The following was sent to Digital Life (an Tuesday insert into the Straits Times). Bold and italicized word/lines was edited out though in the dead tree edition that came out July 3rd 2007. I am also putting out the original OpEd that was significantly watered down.
The one that got printed:
Building for the Future
It has to be acknowledged that the IDA does have a good amount of forward planning in place when they called for the S$1.5 billion Standard ICT Operating Environment (SOE) project.
The tender specifications (publicly available on http://www.gebiz.gov.sg) were silent on the technology choices, letting the four consortia to respond as they saw fit – and that is a good thing.
We are in a sense re-building and renewing the government IT infrastructure that must survive the test of time and the vagaries of technologies. What is compelling in 2007, might be seen to be a big mistake in 2017. We are all aware of the wisdom that comes from hindsight. So, what has all of that to do with the SOE?
For all the technology neutrality, the SOE is being proposed on the back of an non-existent national document format. A national document format is a fundamental pre-requisite for building, manipulating and archiving information within the government and to a larger extent the whole nation.
Knowing what we know today about the fact that many a times, we are not able to retrieve electronic documents created using earlier waves of applications, we have to ensure that this mistake is not repeated.
Let’s consider an analogy. If there was a call to build a road and were totally neutral about the measurement schemes, then it is possible that proposals could come in metric, imperial or some other proprietary measurement schemes. We know that that is not the way to build a road or anything else for that matter.
In that light, the basic measurement scheme of any IT project – SOE or otherwise – is the data format. If the data format is not stated upfront – whether open or proprietary – how are we to assess the value of the proposals? We have an even bigger problem here, because data (whether text, numbers, images, audio, video) needs to be readable long after the applications that created them have been displaced. The concept of document fidelity is extremely fundamental here and no project can ignore it.
As has been gleamed from the SOE specifications, there is no stated document fidelity requirements nor indications of how these documents are to survive applications and be readable in the future.
Industry intelligence indicates that all the solutions proposed for the SOE are based on proprietary, non-published document formats. This is like the road project being proposed in a measurement scheme that is only known to the special measurement tool. If you loose that tool, or the tool becomes obsolete, there is no easy way to recover.
If the SOE had stated that, as a nation, we want to be able to have document fidelity and long term document survivability, then the solutions could compete on what does the job best. There is already an internationally approved and recognized document format called the ISO 26300. If the SOE had mandated that all the solutions must ensure compliance with ISO 26300, then all information gathered during the sunrise and sunset of the SOE project, will be accessible.
The need for document fidelity is well appreciated by the Ministry of Defence (who are exempt from the SOE) and have essentially standardized on ISO 26300 (through their use of StarOffice and OpenOffice productivity suites). So, what would now happen is that government documents will have at least two different, incompatible formats – one for Mindef and one for the rest. Lessons from the tsunami disaster of 2004 in Thailand and the Hurricane Katrina disaster of 2005 in the US where agencies could not exchange data because of incompatible formats seem to be forgotten.
It is still not too late. Changes, extensions and clarifications to the tender specification are still possible before it becomes a serious problem.
An important benefit of adoption and mandating of ISO 26300 is that the government can easily switch out and use applications without any concern of document fidelity. Independence from document format lock-in has to be vigilantly defended. Do we need to have to have a Rosetta Stone for document formats?
The original version:
Building for the Future
It has to be acknowledged that the IDA does have a good amount of forward planning in place when they called for the S$1.5 billion Standard ICT Operating Environment (SOE) project/tender.
The tender that had got four consortia to respond is, unfortunately, exclusively centered on proprietary technologies that does not at all consider solutions that can be built on published open standards and deliverable using free software.
Singapore today does not have a standard by which all documents created by the government and its related agencies has to be in. As of May 2006, there has been an international document standard called the ISO/IEC 26300. This standard guarantees that any document, spreadsheet, drawing or presentation created will be readable forever, irrespective of the application that created it.
What that means is that even if the tools used to create them are no longer available, the contents can be accessed and used with whatever tools available at the time of access- all because the format in which they were created is public and fully documented. This seems to be a very obvious consideration, but sadly, something that seems to have been overlooked by the IDA.
So, if the SOE specified the mandatory use of an open, standard, document format, like the ISO/IEC 26300, then there will be a much wider choice (and possibly lower costs) of applications that can be run on government desktops.
But as it stands, none of the four consortia are known to have put forth any solution that actually creates and uses open, standard documents – chiefly because there is no indication, interest nor incentive from the IDA to do so. To the best of the my knowledge, all of them have proposed solutions that use Microsoft’s office formats which are not fully open and are known to have significant lock-ins.
What would this mean in the long term? The SOE is to last at least 10 years. In 2017, would the Singapore government still be beholden to Microsoft’s tools to read it’s documents? Wouldn’t the Singapore government want independence from vendor lock in and be able to substitute any other tool (be it free and open source software or closed, proprietary ones)? Would we as a country not want to be able to innovate and be able to ensure that our documents will stand the test of time and survive vendors? Why should the tax payer be burdened with, and held at ransom by a foreign commercial entity? Shouldn’t we be maximizing citizens value rather than a commercial entity’s shareholder value?
It could well be an urban myth, but it has been stated in many circles that Singapore is a shoe-in Microsoft shop; that the Singapore government will never deploy on a wide scale non-Microsoft technologies
for fear of the Redmond company pulling out of Singapore. When did we surrender our national sovereignty to a corporate entity? How much money does Microsoft actually invest in Singapore that we are afraid that they will pull out and cause the economy to collapse? By all means use Microsoft’s tools, but do so ONLY with a stated open, fully published document format policy.
It is insufficient to say that “let the industry decide”. Large parts of the IT industry are in a state of paralysis because of documented monopolistic and predatory business practises of Microsoft and desperately are seeking ways out. Continuing to close an eye to this is not helpful for the long term independence and growth of the Singapore IT landscape.
Why would we still want to look at the desktop metaphor as the way it will be for the future? The on line way (did someone say Web 2.0?) of creating, organizing and moving data means that anyone in government in 2017 should be working not from a 2007-vintage desktop, but from possibly handheld devices anywhere, anytime securely.
But for all of this to happen, tools and solutions need to function in a without any concern of document fidelity. Alas, the SOE as defined, does not provide for that.
I would urge the IDA to step forward with bold visions, foresight, technical and thought leadership in, first putting in place a national document format (like the ISO/IEC 26300) and, second in leveling the playing field so that free and open source providers can compete on an equal footing.
Let’s make sure that the SOE succeeds, but equally important, let’s do it right.