FOSS A General Introduction (Case Studies)

From OODA WIKI
FOSS Introduction.jpg

Table of Contents

Case Studies

Case Study: FOSS in Government

Introduction

The city of Largo is one of the earliest high-profile cases of a government administration migrating over to Linux. The IT system of this small city in the state of Florida, USA, supports 800 city workers, including local safety and health services. Implementation began in 2000 and their experience with Linux in the years since then have been nothing but positive.

Motivation for migrating to Linux

In 2000, the IT department of Largo was evaluating upgrade options as problems were being encountered with existing OpenServer and UnixWare products from the Santa Cruz Operation. Various options were evaluated, including Microsoft Windows on personal computers. However, since they were already on a Unix-based thin-client infrastructure, the combination of hardware and software costs involved in such a migration was deemed prohibitively expensive. Additionally, the IT team did not want to be locked into a 2-3-year upgrade cycle, where they would be forced to pay upgrade costs even when upgrades were not necessary.

Ultimately, the decision was made to keep the existing thin-client infrastructure but migrate systems to a Linux system based on Red Hat's distribution.

Implementation approach

A solution was tested and implemented starting in 2000 and completed by mid-2001. Two powerful (for that time) dual-processor Compaq servers delivered the services needed by most users. A variety of FOSS and non-FOSS applications were combined, including Netscape (web browser), Evolution (email client) and WordPerfect 8 (word processor). Heavy-duty database needs were run on a proprietary Oracle database while Microsoft's Excel and PowerPoint were made available to Linux users via a combination of Windows NT and the Citrix Metaframe server. In total, there were about 20 different servers working together, running a mix of Linux, Windows and Unix operating systems.

On the desktop side, things were simpler. The thin-client model requires only the barest minimum from desktop units. Hence, desktop units could be obtained at a relatively low cost. In some cases, the IT team managed to obtain desktop systems for as little as US$5 per unit. With 10-year lifespans and few moving parts, these desktop units rarely broke down and had a longer useful lifespan than normal PC desktops.

Results

The migration to Linux was estimated to have saved the city as much as US$1 million in the first year alone. Largo currently has an IT budget that is only about 40 percent the size of comparable cities. Where cities of a comparable size normally spend 3-4 percent of their city budget on IT, the Largo team gets along quite comfortably with only 1.3 percent of the city budget. The efficiency with which Linux uses hardware has also reaped enormous savings. The IT team estimates that they will not need to upgrade their desktops until 2007.

The reduction in number of personnel required is also significant. The end-user help desk requires only two to three people to support a user base of 800 workers. This low ratio is attributed to the reliability, stability and predictability of the system. The remaining staff members of the city's 10-member IT department are then freed for other tasks, including making additional improvements to the IT infrastructure.

For further reading

Case Study: FOSS in Education

Introduction

The Goa Schools Computer Project (GSCP) was launched in the Indian state of Goa to provide affordable computer labs to secondary and higher secondary schools in the state. The first pilot projects were launched in 2000 and after evaluation a second, larger project was launched in 2002.

The GSCP is a collaboration involving public, private and NGO organizations. The Goa Department of Education, Red Hat Linux, the Goa Computers in Schools Project NGO and the Goa Sudharop Community Development Charity all contributed to making this project a success. Using recycled computers and the FOSS GNU/Linux system, a total of 125 schools received computers that otherwise would not have been available to them.

Motivation

Cost was a primary motive for using the GNU/Linux system, particularly the licensing cost of proprietary software. Because the project decided from the start to recycle computers (also for cost reasons), finding software to place on these systems became a major issue. These systems were typically received with blank hard drives, due to concerns over security of the organizations donating the computers. Purchasing software to run on these systems would have multiplied the costs of using these computers manyfold.

By going with the recycled computer/GNU/Linux combination, the GSCP was able to install systems for as little as US$35 per system, with full computer labs, including networking, costing less than US$500. Proprietary software for a single computer would have cost at least US$400–500, many times the cost of the computer itself.

A comprehensive costing was performed for this project. Based on data from previous projects in other Indian states (Andhra Pradesh, Karnataka, Tamil Nadu and Kerala), it was estimated that the GNU/Linux/recycled hardware model would save as much as 77 percent of a traditional solution (proprietary software, new hardware). Combining GNU/Linux with new hardware would have saved 64 percent of the costs of the proprietary software/new hardware model.

Implementation approach

The GSCP used refurbished computers imported from wealthier, more developed nations. These computers were typically outdated models, replaced in regular corporate upgrade cycles. After testing and refitting as necessary, the computers were installed with the GNU/Linux operating system. The larger installations (labs with more than four computers) used GNU/Linux in a thin-client configuration.

Each computer lab was typically a cooperative effort between GSCP and the local school. GSCP would supply the equipment and teacher training while the school would supply the UPS, wiring and furniture for the computer lab. Once set up, the computer lab would be used by the schools during school hours and by the community at large after hours.

Results

A survey carried out one year after the computers were shipped found that 90 percent of the PCs had been installed and 79 percent of the PCs were operational. Schools using the thin-client model, which were also the schools that received four or more PCs, fared best. Urban schools fared better than rural schools for a variety of reasons, including better support and a larger number of available computers (due to larger student populations).

The schools are now charging 20 cents per student to pay for maintenance and Internet access costs. Pilot experiments are also underway to test the sustainability of charging the community for after-hours access to the computing facilities and the Internet. Experiences from projects in other countries have shown this to be feasible and it is hoped that it will be just as successful in Goa.

For further reading