I was celebrating 10 years after the dreaded Y2K down in Florida this year and was wondering if I would dare to write down my technology predictions. So feeling bold here they are, only time will tell if I’m anywhere close to accurate 
:
- Cloud
- Authorization
- Death of smart clients
- No-SQL and ORM
- BI becomes commodity
- Netbooks – and the Google ‘netbook’
- Deployment
- SVN ==> GIT or Mercurial
Cloud
No surprises here, with Amazon running this model ‘forever’, Google not too far behind and now even Microsoft has started to chime in as well. The are alliances and definitions for SaaS, IaaS, PaaS. This concept is here to stay, and in fact will affect us in significant ways.
Authorization/Authentication
This is the big play that will help make the cloud be really viable, and even work with managing external companies that need to work for your company (think PCI compliance and support). This model will be based on the claims concept that has been knocking around for a while now, all it needed was something to push it into the limelight. Clouds have started to push authorization / authentication! In addition when this matures to a ‘tipping point’ I have no doubt customers will start to ‘outsource’ their claims trust (with contractual backup!) to other companies. For example if your company needs to provide PCI support to a customer, today the customer will have to be provided the peoples names and assign them logon accounts and manage their passwords. This becomes a major overhead and can be relatively easily managed through a claims based authorization system. Naturally this wont happen within two years – but it is coming!
Death of Smart Clients
Companies will realize that Smart Clients (aka Prism, Click-Once web forms) are only a stop gap. Browsers will rule the application space within 5 years, and will have made significant inroads before 2012. Conceptually they all make sense, but in today’s platform varied world they are dead – but it seems not everyone realizes it yet!
No-SQL / ORM
It seems that there are two factors pushing these forwards, but the root of it is that the historical relational database has so much overhead associated with infrastructure, management and DBA’s – people are looking for alternatives. It has become increasingly common for application programmers to fear the rules of relational databases they have to integrate with as they have evolved into their own discipline. This leads into ORM and which really try to ‘ignore’ the storage mechanism and provide plain business value. The No-SQL group are in a similar boat, they are interested in scaling and maintainance and know traditional relational models don’t meet their business needs.
BI becomes a Commodity
Microsoft has been playing the “BI is coming, BI is coming” song for a while now. However it seems with the SQL 2008 R2 release we are starting to get commodity solutions. This in intern will coerce the competitors to step up. The good news for competitors is that Microsoft has got more pricey, its no longer the ‘cheap kid on the block’. Open Source is starting to catch up with its commercial cousins. Either way its going to be an interesting couple of years for BI.
Netbooks – and the Google one
We have heard the term many times before, and even had a couple of false starts. The difference now is:
- Processing power has significantly cheapened – allowing more inexpensive offerings
- Web/Cloud infrastructure is in place. No worries about email or docs being stored on line.
So we have pretty good netbooks now, but I suspect Google will potentially set the standard for the netbooks. If done right (as opposed to the Nexus!) they have a killer combination. This is one I’m really interested to see what happens!
Deployment
Vista didnt get much traction and XP is on its last officially supported legs, as such I suspect that companies will be rolling out Windows 7 over the next 2-3 years. This will stress the deployment tools, and require individuals that know deployment to step in. In addition to this Windows Server 2008 R2 only comes in 64 bit – software houses realize the time has come to admit they are going to have to invest in supporting 64 bit solutions. This is going to be unusual for customers as they have forgotten how to roll out such changes as it has been a while since the last great deployment upgrade. There will be lots of opportunities in this area!
SVN ==> GIT / Mercurial
It seems that the CVS savior is in the process of being usurped by the DVCS products. Specially GIT and Mercurial are the strong runners. Its a shame because I had waited a while for SVN to mature to a point and now GIT & Mercurial are eating up projects. Oddly at the time of writing this even the Microsoft sponsored CodePlex site now supports Mercurial !
Finally the one I didn’t add to the list is software parallelism programming models. I suspect this will mature with all the new multi-core processors that are coming out. However I think this will mature after 2012
Gareth
