile applications. The problem is that MEAPs demo really well in front of the CIO… “Wow! You just hit F5, and your application is compiled and deployed to iOS, Android, Blackberry devices… Where do I sign?” – yet the reality is very different. Namely:

Language Abstraction – Many MEAPs have their own language that claim to be similar to Java or a flavor of JavaScript. Even if it is the same language, there is always something new to learn. Also, developers tend to shy away from learning any language that’s vendor specific (APEX on anyone? ;–)

Language Limitations – When you create a language that abstracts other languages you always end up serving only the lowest common denominator. Often there are ways of coding styles and nuances supported in the native language that the MEAP won’t expose because it’s not on other platforms.

Platform Limitations – Time has shown that platform abstraction doesn’t work (do a search on Wikipedia for cross platform RAD tools if you don’t believe me). Also, platforms change quickly. What happens if/when the next version of iOS is released? You have to wait for your MEAP vendor to catchup before you can use any of the features.

UI Limitations – Following on from the previous point, many of the controls that MEAPs offer also follow the lowest common denominator rule. For example, Cocoa Touch has a neat feature called a “Half Page Curl Transition”. If you abstract your UI to a MEAP vendor, how is this supported on other platforms? Either a) it’s not (so by default you can’t use it on iOS), b) you can only use it on iOS (which breaks the promise of a MEAP) or c) it’s implemented as a hack job on all platforms.

Tooling – MEAP tooling is often Web based, or bundled as an Eclipse add-on – but in both cases it often doesn’t fit in with the other tools that developers use today (e.g. ALM, refactoring, unit testing, etc.). Also, many of the MEAP vendors seem to be aiming their toolsets towards business analysts. When are we going to let this one go? Business analysts have an important role to play in teams, but they shouldn’t be developing UI, and they shouldn’t be writing code.

Debugging – Let’s say you create your application with a MEAP, and then it crashes 1 time in every 10 (i.e. one of those really nasty bugs to find). What are you going to do? a) Reach out to Apple? I suspect they won’t help you. b) Search the web for other people with the same issue? Could be difficult to find. c) Pay the MEAP vendor even more money to investigate your issue? Ah, that’s the one…

Data Abstraction – Many MEAPs offer connectors to databases, XML files, SAP, etc. Not only are these costly, but many are just pass-through connectors, so as a developer you don’t get any control over the connection. Want to implement synchronization? Want to change the formatting or query for better optimization? Probably not going to be possible.

Difficult to Extend – Many MEAPs don’t extend very easily. Found a new JavaScript library that you want to use in your application? How about a native control that you’d like to display? Could be tough. Some MEAPs do offer bridging to native code – but in which case, why not just write the whole thing in native code to start with?

Vendor Lock In – The sum of all the above leads to quite the definition of vendor lock in. Chances are by the time you deploy your first application you’ll be signed up with an expensive support and service agreement, with your developers working on code in a language that they’ll never be able to re-use, on a platform that won’t go anywhere.

I’ve extended this into a presentation that you can find below – which cover the above points in more detail, plus outline some general alternative approaches to creating applications without the dependency on a MEAP framework.

Based on other people’s experience with MEAPs, am I missing anything?

Tips, Tricks, and Recommendations for HTML5/Mobile Web Development

Over the past few months I’ve been fortunate to have worked with several customers creating Mobile Web applications for different devices. These have included an iPad based Mobile Web app for a healthcare provider, a cross-platform Mobile Web app for a commercial real estate broker, and an Android tablet solution to allow scientists to research experiment details from the comfort of their lab.

From these projects, I’ve come up with a list of tips, tricks, and recommendations that I wanted to share if you are looking to develop your own Mobile Web applications. Hope you find them useful…

Understand the differences between Mobile Web and hybrid applications

It’s important to make an early decision on whether your application is going to be pure Mobile Web (the user accesses it through the browser on the device) or hybrid (the user launches an application with an embedded control to display web based content). The decision to use the mobile browser or to have a dedicated hybrid application will likely be determined by the functionality the application requires. For example, obtaining the GPS location from the device can be performed perfectly well inside the browser, but accessing the camera from a Web page needs a hybrid application to provide the functionality.

In my experience there are three common features that drive developers towards a hybrid application: 1. Access to the camera. 2. Access to local contacts on the device. 3. Deployment to the AppStore/Marketplace. The first two are to overcome limitations of what can and can’t be accessed through the mobile browser (and is something that Mozilla is trying to tackle through WebAPI). The latter one is probably the most important to understand however – as while a hybrid solution will enable the application to be published to the app stores, extreme care needs to be taken to ensure that the application will not be rejected. This is most common with applications destined for Apple’s AppStore (if Apple view your application as a single container for web content with no apparent native functionality, they will more than likely reject it).

If you do take the hybrid approach, you really have two options to build the application – you can either set out and build a native application, embed a WebView control and then hook different events to create your own bridge between the two. This can be very useful if you only have a small piece of Web functionality that you need to embed in your application. Alternatively, you can consider using a framework such as PhoneGap (now known as Apache Cordova). Using Cordova is not without it’s challenges (for example, with cross domain scripting and client side certificate authentication), but it can definitely give you a head start especially if the majority of your application is going to be based on mobile web content.

Setup a robust development and testing environment

Setting up a robust development environment might be viewed as obvious for most of us, but tools for developing Mobile Web applications are still relatively immature, so it’s an important area to get right before you start your first app. In the most basic setup, you are going to need three things to be successful:

The first is a great IDE for HTML development. You’ve probably got your favorite already, but a strong IDE with as much JavaScript code completion support as possible will be your friend. Personally, I switch between TextMate, Visual Studio 2010, and WebStorm depending on which platform I’m working with. It’s also interested to see where Adobe are headed with their toolset.

Secondly, you are going to need a browser capable of inspecting the DOM and doing JavaScript debugging. Personally, I use Safari for doing this, but Chrome and/or IE9 will equally provide you the tools that you need. The important thing here is not so much the choice of the browser, but the understanding of how to breakpoint JavaScript code and inspect elements both on and off the page.

Finally, you’ll need some kind of device, or device emulator. Switching the user agent on a desktop browser will only get you so far, and eventually you are going to need to test on something representative of the device that the user will be using. If you are developing an application that will be primarily used on iOS devices, I would certainly recommend a Mac (as the iOS simulator is very robust). If a Mac doesn’t float your boat as your development machine however you could consider a hosted solution such as MacinCloud or simply do all your debugging using an iPod touch. There are a few PC based utilities that claim to emulate mobile safari/iOS, but I’ve had a poor experience with all of them.

For Android, I tend to avoid the emulator whenever possible (it’s a resource hog) and instead have a selection of devices that I can use for debugging on the fly. I’ve found that the key for Android testing is to have a selection of devices with preferably different resolutions and running different versions of the OS (e.g. a separate device running 2.1, 2.2, 2.3, and 4.0 should cover most of the current market, unless you have a need to go back to 1.6 for some reason). If you don’t have access to physical handsets, you may want to consider one of the testing providers such as Perfecto Mobile. They have a variety of devices that you can rent, covering many different platforms.

Use client side JavaScript instead of server side scripting

If you’ve been developing Web applications for some time, especially enterprise facing applications, you’ll be used to having a lot of the presentation and application logic tied up in server side scripts, maybe running server-side ASP, ASP.NET, JSP, PHP, etc. When developing Mobile Web applications however, things need to change. While server-side scripting will of course still work for a mobile application (and might be very tempting as you start to write your first application), it quickly creates a tight coupling between your device and the server. In the old days where you were accessing the server from a desktop browser on the same network, this wasn’t a problem – but once you move the experience to a mobile device with a potentially patchy network connection, the last thing you want to be doing is relying on the server side for generation of pages and views. If your mobile web application needs to do a roundtrip to the server when you hit the “Next Page” button, you’re going to be heading for a world of hurt.

The answer is to move as much (if not all) of the application logic to client-side JavaScript. This might be a painful exercise as you translate the move from an MVC server side framework to handcrafting page navigation using client side JavaScript, but the pay off will be worth it. If you have all of the application logic locally, together with pre-fetching as many of the pages into the DOM when your application first loads, you effectively disconnect yourself from the server side which makes performance much better and also moving to offline much easier. Of course, you still want to communicate with the server side – who doesn’t? – but this should be done via a set of asynchronous REST/JSON calls to bring data into the application rather than relying on the server to generate HTML for your application.

Design for running offline

Related to the previous point, you should also start thinking about how your application could run offline, even if you have no plans to do so. This will put you on a good architectural path, and will also mean that your application will work if and when network connectivity is lost. There are plenty of articles explaining how to enable offline access in HTML5 applications, so I won’t plan to cover here – other than to say you’ll be creating a cache manifest for your application, specifying the files to go into your cache, enabling the scripts for checking the cache when your application starts, and handling control logic accordingly (e.g. when you hit that submit button, you’ll want to know whether there is network connectivity before making the call to the server!).

If there’s one piece of advice that I can offer here however (and we learned this the hard way) it is to design for offline first, but implement last. Do all of the right things to support offline in your application, but don’t enable it until you are close to the end of your development cycle. If you enable it too early, it can cause mayhem with the development team – as your developers could end up debugging against a version of the application that could be slightly older – or the cache doesn’t get refreshed in the way it should. This doesn’t application to local storage – this is fine to access during development – I would just recommend not enabling the cache manifest until you go into UAT.

Select the right UI framework for your needs

Almost every other day there seems to be a new HTML UI framework released, and mobile frameworks are no exception. At Neudesic we’ve had most experience and success using jQuery Mobile, but have equally done work using Sencha, Dojo and others. As all the frameworks offer slightly different combinations of controls, I’ve found one of the tricks for selecting a suitable framework is to first build out a list of the controls that you are likely to need. For example, in one project we developed a set of paper prototypes (side note – I really dig UI Stencils for this) and realized that we would be creating a multi-tabbed application, with many lists on each page. Knowing what controls we were likely to use ahead of time made selecting the framework easier, and also put into perspective what controls we would have to build from scratch.

The good news is that most of the frameworks have demo pages/sites available, so that you can check them out beforehand without having to download anything. This can be a great way of testing the controls across a different number of devices, as well as getting a feel for the performance. My rough order of magnitude is that HTML based mobile UI will feel around 85% of a native control – it will look similar, react similar, but you just know that there’s this 10 – 15% gap that doesn’t quite feel native.

Understand the caveats with implementing maps in Mobile Web applications

Related to the last point, many applications have a need to display maps and pinpoints, and this will likely remain true for many mobile web applications that you might be creating. Based on experience, there are a couple of caveats that need to be understood before you head down this path.

Firstly, performance on Mobile Web map controls is not to par with native controls. Using Google Maps in Safari on iOS seems sluggish when comparing it to using MapKit in a native iOS application. I’m not sure what optimizations have been done in MapKit to give it the performance it has, but the difference is quite noticeable. As a result, if your application has a fundamental need to do a lot of mapping (i.e. if this is it’s primary function) you may want to consider a hybrid approach where you navigate to native pages for the mapping as part of your application.

Secondly, the web control for Google Maps (and this probably applies to other mapping controls) likes to “own” the page – which can put it into conflict with UI frameworks such as jQuery Mobile. For example, when you change pages in jQuery Mobile, the DIV element of your current page is set to hidden, and the DIV element of the new page to block. If your map control is embedded on a page that gets set to hidden this can cause Google maps to have a hard time. We’ve experienced random behavior such as the map control not updating or repainting correctly when it is brought back into focus. This can be overcome by an event on the page load to refresh the map canvas (example below), but it’s one of those bugs that takes a lot longer to find than it does to fix.

$(‘#mappage’).live(‘pageshow’, function() { google.maps.event.trigger(map, ‘resize’); });

Think ahead for non-standard UI elements you are likely to need

While frameworks such as jQuery Mobile will provide standard controls such as buttons, lists, labels, and radio boxes, your application will likely have a need for elements and controls that are non-standard. These can include tables, grids, and image carousels to name a few. Many of these can be created using custom controls, but before going down that path, I always try and ask myself two questions:

Firstly, if the control does not exist, does that mean a better UI nuance is available for the device? For example, there are few mobile web controls for displaying datagrids on mobile devices. Rather than force-fitting a solution that might not work well on a 3.2″ screen, think about what alternative solutions could be used instead. For example, maybe a series of cascading list controls would actually provide for a better user experience over some custom hacked-together table control. When in doubt, try to look at how other mobile applications have used controls to represent a similar data set.

Secondly, if the control does seem to fit, but it’s just not included in the standard UI framework that you’ve chosen, think about searching for a third party solution before re-creating something new. This can apply to image carousels (PhotoSwipe is one of my favorite controls for doing this) as well as charting (there are several really good JS based charting controls out there).

Make your mobile web app look and feel like it belongs with native applications.

Why not go that extra step and have your mobile application look like it belongs with other native applications on the device? A couple of small recommendations for this one:

Firstly, never show the address bar in your mobile application. Simply use a <viewport…/> tag to have your application run in full screen mode. The user will still be able to get to the address bar, but it will be hidden from view by default. (side note – jQuery Mobile now performs this automatically without the need to do “scroll to” events in JavaScript).

Secondly, if you are targeting iOS devices, consider using Cubiq’s AddToHome JS library. This is an excellent add on that will display instructions for the user to add the application to the iOS home screen, enabling the application to have it’s own icon and run completely full screen when launched.

Don’t go creating your own username and password database

Have users sign on to your application can be very useful for security of the application and for storing preferences and other settings. Unless you have a very strong reason however, I recommend that you avoid creating your own username and password database for your application. Doing so can seem really easy at first, but as your application gets more popular you’ll be left with a nightmare situation of having to deal with registration issues, password resets, and all of the other administration that comes with owning your own identity provider.

If you are building a mobile web application for the enterprise, I suspect you’ll already be thinking about this – either hooking in to an existing Active Directory instance or other identity provider (although you’ll likely need to consider how this will work outside the firewall when your mobile users are accessing your application via a public connection).

If you are building a public-facing application however, an efficient solution can be to federate with an existing identity provider. Chances are your users already have a Google, Yahoo, or Facebook account. Why not take advantage of this and use this sign-in information to authenticate the user for your application? This can be done in a couple of different ways – either with a native SDK from one of the providers – or using a federated service such as Microsoft Azure’s Access Control Service. This is a simple service to setup, which will enable you to configure a list of providers and returns a bag of claims to your application which you can use to match up with a list of settings, preferences, etc. There are a few caveats with this approach, especially as some of the providers return the information via a post back (which can really upset your mobile application if you’ve pre-fetched all of the pages in to the DOM beforehand). This is something that I’ve written previously about in this post.

Make your Mobile Web application perform well

It might sound obvious, but if you want your Mobile Web application to come anywhere near to the performance of a native application, you’ll want to keep a keen eye on performance. Generally, I’ve found there are three main causes of bad performance in Mobile Web applications:

First is server-side generation of the UI and round tripping (as discussed in the point on client-side JavaScript earlier). If you are generating any of the UI on the server-side or making any synchronous server-side calls, you are more than likely to run in to trouble.

Second is too many CSS effects/transitions. Many of the UI frameworks, including jQuery Mobile, include several page transitions for when the user navigates between different pages. Unless you are exclusively targeting iOS devices, I would recommend turning all of these off. Many of the older Android devices don’t implement a version of Webkit that supports hardware acceleration – plus they are running on less capable hardware – so to avoid creating an application that stutters on Android devices, it’s best to disable any transitions and effects.

Finally, watch out for too many registered JavaScript events (especially on the class level). The JavaScript performance on mobile browsers can be much lower than desktop equivalents, and if you are not careful you can end up where the performance of the browser is being compromised by too many registered events. Often this will only add 100ms to various events, but combine this with specific events (such as a button tap) that must immediately navigate to another page, any delay over 100ms might be enough that your user thinks that the button wasn’t pressed correctly. They’ll press it again, and before you know it you have two JavaScript navigation events that you’ll need to handle.

Wrapping up

A bit of a lengthy post, but I hope that this is useful if you are heading down the path of developing Mobile Web applications. If you’d like more information on this, or maybe to share with colleagues, I’ve uploaded this presentation to SlideShare that goes through the above list in some detail.

Extracting and Using Facebook OAuth Token from ACS

A couple of my previous blog posts have shown how AppFabric ACS (Access Control Service) can be used as part of the Windows Azure Toolkit for iOS to enable federated authentication with Facebook, Google, Yahoo, and other providers.

I was recently asked whether it’s possible to extract an OAuth token as part of the ACS sign in process that can then be presented to Facebook’s Graph API in order to access details such as friends lists, photos, etc. In this post, I’ll cover how this can be done.

The first step is of course to display the authentication page. As you’ve seen in previous posts, this can be achieved using the following:

WACloudAccessControlClient *acsClient = [WACloudAccessControlClient accessControlClientForNamespace:@“iostest-walkthrough” realm:@“uri:wazmobiletoolkit”];

[acsClient showInViewController:self allowsClose:NO withCompletionHandler:^(BOOL authenticated) { if (!authenticated) { NSLog(@"Error authenticating"); } else { WACloudAccessToken *token = [WACloudAccessControlClient sharedToken]; NSString *securityToken = [token securityToken]; ... }

In the application, this will display the federated login dialog, and prompt the user to enter their Facebook credentials.

myWPEdit Image

You’ll notice that the ACS client returns a cloud access token, of which a security token can be extracted. This security token is a set of claims returned from ACS. Here’s an example:

If we want to make follow on calls to Facebook’s Graph API, we are going to need to present the Facebook User ID and an OAuth Token. Fortunately both of these can be extracted from the token.

To extract this, we first HTTP encode the token:

NSMutableArray *httpEncoding = [NSMutableArray arrayWithObjects:[NSArray arrayWithObjects:@"%3a",@":",nil], [NSArray arrayWithObjects:@"%2f",@"/",nil], nil];

while ([httpEncoding count] >= 1) { securityToken = [securityToken stringByReplacingOccurrencesOfString:[[httpEncoding objectAtIndex:0] objectAtIndex:0] withString:[[httpEncoding objectAtIndex:0] objectAtIndex:1]]; [httpEncoding removeObjectAtIndex:0]; }

NSError *error = NULL;

Using a simple RegEx search we can extract the Facebook User ID:

NSRegularExpression regex = [NSRegularExpression regularExpressionWithPattern:@“[0-9])” options:0 error:&error];
NSTextCheckingResult *match = [regex firstMatchInString:securityToken options:0 range:NSMakeRange(0, [securityToken length])]; firstRange = [match rangeAtIndex:1]; fbuserId = [securityToken substringWithRange:firstRange];

And the required OAuthToken:

regex = [NSRegularExpression regularExpressionWithPattern:@“[A-Za-z0-9]*)” options:0 error:&error];

match = [regex firstMatchInString:securityToken options:0 range:NSMakeRange(0, [securityToken length])];
NSRange firstRange = [match rangeAtIndex:1];
oauthToken = [securityToken substringWithRange:firstRange];

Now it’s just a case of calling the Graph API using these credentials:

NSString *graphURL = [NSString stringWithFormat:@“”,fbuserId,oauthToken];

For the purposes of this post, let’s take a quick look at my list of friends.

NSURLRequest *request = [NSURLRequest requestWithURL:[NSURL URLWithString:graphURL]];
NSURLResponse *response = NULL;
NSData *data = [NSURLConnection sendSynchronousRequest:request returningResponse:&response error:&error];
NSString *friendsList = [[NSString alloc] initWithData:data encoding:NSUTF8StringEncoding];
regex = [NSRegularExpression regularExpressionWithPattern:@“id” options:0 error:&error];
NSUInteger friendCount = [regex numberOfMatchesInString:friendsList options:0 range:NSMakeRange(0, [friendsList length])];

Displaying the friend count in a UIAlertView on the screen shows how popular (or not!) I am:

myWPEdit Image

A simple example, but hopefully this shows not only that you can authenticate against Facebook using AppFabric ACS, but also how the returned Facebook User ID and OAuth Token can be used to make further calls to Facebook using the user’s credentials.