Cloud, Spirals and Skeuomorphism

Laurent Bede 24th August 2015

Imagine a team of designers from the early 2000’s tasked with creating an application to replace a spiral notebook. Caught between the proverbial “rock and hard place” of facilitating the adoption of your future application and limited technology, the end result would likely be plenty of white “pages” with faint blue transecting lines with a spiral bisecting the screen.

This book, making the transposition from print to digital, is a good example of skeuomorphism, with its detailed reproduction of elements that were necessary to the function of the original object (see wikipedia).

Apple, quite notably, has used and abused this ideal until iOS 6 recently abandoned it in favor of a flat design, marking a milestone in the maturation of the use of technology.

Far from HMI design, cloud speaks of virtual machines (VMs), vCPUs, amounts of volatile or nonvolatile memory, network adapters. Taking the concepts of the real world of which it is a digital implementation, this is yet another fine example of skeuomorphism.

For the same reasons as the spiral notebook, the visual reminders persist even though the technological constraints account for the largest share.

Indeed, we cannot rid ourselves of the distinction between local and remote resources, which complicates many architectural drawings. As long as the network speeds remain 10, 100 or a thousand times slower than the interaction between CPU and RAM, the Cloud will propose a segmented paradigm and will encounter this limit just like a goldfish in a bowl.

These constraints, certainly temporary, align marketing a business that meets the need to not scare a CIO who has already espoused cloud in an effort to abandon server rooms with the  expectation that he not be required to surrender, conceptually, that on which he founded his computer expertise.

So let’s take our cloud CIO and his ecosystem of software vendors by the hand and take them out of the Ether.

The next (r)evolution of cloud will coincide with the advent of distributed operating systems, multi-machine application layers and offer homogeneous software abstraction, smoothly and without capacity constraints.

Providers deploying engineering resources to accelerate this transition, can rely on their partnerships with manufacturers to design hardware and specialized middleware optimized for Cloud use.

As usual, the world of open source will play a pioneering role, with its solutions waiting for official recognition of a tech giant to become “prime time”.

The fight for the Cloud OS standard will be fierce and the winners will be those that are least disruptive, much like emerging databases providers “NoSQL” and “Big Data”, which made the both clever and realistic choice to retain SQL access.

The challenge is to open an ocean to our brave goldfish, where it can develop without suffering genetic modification.

Meanwhile, IaaS, nearly out of adolescence skeuomorphism and entering its “flat design” adult phase, is available to build on its tangible benefits, leaving it free from waiting for the server, with its storage in the garage next to a box full of old spiral notebooks.