For all the hype (both justified and not) around cloud, one topic that’s been given lip service is that of standards.
I call it the one word oxymoron because the plural invalidates the concept, ipso facto. However, I think there’s a lot to be gained in the cloud space in this arena.
Few would argue that a lack of any standard is preferable. Manufacturing, transportation, telecoms, retail have all benefitted from standardisation.
Now, one can argue that many of these early standards are not truly global in nature, but each in its own way, helped propel an industry forward.
Enter the cloud. It’s not surprising that there’s a rush to define standards in this space. No one wants to have to design their systems to multiple vendors’ offerings.
You can argue that in some senses cloud adoption has been inhibited by a lack of standards (or let’s at least agree on interoperability for the purposes of this scribe).
But, at one point, I took note of more than twenty organisations trying to define the standards for cloud.
They ranged from true standards organisations to industry groups to technology companies to consultancies. However, most tellingly, the participant NOT in the process are those companies which one could consider at the forefront of cloud scale.
I’ve heard the arguments that standards applied too early stifle innovation. Those who point out that many so-called standards are merely thinly veiled attempts to freeze a market are likely right in some cases.
Then again, so are the people desperately wanting some minimal set of standards because of a need for multiple providers.
It makes sense from a legal and commercial perspective to have commonality, to have options, and to have a plan B (and C and for the truly paranoid plans D through J—I picked J because that’s where I normally stop….)
I don’t know when we’ll get to the level of standards that many aspire to. But interoperability can’t come soon enough for me. I’d really like to see the industry coalesce around some basic interoperability sets.
I’d urge providers to at least include shims to something like this, or, come to consensus on something similar. Having divergent API stacks just delayed adoption.
We’ll get to the ISO –style standards in due time, but I want to be able to write code and systems that run on a multiplicity of cloud providers. Accenture’s clients want that too. It doesn’t seem like a crazy request.
So, do we care about innovation or adoption of the technology? It’s only when we settle this question that we’ll be able to have the truly interesting discussion about unlocking the value in cloud.
Joe Tobolski is Global Lead of Infrastructure at Accenture Technology Labs