There are two big trends in ICT; Service a a Service (SaaS) and Open Source Software (OSS). And I do think those two go hand in hand and are reflecting changes that are coming towards the ICT landscape. When I say "hand in hand", I do mean that they are complementary (like people are in a relationship) but also that the are in contrast of each other (like in many relationships). It is my opinion that you either outsource
(parts of your) IT activities, or build the solution yourself. You will either use a cloud to deliver your needs, or make your own cloud. It will either be service from the Microsoft's or you will have to build and manage the solution by yourself. You either well use a cheap commodity service with limited customisation to have the complete freedom to fit the software to your business objectives. Let me try to make clear why this is my opinion and how this influences your role as a user or provider of ICT services.

Recent trends have shown that release cycles of software have become shorter and shorter; to keep ahead of our competitors you have to be able to release early and often.

Release early and release often as a way to be able to quickly add new features for the future, fix problems for the current software and to prevent that software becomes obsolete ("end of life"). The perpetual beta as the new adagium. Where flickr was able to deploy during it's booming period a new codebase every 15 minutes, Microsoft was able to make a new operating 7 years. One is using the "service" (cloud) way of offering it's software, the other is using the "fat client" approach. So on one hand proprietary fat client software s facing competition from cloud based services. Sure Google Docs is not as feature rich or reliable as the Office suite of Microsoft but most agree that this is just a question of time and network reliability; in due time Google Apps will be good enough for the masses. Mind you, most if not all cloud services are proprietary and are doing well; salesforce as the most prominent example.

On the other hand, proprietary software faces problems from the Open Source alternatives. OpenOffice.org is a real competitor for MS-Office, the Ubuntu distribution beats Microsoft in many areas and MySQL is giving the absurd licence fees of Oracle a hard time. If proprietary closed source software wants to stay in business, they have to move. Not to a "long tail" niche but in the other direction, to the left side of the tail where you can offer a highly standardised yet customisable version of their product. That way they are able to release early and often and go for a low margin per product sold but sell a lot. So I do think that closed source software has to move towards a service model, away from the client into the data centre.

This means that the other trend (Open Source Software) -that has written "release early and often" written all over it- will dominate the Do It Yourself area. OSS will be used by people and companies that have time and resources to fulfill their needs via highly customisable software. You will see this first with applications that are by nature webbased; the can move to the cloud with less legacy baggage. Software with much interaction with local legacy products will follow later, much later in some case So Office Automation for existing companies will take some serious time to migrate to the cloud since hybrid solutions (some data local, some in the cloud) will be rather expensive and complex to many from security, identity and manageability point of view.

One of the webbased applications that will dominte the "DIY" will be Drupal. It i already the best Content Mangement System ("looking outside") on the market and it is moving more in the direction of the core of business processes ("looking inside"). Drupal will more and more be used as both a frontend system and a backend system; a system where you can aggregate and enrich data for internal use that can be pushed towards for example an external Drupal site.

If you follow this logic (proprietary moving towards commodity cloud service, Open Source solutions towards customisable client service) you might conclude with me that Open Surce CMS-es have nothing to fear from closed source CMS-es like sharepoint. Sharepoint will be the shell around your office data if you want to use that from a cloud perspective, Drupal will be used by enthusiast and enterprises that need more power and have more resources to kickstart and operate that power.

So some people will use an iPhone and the cloud service "Mobile me", others will build Android. Some will use digital TV solutons from their cable providers, others will build MythTV. Some will run an OpenID service themselves, others will use it from a Google/Yahoo! And some will use voicemail (the most used cloud service in the world) and others prefer a local answering machine. I, I use all kind of differtent services, cloud and local, like most people will do.

PS: This posting as very late for last years' Drupal prediction posting or very early for next year, whatever makes more sense to you

PPS: Sure, you can have Open Source "SaaS" solutions as well, for example hosted and managed Drupal instalations but it will be a niche crossover, if that makes sense to you. Also, when I say "build", it can also mean "let other build", aka buy.

PPPS: I do think that SaaS is a complete wrong term; it is a technological acronym. First, people do not want "Software" as a service, but they want a service (as a service). As long as the ICT things about acronyms like SaaS, true adoption of using a "Service as A Service" will only stall. It is time to stop the technology lingo where it should stop; at the door of the customer and think of services instead of software. Second, Software as a Service is a very limited view on what truly can be accomplished with services; it might be disk capacity from the cloud (like S3, Storage as a Service), it might be CPU capacity (like EC2, CPU as a Service), it might be housing (Rackspace as a Service), hosting (Linux box as a Service) or to give an everyday example we are used to, voicemail (Answeringmachines as a Service). Therefor I plea to stop using the term SaaS and use XaaS ("Anything as a Service") or use SaaS for the acronym "Service as a Service", whatever makes more sense to you.

Via the Twit network I found out about a great serie of Podcasts over at Magnatune. They are free (as in beer), DRM free, advertisement free and well categorized. Sure worth listening to it on your iPod.

Talking about iPods, yesterday I bought the new fat Nano, my second iPod and .. my first Mac; an iMac! I ordered them online and hope to have them in a week or so. In even more unrelated news, Apple is worth more then IBM! (well, maybe not, but still)

As said before I really like the Many Eyes service of IBM. Today I wanted to get some historical data on the temperature in the Netherlands and it took me real long to find it, but in the end at KNMI.nl there was the data per month from 1706 to now.

I uploaded it to Many Eyes and here are some graphs (click on the picture to play with the visualization)

And a different way of looking at the same data:

Tomorrow I'll blog about why I needed this data in the first place.

I have been playing lately with the excellent Many Eyes site of IBM. And it is really great what you can do with data in time. A very small example is shown above. Note that during December 2006 I disabled a block with lost of pictures (resulting in less KB transferred and hits but not pageviews) and the data for February 2007 is .. well not yet complete.

Another way of looking at the data:

I would love to get my hands on the Drupal.org apache data so we can do even more analysis, IP to country to create an easy GIS data in time to see where visitors are coming from etc. I know Analyics (I have an account) does the same but having this data public available so people can reuse it themselves and make graphs of it themselves would be a better -more open way- for an open source project without targeted banner advertising like Drupal.

So what would you like to do with the webdata of a top1000 like Drupal.org if you would get your hands on it?

RSS readers: most likely your aggegator / RSS reader filters out the images, in that case click the tittle to visit this posting)

The size of a large refrigerator.
5 Megabytes on fifty 24” disks.
2KBits/sq. inch areal density; 20 tracks/inch; 100 bits/inch
8.8 kbits/second transfer rate.

On flickr

Also be sure to read the research site of IBM regarding "advances in disk drives to corresponding trends in storage systems and projects where these trends may lead in the future."

I just ordered a big NAS with some Terabytes for our webhosting proposition. Now by the time it hits the datacentre, it will be half the value worth I paid for it.