Polishing old gems

As I have done a few snippets of code here and there to make my life a little easier, I still have not found the ease with which I did things when I used my own Trollsilm SQML, not to be confused with SQml^TM - Which reminds me I should change name of the language while I still have the chance.

I wrote the SLOB database in 1999, the markup language in 2001, and stopped developing it ca 2005 due to time constraints. The tool is still being used by an ISP in Norway, hoping that development will commence. Because just like myself, their experience is that they just haven't found any alternative that allows you to whip up full blown web applications in such a straight forward and intuitive fashion than this.

This might sound like bragging, but as I am now faced with new challenges at work, I long for this tool. I looked up "the competition" (Ruby-on-rails), and found out that they were not really competition after all. Does SQML - or should I call it "TrollML?" - have the right to live again? Should I spend time continuing its development?

You betch'a! There are caveats, of course. Currently, the SLOB database is the one getting the most out of the language. I need better SQL integration that makes it as easy to deal with mySQL as it is to deal with SLOB. And most of all - an Apache module release. I have a good list of things that need to be done, and in which order. And most importantly, I can feel my fingertips getting tingly in anticipation to work on this project again.

Business model? Open source. It's not the technology itself I wish to monatize on in the future, but assisting people who wish to use it.

CIPA part 5: National solutions

Consultants in Public Administration: Part 1 - Part 2 - Part 3 - Part 4 - Part 5

Employees of the Norwegian Wealthfare Agency (NAV) complain about how their systems are outdated. Some must even "learn DOS commands", which we know is the umbrella term for black windows with white text where you must actually use the keyboard.

NAV is a "recently" established cooperation between local and state government bodies. Some of the IT resources are owned by the municipalities, while others are owned by the national government, and there is little information being passed between these as a result of content rights and personal data protection directives. Setting up a NAV office requires some planning to enforce all the information flow/blocking policies. Even copy/paste is disabled across the platforms, so you can't copy information between the municipal and the national systems.

And this is where the pain begins.

As each municipality runs their own systems, NAV must attempt to integrate all of them into the big conglomerate. But neither state nor NAV has the authority to override what is installed at the local municipalities as long as it conforms to certain integration standards. Because if national authorities dictated a shift of technology, then national authorities would be liable to pay the bill. In the short run, the cost of integration seems to be smaller than the cost of conversion. By publishing the integration interface, much of the integration cost is now pushed over to the municipalities as part of their annual software license fees.

The cost of convertion, as such, is what we my refer to as "the cost of exit." As each system has an entry and a maintainance cost, there is also an exit cost. What does it cost to change system? And what system should be our standard? After all, setting a national standard would kill the business of all the other tailored government system developers. Some commercial actor would be able to monopolize. So how do we get out of this?

In 2006, the Swedish police turned to an open source platform and hired its own software developers to build their own mySQL and JBoss based information system. From this, they gained tailored software, a flexible system that can quickly be altered to reflect the needs of the police, full ownership of how the system works with their own data - and economically? The project pays far more than the investment in software developers. Reduced licensing and consulting costs equals approximately 400 fully equiped police cars per year.

A similar approach may be used in other government bodies, where specific tasks have been defined. This is what you do, let's all do it on this one system which we all own. If a law changes, we can change our own software - or pay huge sums to 4-5 different software companies to modify their software for the new law.

And one could take this a step further by setting a national standard for ESB - or would that be a GSB (Government Service Bus)?

Consultants in Public Administration: Part 1 - Part 2 - Part 3 - Part 4 - Part 5