Requirement Analysis And Defining Use Cases

Published Date: 02 Nov 2017

The aim of this chapter is to provide a description of the study area. Basically the description is centered around the relevant information required to understand the nature of the problem being addressed in this study.

Papua New Guinea lies 06° 00' S and 47° 00' E of the equator with total land area of approximately 462 000 square kilometers. It covers the Eastern half of the island of New Guinea. PNG is incredibly diverse in terms of linguistics, culture and biodiversity. The population was estimated to be 7 million in 2010. Almost 80% of the population lives in rural areas, scattered across the rugged terrains and numerous islands which makes service delivery very challenging, chaotic and costly. There are 800 different languages spoken in the country. The official languages are however English and Pidgin (creole). The government system being democratic is based on the Westminster Model. There is a decentralized system of government where the three levels of administration at the subnational level are: provincial, district and local. There has been some economic growth in recent years. However, this has resulted from the increase in prices of mining products on the international markets rather than improvements through internal performance (Government of Papua New Guinea 2010). Three neighboring provinces have been chosen for the prototype, namely West New Britain, Morobe and Madang (see figure 5 below). These three provinces were chosen because they had a more complete health facility attribute information compared to the rest of the provinces. For all these provinces, focus was on vector data.

Figure 5 Map of Papua New Guinea showing the prototype provinces

Madang province is one of the four provinces that make up the Momase electoral region of Papua New Guinea. It has a land area of 29 000 square kilometers with a total of 6 districts, 19 llgs and 45 wards. The total occupied area is 16000 square kilometers with a population density of 22.6 per square kilometer. The population in 2010 was estimated to be 430 000. The main small scale economic activity involves the sale of cocoa, copra, livestock and food crops. Larger commodities includes Ramu Sugar and several mining and forestry operations (NRI Research Paper - PNG Provincial Profiles). Madang has a total of 229 aidposts and 47 health centers in 2010. These figures include both the open and closed facilities. The total length of road is 1900km. Figure 6 below shows the provincial profile snapshots.

C:\Users\barbara loked\Desktop\Capture.JPG

Figure 6 Madang province snapshot provincial profile Source: PNG NRI

Morobe Province

Morobe province is also one of the provinces apart from Madang that makes up the Momase electoral region. It has a land area of 34 000 square kilometers with a total of 9 districts, 33 llgs and 547 wards. The total occupied area is 11 708 square kilometers with a population density of 46.1 per square kilometer. The population in 2010 was estimated to be 700 000. The main small scale economic activity involves the sale of betelnuts and fresh food from those living in the valleys and fish, coca and copra from the coastal inhabitants. Larger economic activities include the numerous mining operations, forestry and cattle farming (NRI Research Paper - PNG Provincial Profiles). Morobe has a total of 289 aidposts and 49 health centers. These figures include both the open and closed facilities. The total length of road is 2370 km. Figure 7 below shows the provincial profile snapshot for more information about Morobe.

C:\Users\barbara loked\Desktop\2.JPG

Figure 7 Morobe province snapshot provincial profile Source: PNG NRI

West New Britain Province

West New Britain province is one of the island provinces that makes up the New Guinea Islands electoral region of Papua New Guinea. It has a land area of 20 500 square kilometers with a total of 2 districts, 11 llgs and 111 wards. The total occupied area is 5807 square kilometers with a population density of 31.08 per square kilometer. The population in 2010 was estimated to be 260 000. West New Britain province can be classified as rich with high earned incomes from Oil Palm productions as well as cocoa and copra. There is also high presence of forestry/logging operations taking place in the province (NRI Research Paper - PNG Provincial Profiles). This province has a total of 119 aidposts and 32 health centers. These figures include both the open and closed facilities. The total length of road is 2729 km. Figure 8 below shows a provincial profile snapshot for further information about West New Britain.

C:\Users\barbara loked\Desktop\43.JPG

Figure 8 West New Britain province snapshot provincial profile Source: PNG NRI

4. METHODOLOGY

As per the objectives of the study, the workflow procedures have been divided into four major phases which will be elaborated upon separately in the consequent major sub-sections. The workflow procedures for this study were implemented in the order shown in figure 9 below.

Figure 9 General workflow procedure followed in this study.

The Web based GIS application prototype developed in this study follows the three-tier Web GIS architecture where there is a client, business/logic and data tier as shown in figure 10 below. The client comprise of HTML containing JavaScipt. The business/logic comprises of Tomcat Apache and GeoServer. The data tier comprises of PostgreSQL/PostGIS. Actions performed by the users are converted into HTTP requests by JavaScript. On the busines/logic tier, Java forms the interface standards that facilitate the communication between Tomcat Apache and GeoServer. On the data tier lies the PostgreSQL/PostGIS object relational database management system where the data is stored and served to GeoServer via the structured query language (SQL).

Figure 10 Web GIS Architecture followed in this study.

4.1. Requirement Analysis and Defining Use Cases

In Papua New Guinea over 80% of the population lives in rural areas, scattered across the rugged terrains and numerous islands which makes service delivery very challenging, chaotic and costly. In order for the government and donor agencies to allocate funding equitably for health services delivery, they need to take into consideration the influence of geography on the cost. Given that health facilities and population both have geographic space; the urgent need of a spatially enabled health information system cannot be stressed any further.

Defining Use Cases

It was realized that users don’t really know what they want from such an application, as is the case in Papua New Guinea. Unless you create and deliver it for use, they usually don’t know what they want. As advised, a better approach would be to create a prototype and distribute it for use in a way to get feedback from users. However several use cases have been defined for starting. The approach for this study is to create a working prototype for possible future extensions and further development. The target audiences are professionals from the public, private, churches and donor agencies who play a role in health service delivery in Papua New Guinea.

Below is a list of use cases:

Donor agency is in charge of the distribution of medical supplies to rural health facilities. They need to know where the health centers and aidposts are located. They need to know how to get to those locations – the transport modes from point A to point B (road all the way or road and then walk or by air and then road) all of which have costing effects. They need to know if the facilities exist or are operating or not. They need to know which aidposts are supported by which health centers. They need to know accessibility and population so they can prioritize their intervention activities.

Planning officer within the department of health need to know the spatial distribution patterns of health facilities in relation to population distribution patterns. They need to know staffing levels in a particular district in relation to population and facility so they can plan the deployment of staff. If facilities are closed, the necessary actions should be taken to re-open for use based on the population aggregates within a particular area.

Provincial health officers also need to know the same range of information as for the donor agency and planning officers so they can ensure there is alignment of planned activities with higher levels of government. They need information of facility locations and accessibility modes in order to plan supervisory visits, immunization patrols and medical supply distributions once it gets to the provinces.

Church organizations who generously assist GoPNG with the delivery of health services also need to know the same range of information as mentioned for donor agencies, national and provincial health offices.

Ideally, if the prototype is accepted and further developed, there will be a central management unit within the National Department of Health with varying levels of authentication and user/administrator access rights to provincial health offices, donor agencies and church organizations to add and update data attributes as well as access data. When there is central data management and distributed access with varying levels of access rights to different stakeholders based on their service delivery functions, this may also promote the alignment of planned activities so that one stakeholder complements the other.

Further descriptions of the Use Cases can be seen in Table 1 below.

Table 1 Use Cases.

XXXXXX

**Additional stakeholders could include resource developers like mining and petroleum compaines who set up mining townships in the remote areas. They would need such information as well to plan how best they can contribute to delivering health services within the areas they operate.

Up to this stage, it can be gathered from table 1 that the main advantage of this Web GIS Application to its different Use Cases and stakeholders is to promote realistic planning of health services delivery taking into consideration the impact of geography on the cost involved in the different activities. Although there is no information on cost provided literally however the different attributes provided have cost implications. For example, if a health facility is supplied by air, that would cost more than that supplied by road. As such, provinces or districts with very remote facility locations should be given more funding to cater for this. The lack of such kind of information can lead to underfunding of activities resulting in no outcomes. Or it can happen the other way around where certain activities can be overfunded leaving other priority areas to be underfunded. Therefore as mentioned before, the advantage of this application, although it is not built to spit out answers, a user with sound analytical skills can be able to draw a lot of useful information which can be an end point or those that should lead to further investigations.

Objectives of the Web GIS Application

The application is intended for use by officials at the National Department of Health, Donor Agencies, Churches and Provincial Health offices for planning the delivery of health services to the rural population of PNG. It is not intended for use by the rural population as a tool for finding nearest health facility. The Web GIS Application should allow the target users to:

View and create maps of health facilities with other data layers provided depending on their layers of choice,

view data attributes for the data layers provided,

overlay the different available data layers to see spatial relationships and

perform simple queries or facility name search

Data

Based on the above mentioned use cases and the objectives of the application, the data were collected accordingly however one major setback is insufficient data attributes which some agencies did not have readily available. However the main data layers and some useful attributes were gathered. The data provided by the different agencies were given on the grounds of supporting research. Additional data layers which are not so relevant to this study were also gathered and listed in table 1. Please refer to table 1 for further details of the data layers with their attributes collected as part of this project. All these mentioned data layers are georeferenced and can be displayed on a map. The data are in shapefile format. In consideration of scalability for this application prototype, not all the data layers mentioned in table 2 is included. The data layers of high relevance that will be included are the health centers, aidposts, roads, vct sites, census units, areastore hc links, hc ap links, airstrips, major towns, provincial, district and llg boundaries.

Table 2 Description of data layers collected for this study.

Layers

Health Centers, Aidposts, roads, administrative boundaries, census points, catchments, boat routes, walking routes, airstrips, mining leases, petroleum leases, drug distribution links and HIV Voluntary Care and Testing sites (VCT).

Potential Map Use

Planning drug distribution, immunization outreach, facility planning, service accessibility map, emergency response map, facility spatial distribution map, road and airport maintenance planning map…

Data Sources

the National Department of Health, National Department of Works, National Statistics Office, The National Economic and Fiscal Commission and the University of Papua New Guinea Remote Sensing Center

Type of Representation

Points: Health Centers, aidposts, census points, airstrips, VCT sites

Polylines: roads, walking routes, boat routes

Polygons: Administrative boundaries, Mining leases and petroleum leases

Spatial Relationships

Roads must be connected, no gaps

Map Scale

Varying

Symbols and Labels

The above layers to be shown in different distinguishable colors

Attributes

Health Centers: Health facility Code, Name, Type, Agency, Lat, Long, Status

Aidposts: Name, Supervising Center, Admin District, Political District, Staff Male, Staff female, Status

Roads: Name, Section ID, Type, Surface, Condition, Gravel Definition

Census Units: Name, Geocode, Province Code, District Code, LLG Code, Ward Code, Ward Name, Lat, Long, Registered Household, total population, Male, Female, Growth rate, Population estimates 2008

Provincial Boundary: Name, Unique ID, Area in SqKm, Total Population, Male, female, density

District Boundary: Name, Unique ID, Area in SqKm, Total Population, Male, female, density

LLG Boundary: Name, Unique ID, Area in SqKm, Total Population, Male, female, density

Catchment: Name, Area in SqKm; Boat Routes: Length; Walk Routes: Length in Km; Airstrips: Name, Type, Class, Runway Dimension, Stopway Dimension, , Strip dimensions, Slope, runwary suface, administrative authority and status

Mining Leases: Lease Type, Lease Number, Lease ID, Lease Name, Sub Block, Holder 1, Holder 2, Holder 3, Equity 1, Equity 2, Equity 3, Bounds ID, Grant Date, Last Renewal and Expiry date.

Petroleum Leases: Name

Drug Distribution Links: Facility name, method of distribution and distances, time taken, cost involved per distribution

4.2. Tools and Software Review

Prior to the implementation of this study, a decision was reached to use FOSS for GIS. Since this study takes more of a proof-of-concept approach, FOSS was opted for to address economic and technical feasibility at this proof-of-concept level. This is not to understate the fact that fully fledged Web GIS applications can incur some cost and may require high level of expertise even with FOSS, however it is to show that a prototype for further improvement can be economically and technically feasible. In this discussion, the term ‘open source’ will be used interchangeably with ‘free and open source software (FOSS)’. The term FOSS is made mention of as it is widely used in the GIS realms. This section is structured as follows:

First of all, the terms FOSS, freeware and proprietary are defined.

Secondly, FOSS advantages and relevance to developing countries (Papua New Guinea) and their institutions will be discussed.

Thirdly, the tools and software review

Finally, tools and software selection and description.

4.2.1 FOSS, Freeware and Proprietary

As stated by Steiniger and Bocher 2009, a Free and Open Source Software (FOSS) is one which grants freedom to the user through its license to (i) run the source code for any purpose; (ii) study the source code; (iii) freely copy and distribute the source code; (iv) modify the source code, and (v) distribute the modified version. As part thereof, FOSS licenses grants the user to sell the modified versions hence the opposite of FOSS is not commercial but proprietary which puts emphasis on ownership. (Steiniger and Bocher 2009). Examples of FOSS for GIS would be QGIS and uDIG.

Apart from FOSS, there exist freeware, a term with no clear accepted definition that is used to refer to software that is freely redistributable but not open source. A freeware is sometimes developed and distributed as a component compatible to a particular proprietary and commercial software with the intention of promoting the demand of the proprietary/commercial product. In other terms, a freeware can be looked at as promotional components of the business models of commercial and proprietary software products. Hence the term freeware is attached more to proprietary software products. An example of a freeware would be ArcGIS Diagrammer which is only compatible with ESRI's ArcGIS Desktop product REFERENCE.

Proprietary software is the opposite of FOSS. This means that for a proprietary software, there is no access to the source code at all. Proprietary software is licensed under exclusive legal right of the copyright holder where the user is granted use under certain conditions and must accept an end-user license agreement (EULA). EULA is a contract between user and publisher, in order for an application to be installed on a hard-drive. Most proprietary software are commercial meaning the user must pay some fees for the license. Examples for proprietary GIS software would be ArcGIS Desktop and MapInfo Professional REFERENCE.

4.2.2 FOSS for GIS in Developing Countries

While GIS is vital to achieving socio-economic progress, one of the key constraints in implementing GIS in developing countries is the high technology price of the required software. Hence the impact of GIS has been marginal in some areas since most GIS related projects using proprietary software comes to a standstill when funding support comes to an end. As such it is important for developing countries and their donor partners to carefully review software procurements to ensure low-cost and options for open-source software products are properly considered and critically evaluated in terms of costs and benefits in the long run.

One of the impressive case study of the use of FOSS for developing countries is evident in the South Pacific Region where FOSS has been used for most GIS and Remote Sensing work. This was led by the Pacific Applied and Geoscience Commission (SOPAC) who made source code available to its stakeholders in the GIS for Utilities Project in 1997. This resulted in the processing and delivery of timely GIS outputs which are external vendor independent or free of licensing constraints. Another outstanding impact is the use of FOSS in universities with reference to the GIS and Remote Sensing courses offered at the University of South Pacific in Fiji. Students are exposed to different FOSS Desktop GIS for their lab practices. This has also resulted in non IT personnel increasingly becoming familiar in the use of Desktop GIS software thus increasing the usability of these software to a wide range of users. This brings about an opportunity for projects and institutions with little budget to develop their capacity.

In the case of Papua New Guinea, generally GIS is still at an infant stage as mentioned before. The GIS software used in the University of Papua New Guinea and University of Technology is MapInfo by Pitney Bowes. Likewise, MapInfo is also used in government departments like Planning and the National Statistics Office for basic desktop tasks like basic thematic map generation and viewing on an ad-hoc basis. However having said that, the issue in PNG is that the full use of GIS is still lacking in most institutions. This means first of all, there is a need for sufficient promotion of the usability of GIS for different sector specific applications or uses. Once that is achieved, the next task is to realize the technically and economically feasible software solutions to achieving these sector specific tasks. PNG being part of the developing world with a lot of budget constraints should opt for FOSS as an alternative to proprietary solutions. PNG NDoH is one of the departments with limited budget and many health outcomes to achieve. As such, the use of FOSS for GIS incorporated into its information systems to support decision making is highly beneficial and cannot be emphasized any further. When a government mandated body has fully realized the benefits of GIS and adopts the use of FOSS, then it will be required that universities teaching GIS should also use FOSS in their curriculum.

The motivations behind the use of FOSS apart from the obvious minimized cost can also be further looked at in the perspectives of independence and security. For instance, by promoting the use of FOSS, local IT and Computing firms can be given the opportunity to enhance their support and maintenance capability where and when needed by institutions using FOSS. This creates jobs for university graduates in GIS, computing and software engineering courses. There is no need to depend on foreign expensive software support and maintenance by proprietary software companies. This creates an environment of independence. In terms of security, open source software has been said to be more reliable as there are more expats checking the code, finding and fixing problems. Typical Government's tasks to provide free access to public information can be realized by the use of standards and open formats which is mostly instrumented in FOSS.

4.2.3 Tools and software review

In any GIS project, a careful assessment of the range of tools and software available is a vital prerequisite to choosing which ones to use. Decision has to be made based on project requirements, nature, tasks and scale. To achieve maximum output, it is necessary to assess tools and software and choose those with the required capabilities to execute project tasks.

Several studies have been conducted in the past to compare FOSS for GIS against either proprietary GIS software or other FOSS for GIS products. In such kind of studies, the evaluation criteria may vary depending on whether comparisons are being made among FOSS products or between FOSS and proprietary products. For example, Donnelly 2010 assessed FOSS product for GIS against ArcGIS desktop in their capability to create maps within a library environment. It was stated that the FOSS products had their own strengths and weaknesses with respect to thematic mapping and were generally weaker compared to ArcGIS in terms of support for various projections and coordinate systems, joining of attribute data and labeling. However FOSS advantage is that it was free of cost and licensing restrictions. The addition of plug-ins and helper applications increased the viability for FOSS. Another study by Hengl et al., 2009 involved a comparative analysis of GRASS and SAGA GIS (both FOSS) for automated analysis of elevation data. The result of the study showed that SAGA was more computationally efficient while GRASS generated more stream networks and offers more complex analysis of stream network data. Generically, there is no rule that dictates which software should always be used in any GIS related project. Choice of software is a subjective matter and is based on the user requirements and what they aim to achieve.

While there are many available FOSS products, the following sub-sections aim to give a review of the main ones in the game, as a prerequisite to choosing which ones to use. Relevant FOSS tools and software typical to Web GIS implementation that will be discussed in this section includes database schema/modeling software, database server/RDBMS software, web map servers, web servers and web mapping APIs. The advantages and disadvantages of the following tools and software are also based on user requirements and what they aim to achieve hence will not be emphasized in the following discussions.

Database Schema/Modeling Software

A database schema/modeling software in this study refers to a software that offers the capability of drawing up the logical schema of a 'spatial' database. Spatial database supports geometry and spatial reference systems. The logical schema, as the name implies, basically sets out the logic or real world behavior and relationships between the different data layers and their attributes. Logical schema is the end product of the logical design phase which is a three step process including conceptual design, ER entity relationship diagrams and normalization. Two of the known software that can be used to design the logical schema of a spatial database include ArcGIS Diagrammer and Moskitt Geo. Other highly functional tools include Open ModelSphere, Database Designer for PostgreSQL and Power Architect however their spatial support is not known. ArcGIS Diagrammer is a freeware developed by the Application Prototype Lab of ESRI Redlands and used as a productivity tool for ESRI Workspace documents REFERENCE. As a plug-in to the Moskitt database design software, Moskitt Geo which is open source enables the designing of spatial database by offering geometry support, EPSG based SRS and spatial index. It comprises a set of tools including UML, database modeling, work breakdown structure modeling, UI modeling, transformation and synchronization from UML to database and reverse engineering. The design process starts from executing a new UML Class Diagram to a Data Base Diagram Spatial Transformation then to the last Transformation depending on the final selected data base with spatial support where the Database Schema become a DDL (Data Description Language). MoskittGeo has support for Oracle 10g locator and PostGIS REFERENCE.

Database Servers/ORDBMS

In simple terms, a database stores an organized set of information that can be queried and a database server software is designed to handle the distribution of these information across a large group of people. As there has been progress in the development of databases software over the years, several approaches for storing and linking information have emerged. A relational database is one that is designed to follow the rules of normalization. However as the use of databases have evolved over the years as well, the use of relational database was seen to have several deficiencies which made it unfit to support the changing database use requirements. The first deficiency is that SQL-92 supports a restricted set of built-in types that accommodate only numbers and strings. However today, many database applications have begun to deal with complex objects such as geographic data. The second deficiency being that relational tables lack good support for nested structures, such as sets and arrays. The third deficiency is that RDBMS did not take advantage of object-oriented (OO) approaches to software engineering which have gained widespread acceptance in industry. These deficiencies in a relational database have caused the birth of Object Relational Database Management Systems. The Object Relational Database Systems extend relational database systems to support a broader class of applications, providing a bridge between the relational and object-oriented standards. With regards to FOSS for Web GIS, two of the most common database management systems are PostgreSQL and MySQL. MySQL owned by ORACLE is a relational database management system (RDBMS) first released in 1995 under the GNU GPL V2 license. It is cross platform and written in the C++ language. SimaËœo et al 2007 have employed the use of MySQL to store the alphanumeric data in their development of a Web-based GIS for collaborative planning and public participation for the strategic planning of wind farm sites. MySQL is used also by Google, Flickr, Wikipedia and Facebook. PostgreSQL developed by the PostgreSQL Global Development Group is an object relational database system written in C and is cross platform. PostgreSQL is released under the PostgreSQL License and has a strong reputation for reliability, data integrity, accuracy and security. PostgreSQL comes with many sophisticated features and has been tagged by most as the world's most powerful open source object relational database management system. Several Web GIS studies have employed the use of PostgreSQL due to its robustness including that of Puyam et al 2012, Green et al 2007 and Brovelli and Magni 2004. Some companies and institutions using PostgreSQL includes BASF Agricultural Product Division, Berkman Center for Internet & Society at Harvard University, Sternberg Astronomical Institute in Russia, University of New South Wales in Australia, University of Oslo in Norway, United Nations Children's Fund, United Nations Industrial Development Organisation, U.S. Agency for International Development, U.S. Centers For Disease Control and Prevention, GreenPeace, Cisco, Skype and Optus.

Web Map Servers

A web map server is designed to serve maps over the web. Essentially, a Map Server should be able to do three things. First of all, create a map in the form of a picture, or a series of graphical elements, or a packaged set of geographical feature data. Secondly, it should be able to respond to basic queries about the map content and thirdly, communicate to other programs what maps it can produce and of which ones can be queried further. Examples of two commonly used open source web map servers include GeoServer and MapServer. GeoServer is developed by OpenGeo, GeoSolutions and Refractions Research. It is written in Java, supported in Linux, Mac OS X, MS Windows and POSIX and released under GPL License. GeoServer allows users to share, process and edit geospatial data as well as its designed for interoperability and open standards. GeoServer being the reference implementation of the OGC WFS, WCS and WMS makes up a central component of the Geospatial Web. GeoServer has been used in several Web GIS studies including Jungialern et al 2012, Sarup and Shukla 2012 and Pascaul et al 2012. Companies and institutions using GeoServer include: Sweco, Greek Regulatory Authority for Energy, City of New York, EB Server AusNet, Landgate, Farrallon Geographics and Tike. MapServer is also an open source development environment for building spatially enabled Internet-Web applications. It is cross platform and released under the MIT license. Written in C++, MapServer uses special libraries enabling it to access various raster and vector data formats without the need for data conversion. Web GIS studies using MapServer includes Boulos and Honda 2006, Rathore et al 2010 and Flemonsa et al 2007. Companies and institutions using MapServer include: Hectares BC, Western Forest Products, US Army Corps of Engineers and AT and T Wireless.

Web Servers

A web server is a computer program that delivers web pages upon request by the user. For instance, a user enters a URL into a browser which sends off the request to view a certain web page. Web server is the computer program that responds to the request and delivers the web page content back to the user. This is made possible as every computer connected to the internet has an Internet Protocol address which enables one computer to discover another and communicate with each other across the network. There are various open source web server software including Apache, Nginx, Lighttpd, Cherokee, HTTP Explorer and HFS HTTP File Server. Among them, the world's most commonly used is Apache HTTP. The Apache HTTP, a product of the Apache Software Foundation is written in C and mainly used for serving static web pages. However with add-on modules, the Apache HTTP can modify static contents and serve dynamic contents written in Perl, PHP, Python and Ruby. The Apache HTTP is used for general web purposes hence is not of interest in this study. Another product of Apache Software Foundation which is used in most Web GIS studies with the aim of serving dynamic contents is Tomcat. Tomcat is a Java Servlet or Java Server Page (JSP) container used to serve servlets and JSPs which are dynamic. Tomcat can also host static contents on the web. For Web GIS, Tomcat has been used in the works of Sarup and Shukla 2012, Jungialern et al 2012, Harper 2006, SimaËœo et al 2007, Giridhar et al 2011 and Padmanadh et al 2010.

Web Mapping APIs

Often in life, one has to rely on others to do certain thing that one lacks the capability to or may not be permitted to. The same phenomenon also applies where all software requests other software to do certain things for them. To make this happen, the software sending the requests uses a standardized set of request protocol called application programming interface (API) that are defined for the software in which the request is made to. Literally, an API is a source code interface that a software program library provides to facilitate requests for services to be made to it by a computer program. An API is specified in terms of a programming language that can be compiled when an application is built. The software offering the functionality described by an API is referred to as an implementation of the API. So a Web Mapping API refers to APIs used in building web mapping applications. Examples of open source web mapping APIs include OpenLayers (JavaScript), OpenScales (ActionScript 3 and Flex), Leaflet (JavaScript), MapQuery (JavaScript) and MapFish (Python). The choice of which API to use really depends on the user requirements. OpenLayers is one commonly used API in building web mapping applications. OpenLayers is a JavaScript library which implements the JavaScript API. It is used in displaying maps in web browsers with no server-side dependencies. OpenLayers generally is similar to Google Maps for example except that it is open source. OpenLayers has been used by Sarup and Shukla 2012 and Jungialern et al 2012. It has also been used by GIS companies like Farallon Geographics and NuMaps. Farallon Geographics is a GIS company in US with focus on web developments aspects. They have used OpenLayers to build Web GIS application for their clients including: Colliers International Bank Site Data Management Application, Bay Area Air Quality Management District - Geospatial Components Integration With Mobile Field App, Union Sanitary District Plant GIS, GIS Web Application to Identify and Manage Critical Geological Hazard Data and MarineMap web application for identifying marine life protection areas.

4.2.4 Selected tools and software

Following the review of the available common FOSS that are typical to most Web GIS implementation, the selection of tools and software for this study was done. Table 3 below lists the selected tools, software, mapping client libraries and web development tools for this study.

Table 3 Description of the selected tools and software.

Software

License

Brief Description

Phase 3: Spatial Database Implementation and Testing

ArcGIS Desktop 10.1(Data Preparation and Geodatabase creation)

Proprietary

One of ESRI's primary software product used by GIS professionals to put together, use and manage geographic information.

ArcGIS Diagrammer 10 (Database Schema)

Freeware

ArcGIS Diagrammer is a productivity tool or a freeware for creating, editing or analyzing geodatabase schema. It is a visual editor for ESRI’s Xml Workspace Document which are created by ArcCatalog.

FWTools 2.4.7 - OGR (importing .mdb into PostgreSQL)

FOSS

Developed by Frank Warmerdam, FWTools is a set of binaries for Windows (win32) and Linux (x86 32bit) systems. FWTools include OpenEV, GDAL( OGR), MapServer, PROJ.4 and OGDI. OGR is a C++ library and command line tool for Simple Features providing read access to a variety of vector file formats including ESRI Shapefiles, S-57, SDTS, PostGIS, Oracle Spatial, and MapInfo mid/mif and TAB formats.

PostgreSQL 9.2/PostGIS 2.0 (Realizing 'Spatial' Database)

FOSS -  PostgreSQL License, a liberal Open Source license

Developed by the PostgreSQL Global Development Group and supervised by Red Hat and Enterprise DB, PostgreSQL is the most capable and reliable open source database. PostgreSQL is an object-relational database management system.

QGIS 1.8 and uDIG 1.3.2 (Testing)

uDIG - GNU LGPL

QGIS - GNU GPL

uDig is a desktop application framework built with Eclipse Rich Client technology. It can be used as a stand-alone application as well as a plug-in in an existing Rich Client application.

Development started by Gary Sherman in 2002, QGIS is an open source desktop GIS application that provides data viewing, editing and analysis capabilities.

Phase 4.1 Map Authoring, Publishing and Testing

GeoServer 2.2 (Map Authoring and Publishing)

FOSS - GNU General Public License or GPL

Developed by a diverse group of individual all over the world, GeoServer is a software server written in Java that is used to share and edit geospatial data.

Apache Tomcat 6.0 (Java servlet container housing GeoServer)

FOSS - Apache License Version 2

Developed in an open and participatory environment with the logo as a trademark of the Apache Software Foundation, Apache Tomcat is a servlet container that works as a Java applet on a server. It uses the Java Servlet and Java Server Pages (JSP) to provide a Java Web server environment that allows Java code to be run as efficiently as possible.

Atlas Styler 1.9 (Styling)

FOSS - GNU Lesser General Public License (LGPL).

AtlasStyler is a stand-alone application for styling geospatial data. AtlasStyler allows the styles created to be saved as XML files that follow the OGC SLD/SE standard 1.0. Supported formats include ESRI, Shapefiles, OGC WFS and PostGIS data sources.

QGIS 1.8 and uDIG 1.3.2 (Testing of WMS & some Styling)

Phase 4.2 Client Development

OpenLayers 2.12

Free, Open Source - FreeBSD (Berkeley Software Foundation)

OpenLayers is a JavaScript library for building rich web mapping client applications.

Extjs 3.4.0

Free, Open Source - GPLv3

Initially developed as an add-on library extension by Jack Slocum, Ext JS is a pure JavaScript framework designed for building interactive web applications utilizing techniques such as Ajax, DHTML and DOM scripting.

GeoExt 1.1

Free, Open Source - BSD

GeoExt combines the strengths of OpenLayers and Ext JS to build powerful desktop style Web GIS applications with JavaScript.

Heron 0.7

Free, Open Source

The Heron Mapping Client (MC) enhances the possibility of developing browser-based web mapping applications with GeoExt. Heron provides ammunition to GeoExt by offering high-level components and a convention to quickly assemble applications through configuration.

Google Chrome 25.0 Developer Preview version

Freeware

Free web browser by Google

Firebug Lite 1.4.0

Freeware

Web development tool for live editing, debugging and monitoring of CSS, HTML and JavaScript in any web page.

Chrome Developer Tool

Freeware

Also, a web development tool that allows web developers and programmers to dig into the internals of the browser and their web applications

JSLint Edition 2012-12-17

Freeware

A JavaScript code quality tool, JSLint is a syntax checker and validator. JSLint scans a JavaScript source for styling conventions as well as structural problems. Whenever it finds a problem, it returns a message describing the problem and an approximate location within the source.

JSBeautifier

Freeware

A tool for Beautifying, unpacking and providing clarity to JavaScript and HTML, it also makes JSON/JSONP more readable.