Preserving iPad apps

Graham Purnell is Digital Preservation Assistant at the National Library of Scotland.

As part of a practical exercise to investigate preservation of modern digital objects at the National Library of Scotland (NLS), Graeme Hawley, our Head of General Collections, drew up a short list of items for investigation and possible addition to the Library’s collections.  This list partly emerged out of discussions with partner Legal Deposit Libraries, including the British Library. A small team from across NLS was assembled to investigate best methods and discover technical and legal barriers. The Bytesize Collection Group was born.

The first fully investigated object was an iPad app called "Goldilocks and Little Bear" from children's book publisher Nosy Crow. The app shows the story from Goldilocks' perspective in one view and if the screen is rotated 180 degrees, it shows what's happening to Little Bear. The app won FutureBook Award for Children's Digital Book of the Year in 2016. You can watch YouTube videos to give you a flavour of its content:

https://www.youtube.com/watch?v=TpjN5ph15AA&t=1s

https://www.youtube.com/watch?v=wNagspyCU9s

Unfortunately, Goldilocks and Little Bear is no longer available on the Apple App Store and can no longer be downloaded to a device from Apple. Fortunately, the British Library has a copy of the app received from the publisher under legal deposit, so we tried to see if we could upload the files to a device, to potentially provide access in Library reading rooms.  Results were to be reported back to the British Library and other partner Legal Deposit Libraries. 

The only way to run the app effectively would be on a tablet device with gyroscopic capability, so that the screen can be rotated. There is no way to emulate iOS apps on non-Apple devices, so this restricted us to loading the files onto an iPad using Xcode, the development software for Apple products that only runs on MacOS. Our IT department loaded the files onto an iPad but could not get it to run; it seems the app wouldn't run on a more modern version of iOS and IT were understandably unwilling to seek out, deploy, and run apps on older iPads due to network security issues.

Apart from not being able to emulate on non-Apple devices, and the security issues associated with running obsolete software, iPads also have built-in obsolescence. Even if we could provide access using old iPads, they will eventually break down and replacements will be harder to obtain. This is a problem for any digital content that needs a specific device to interpret digital content, to be a translation layer between code and user.

This left us with a problem of how we might archive the app and also the experience. Interactive gameplay was ruled out; any interaction with the app was proving to be impractical from a technical and security perspective. The solution we came up with, though not ideal, was to archive the app files and to separately and passively archive gameplay, to shoot video of a person using the app so that future generations could understand how an iPad was used (not a given in fifty years' time) and also to archive the video feed from the iPad itself.

Some video content is HDCP protected (HDCP = High-bandwidth Digital Copyright Protection.) Such content cannot be played without a "handshake" connection to another HDCP-compliant device. Some software is coded with HDCP and any attempt to capture video will not work. HDCP has to be licensed by software developers and is not applied to all apps; it is mainly used to protect streaming video, from companies such as Netflix, and iTunes purchases. Most game developers have no interest in HDCP because video capture doesn't steal the game; it can't be interacted with and gameplay video platforms, such as Twitch and YouTube, play a large part in building fan communities and marketing game titles. So the Nosy Crow apps are a distinct possibility for gameplay video capture. iOS versions greater than iOS 11 can record screen output directly to the device, which can then be shared. Unfortunately, Goldilocks and Little Bear requires an earlier version of iOS so this wasn't possible. As seen in the recent Digital Storytelling exhibition, the British Library experimented with creating gameplay videos for a few of the narratives on display, using facilities at the City Interaction Lab within the Centre for Human-Computer Interaction Design at City, University of London. However, there have been difficulties capturing a video stream from the Goldilocks iPad app.

I’ve suggested that the British Library should purchase or borrow an Apple TV device and try to stream video output from iPad to Apple TV using Apple's Airplay. Airplay has been available since the 1st generation of iPads, so video capture from either the Apple TV or its connected display should be possible using HDMI or other video connectors. This hasn't been tried yet, but I will update this blog when I have answers.

Although archiving this app's code could provide a solution for future users to run the game interactively (if they can surmount the difficulties mentioned above) this will not be possible for all iPad apps. Goldilocks and Little Bear is a standalone app and all it needs to run is in the code. This is not the same for all apps, many of which provide an interface and interactivity with resources on the internet. For distributed content like this, where much of the necessary code and functionality is web-based, archiving the app code (.ipa format) will not archive functionality if the web resources go offline. This is not a theoretical risk. Many apps on my iPad2 no longer run because new content no longer runs on old versions of an app, and new versions of the app are not available for old versions of iOS. Some apps no longer work because their web resources are no longer online. Some apps no longer run on new code, so updating a device could make certain apps unavailable. The only way to capture some of these apps is passively, using video capture, while they are still playable/functional. This places a narrow time constraint on when objects of interest can be archived and documented.

Because of the difficulty of archiving certain interactive digital objects, due to the issues mentioned above (and more,) video capture is likely to become a growing element in archiving digital experiences, especially if the experience requires a specific device with in-built obsolescence, or the object requires distributed services (such as online gaming.) The choice will very often be "should we archive this experience on video or quickly lose any capability to archive the experience at all?"