With its roll-out of iCloud, Apple has moved past MobileMe to offer fuller-fledged Cloud Storage, putting into practice their vision of Cloud-based device synchronisation. Yet, what does iCloud really do?
iCloud enables iOS and MacOSX apps to store documents (content) and key-values (metadata) to Apple's Cloud, such that users can view or edit content from any device, without having to sync or transfer files.
Most apps will use iCloud document storage to share documents, enabling photos, videos, and documents to be accessible across devices; the key-value data store manages metadata, such as storing application state, settings, and other important information that delivers a better user experience. The iCloud service handles storage searches, change notifications, version control, conflicts, and security for applications that integrate with a user’s iCloud storage account.
This architecture is nothing new, but implemented as part of the iOS and Mac OSX platforms, it removes the MacBook or PC as the central hub for managing content, enables over the air (OTA) synchronisation, and is tightly integrated into a slew of Apple apps.
What is novel is the extension of metadata-driven content management via a Cloud-based repository onto devices to provide de-duplication. Rather than require the upload of data that already exists elsewhere, such as in the iTunes store, iCloud enables the matching and de-dup of such data.
This is the first step in realising a long-desired model for the consumer market. Can and will this extend to enterprises? I see several challenges to overcome in order to do so:
- De-duplication of arbitrary data requires that all content carry enough contextual metadata to be efficient and accurate. This requires either a normalised view of metadata for arbitrary data formats to be supported by iCloud, or that Apple’s iCloud key-value formats be opened up so that they can be adopted. I find it difficult to see either happen.
- The intelligent synchronisation of Mail, Calendar, Contacts, Documents, etc. has been further advanced in other SaaS and PaaS platforms such as Google’s. Apple’s model has the right idea in mind - the apps work with local data and sync when there’s connectivity. The problems of conflict resolution, caching, and offline devices/modes are not trivial, and in all likelihood, Apple will address them through tight integration and control over what goes into iCloud. This will be hard to sell into the enterprise.
- Transparent, automatic sync or streaming/push of arbitrary content across devices is a difficult problem. It has taken years of work and certain technological developments to address the performance and security concerns in this space. Apple will certainly encounter these difficulties as it develops iCloud for prime time.
- iCloud can only be fully experienced within the Apple iOS ecosystem. Both history and Apple’s model indicate that it won’t be opened up to any competing content platforms anytime soon.
Whilst it will guide the direction of Cloud Storage and re-enforce the validity of the device-centric model, iCloud is clearly focussed on personal content, which plays to Apple’s strengths. This leaves the enterprise market wide open.
Comments