Friday, July 04, 2008

RedMonk talks clouds.

I've been busy with my new project, but I've just picked up all the cloud talk going on at RedMonk.

I've left a brief comment, so I thought I'd post a copy here as well.


Good post.

I've talked about this stuff for a long time but to cut a long story short, standards (as in open formats and APIs) are not enough to create portability and interoperability between providers nor do they solve the additional issues of competition and strategic control for vendors (and a host of others).

We are likely to only achieve competitive utility computing markets where we "write and run enterprise applications that you could move from cloud to cloud" if the markets are based upon open sourced standards.

Since IT is moving from a product to a service based economy with competition on Price vs QoS rather than product differentiation, the use open sourced standards is obvious for the service provision of ubiquitous activities.

It is however unappealing for those vendors who are wedded to a product world with competition based upon features. For those willing to accept the new world, there will be opportunities from brokerages to exchanges to compliance authorities and so on.

Posts you might find of interest.


As for Alistair Croll's question on whether there is a way of solving the SaaS lock-in (or more correctly lack of second sourcing) issue - well yes there is, it's called open source. The real question that should be asked is when will the *aaS vendors realise that they are competing in a service and not a product based economy?

One final warning, as I've said before there will be more open sourced standards at the application layer of the stack than the framework (now called platform) and the hardware (now called infrastructure) layer. If your "product" doesn't become the open sourced standard when a market emerges around that service offering (whether it's CRM, HR, ERP or JavaScript Development Environments or VM etc), then the odds are it's game over for you. Standardisation and consolidation go hand in hand in the service world.

This is a disruptive change that is going on, and there will be casualties from the product world.


Anonymous said...

I'm not sure open source is the answer. Most cloud vendors operate on open source platforms, but still lock you in through scale or APIs somehow.

I think what we need is Open Services. I'm not sure what this means yet, but I suspect it's something like this:

Electrical utilities are "open service" because you can plug into the grid and generate power. So anyone can become a producer. Of course, electricity is a lot easier to standardize than a computing platform, so in the cloud we're likely talking about a set of fundamental functions that are portable and plug-compatible, and priced in an "open" way.

Then, just as OpenOffice keeps Microsoft Office's prices in check, so an open cloud alternative provides push-back against cloud lock-in.

swardley said...

Hi Acroll,

"but still lock you in through scale or APIs"

Whilst there is an issue regarding lock-in through network effects, what you describe as "open service" would seem to be in practice the same as what I describe as "open sourced standards" - i.e. based upon an open sourced reference model which is operational.

The issue is that most services today may well be based upon open source platform but the service itself is not based upon an open sourced standard .i.e. the service itself isn't open.

I would completely agree with you that what we need are more open and utility like services in IT, that's commoditisation for you.