Realization of FLOSS in the Government of Maldives (GOM)

Republic of Maldives is a small group of islands in the Indian Ocean with a disperse population of 300,000 to 200 islands. Out of this the capital, Male', hosts about 70,000 and is the center of all economical and government activities. Given the nature of the nation and it's economy, in the recession of 2009, Government of Maldives (GOM) is looking at the possibilities of turning to Free Libre Open Source Software (FLOSS) as a means of cost cutting, stability and freeing from vendor locks. On June 2009, Communication Ministry (CiviCom) of Ministry of Civil Aviation and Communication (MCAC) migrated all 12 desktops to Ubuntu as a pilot project. This paper describes this modest initiation of realization of FLOSS in the GOM.

Organization and Political Background

With the help of Maldives Open Source Society (MOSS), an NGO advocating and promoting FLOSS in the Maldives, the GOM has decided to see the possibility of using FLOSS within the Government. After a presentation by MOSS at the National Center for Information Technology (NCIT) along with Mr. Mahumood Razee, Minister of MCAC, officials from President’s Office, Ministry of Finance & Treasury, Communication Authority of Maldives, Customs and Police on 22nd June 2009, Mr. Razee decided to migrate CiviCom as a pilot project to see the practical feasibility of the use of FLOSS in the GOM. The GOM expected, in particular, positive security effects (less vulnerability to virus and hacker attacks) and lower licensing costs in this move.

The GOM opted for FLOSS primarily because of the costs reduction, the fact that it is legal and the persistent support and backing of MOSS. The idea was easily taken up as it was a time when the Government was looking for huge cost savings across all areas. Mr. Razee explains "We definitely have to find ways to see if we can lighten the huge burden IT expenses place on the Government budget. So I support the use of FLOSS in the GOM and with thanks to the initiation of MOSS, let us do a pilot project and see how this works in a practical fashion." Hence, this would be the reference project for the use of FLOSS on client environment throughout the GOM.

The pilot project was initiated on 28th of June 2009 with the presence of Mr. Razee after a brief introductory meeting with the Permanent Secretary Mr. Latheef and Senior Staff of CiviCom. CiviCom was chosen as a pilot project because of the support and willingness of Mr. Latheef.

Migration Process

Before the migration all clients were using proprietary software applications. The applications on daily use were in the fields of word processing and spreadsheet management using Microsoft Office package, communication and sharing of information through the use of Firefox via Internet as well as file storage via Network Attached Storage (NAS).

The migration to FLOSS was led by MOSS experts, due to the lack of open source knowledge, employees of CiviCom were not involved in leading the project. In-depth preparation for the migration made sure that daily work processes were not affected considerably.

As this was a pilot project of 12 computers, it was decided to make an image backup of all partitions of all hard drives of 12 machines before the migration. So, in case if needed, things can be reverted as they were with less effort. Next, all user data, including documents, spreadsheets, emails, bookmarks and others were backed up to the NAS.

For the user Desktop, Windows XP was replaced with Ubuntu. Ubuntu was chosen because of it's popularity, stability, user friendliness and specifically for it's short and continuous release cycle of every 6 months which comes with enhancements and bug fixes as they become available. In addition to the fact that commercial support is available when required. OpenOffice.org was chosen as an office productivity suite, as it has an active community based with good support and a host of features that are similar to Microsoft Office. Firefox for Linux remained as the application for browsing and sharing information via the Internet. For email client Evolution was chosen to replace Outlook as Evolution was much similar in features and user interface to Outlook with an active development team and supportive community.

With expectations from the senior management, all users welcomed the change and were comfortable with the Ubuntu system from the very first day. Day to day tasks such as Internet browsing, file management, email and others were seen as fairly seamless. However, in the office productivity department there were some minor challenges due the differences of Microsoft Office and OpenOffice.org. Ad-hoc training as and when needed was provided to tackle the formatting changes that were required in converting the current spreadsheets and documents to open standards.

On the field as and when needed employees obtained training adapted to their fields of applications. Generally the focus of training was based on an introduction of the office application OpenOffice.org. MOSS support was also provided after the migration process to compensate for the insufficient in-house open source expertise and to supervise the employees.

Budget and Funding

The budget of CiviCom is entirely founded by the GOM. CiviCom budget is tight and with the introduction of FLOSS the budget remained the same, mostly due to the volunteer support of MOSS no additional costs were involved. With regards to the the sum spent on IT by CiviCom, no statements could be made, however with an estimate with respect to the acquisition of license software, the savings generated through the use of FLOSS would amounts to over 60% of the IT costs. As such, right now for CiviCom with no future upgrade license costs it is very beneficial, and certainly operating in an economically efficient way.

Software
 		Proprietary
 		Open Source
Desktop Environment X 12 Windows Vista Ultimate Rf49,189.80
 Ubuntu Free
Office Productivity X 12 Microsoft Office Professional Rf76,945.80
 OpenOffice.org Free
Antivirus X 12 Norton Corporate Antivirus Rf6,013.80
 ClamAV Free
Vector Graphic Tool X 2 CorelDraw Suite Rf5,885.30
 InkScape Free
Bitmap Graphic Tool X 2 Adobe Creative Suite 4 Rf64,224.30
 GIMP Free


Rf202,259.00
$0

Table 1: Savings by using FLOSS

This pilot project only concentrated on migration of the Desktops, however, in the case of proprietary software a manger portion of the IT budget is consumed by any back-end server implementation which can also be replaced using proven FLOSS implementations to further reduce cost of acquisition and operation.

Technical Issues and Challenges

The main challenge in the migration to FLOSS was the tedious formatting problems of the daily usage of forms and documents. Users complained about the shift in ways they were used to work in comparison with the proprietary equivalent. Some compatibility problems complicated the exchange of files with external agencies who are still using proprietary formats and software, and it should be noted that this compatibility problems are due to the lack of adequate support for open standards by the proprietary software applications being used. In the first few weeks users perceived that this lack of compatibility problem increased the time spent to carry out their required tasks. Due to the limited in-house knowledge of FLOSS, MOSS had to continue providing support and adhoc training. However, such support and training is mainly only important at the initial stages and becomes less necessary as users become familiar with the FLOSS environment. Positive effects of the migration were an increase system security as well as some application specific advantages such as multiple desktops.

After the event, the migration team came to the conclusion that better integration of the CiviCom employees in the migration process would have resulted in higher productivity and higher user satisfaction levels. The role of the users were limited to testing and error reporting. By integration numerous initial difficulties and problems could have been prevented. As this case of migration shows that integration of user involvement, especially at the early stages, a policy of actively disseminating information, and individual training are important. It is also important to have discussions and knowledge sharing to convince migration opponents. Moreover, it is important to build up the knowledge of the in-house IT administrators in order to reduce dependence on external personnel.

Office Suite Compatibility

As an office suite, OpenOffice.org, which offered much the same functionalities as Microsoft Office was used. At the time, OpenOffice is at it's version 3.0 and feature rich. Also, OpenOffice.org handled Thaana, Dhivehi Language, specifically Right-to-Left writing much better than other open alternatives.

The most important point to note is that OpenOffice.org by default use open file formats. By using open standards, we guaranteed that the files can be readable anywhere in the future, even within the next 50 years. This important requirement was covered by OpenOffice with the use of XML-based open format.

Other technical issues related to the use of OpenOffice.org were possibly caused as still the current de-facto standard was to use Microsoft Office documents in the GOM. A total migration is the obvious choice in the future, as this will solve several problematic issues that remained. While compared to Microsoft Office certain functionalities are still limited in OpenOffice.org to which we were able to come up with work around solutions. One example was when copying and pasting from OpenOffice.org spreadsheet to documents it is handled very differently than when copying and pasting from Excel to Word. However, OpenOffice.org next version has highlighted developments which clearly indicate minimization of these sort of important differences.

Change Management

Before the migration the most important factor looked into was which functions does the users use. As FLOSS environment give numerous alternatives, several different applications where in question and so testing some of them before selecting was practised. More importantly not many people had working experience of FLOSS applications in a productive environment. Since the aim of the pilot project was to change current usage of software with an alternative from a host of alternatives, it was mostly choosing software which works and looks similar to the current applications to which the users are accustomed to. However, note should be taken to the fact that there are many other varieties of applications to choose from, if need be.

After a short presentation of why the move to FLOSS has been considered, a user experience demonstration of the software for the users was presented. So in order to tackle any resistance, mostly to the change, an installation of Ubuntu and applications to two shared computers gave the users some time to get used to the user interface. This way users acceptance for the FLOSS grew with less resistance, and any fears and discouragements was diminished. At the same time, it showed the users the added values of the new software. Some eye-candy, 3D desktop rotation and instant desktop flips without any performance issues were a factors that were important in gaining user acceptance.

More importantly the security aspect was highlighted, FLOSS do not get attacked by viruses like Windows does and is generally more stable. As the CiviCom IT Administrator said mostly referring to the virus based issues "This will make my life free from useless headaches!"

Training

To save time and given the small number of users, no initial training were given to the users before the migration. Since all users were fairly comfortable and experience in using IT resources the new software was not considered too difficult. Short ad-hoc trainings on the use and interface of Ubuntu, where and how to find their files, how to access the NAS and the use of OpenOffice.org was enough for the users. In case additional questions were raised experts of MOSS were available for help and assistance. The functionality of FLOSS applications compared to proprietary software was much the same, so although the interface was a bit different it was seen that user easily found their way logically.

Effect of the Services & Corporation with other bodies

There was a little resistance to the migration process as far as the daily work was concerned, mostly due to the the extra work required in formatting the current data to be usable in the new applications. However in the case of functionality the users did not find any problems, slow down or bothered in the way the work was done. Even in the issues related with sharing the data with other agencies were not considered a problem, they just learned independently to convert the files to required formats before sending them. When the staff who was on holiday while the migration was done returned to office, a brief introduction was enough to get productive.

One major problem was seen in the usage of the custom designed Microsoft InfoPath files that was made mandatory by the Ministry of Finance. Since there is no direct equivalent software for Microsoft InfoPath and it does not run via Wine under Linux, we had to use virtualization using Sun Virtualbox. One other major issue was a complex budget report which was automated using Excel spreadsheets using some complex Excel formulas, another mandatory report that has to be submitted to Ministry of Finance. However with the upgrade of OpenOffice.org to version 3.1 (from the default version 3.0 that comes with Ubuntu 9.04) apparently a lot of compatibility issues with Microsoft Office were addressed.

One has to agree that these incompatibility problem are mostly organisational rather than technical.

Evaluation

As in the past Windows 3.11 to Windows XP was an overnight migration with users adopting and learning to the needs. Similarly and as experienced by the pilot project, an overnight migration to Ubuntu desktop is also a feasible option. However, the pilot project should be considered within the context of an stepping stone and every future step should be carefully planned for the successful deployments in the future.

As seen by the pilot project, it is conclusive enough to say that there have not been any complaints or reports of significant difficulties from end users so far. However, as the pilot project is a minor operation with only 12 machines, no definitive answer can be given in this respect. The next step would be a meeting with officials of GOM to discuss both the future of the IT infrastructure and evaluate their opinions. The pilot project has provided some very important feedback, which has given a good view of possible problems and it looks rosier at the moment.

From a financial point of view, it should be noted that though the option will be less burdened on the budget, it would always be higher than the current status quo. Nonetheless it is very important to look at the cost in comparison. As far as figures are concerned the cost of accruing Windows licenses would exceed far greater than going for a FLOSS migration, as not only hardware updates would have been necessary, but future software updates involving costly licenses would further add to the figure. Since FLOSS have relatively low hardware requirements comparatively, it can even make the older machine functional, and can be installed at zero cost. Hence, the cost of such a solution would always come with cost savings. Moreover, Linux has been fairly stable for the past 10 years, reducing the administration and maintenance cost and time.

Achievements

The migration of 12 computers of the pilot project has been accomplished successfully. The openness of the FLOSS and the community support has made it possible to adapt and improve the system to the needs, which has brought many advantages compared to proprietary software to CiviCom.

For MOSS the project is a very important and unique source of achievement. As there are very few similar ventures in the Maldives, as far as desktop are concerned, the insight gained here will be valuable and shared to any future deployment. As one example, where the firefox browser had some problems accessing the Civil Service Veyo, the MOSS community was able to come up with a solution and shared the solution on the MOSS mailing list. For any large scale deployment this is important, as the occurrence of the same problem could be tackled with less research at the same time increase the confidence in the use of FLOSS. Hence, the pilot project was a mutual success to the GOM and MOSS alike.

Future Plans

With the success of this pilot project and the support of other Government agencies, this will lead the way for a larger scale nation wide deployment of FLOSS in the GOM and thus use of FLOSS is expected to be replicated in the rest of the GOM in the near future.

For the decision to migrate to Ubuntu, no legal steps are required if in-house expertise are used. With regards to the pilot project, external support was accrued free of charge, so no legal steps were required. However if the GOM decide to use contracted expertise of a commercial partners, additional cost will be involve in the process hence, legal procedures will come in effect. Hence, for any future large scale deployment with the help of a contracted partner the GOM is obliged to follow the standard procurement steps:

  • Create clearly defined public RFP with all requirements and specification.
  • Finalize and approve both the migration process and budget with the cabinet.
  • Publicize the RFP.
  • Collect commercial responses to the RFP.
  • Evaluate and select the most appropriate partner.

With the success of the pilot project, it is enough proof of concept, that a large scale migration will not only dramatically improve the GOM IT system, but save huge sums of money that otherwise would have been spent on licenses.

Conclusions

In the pilot project the challenges are natural as in any migration process to a different platform. These were mainly technical issues related to the transition that occur in the early stages of such a project. The continuous support of the Minister, Mr. Razee, the Permanent Secretary, the willingness of the staff and a little time spend on optimizing contributed to the success of the CiviCom transition. With similar supports and efforts, a large scale deployment is easily doable, which would result in huge benefits to the GOM that would shine through the years to come.