Insert author details here, using the following paragraph as an
example.
Ms Emily Feher, (acting) Manager - development and integration, Flexible Learning and Teaching Program, emily.feher@its.monash.edu.au
The demand for flexibly delivered services has exploded, with students and staff expecting to be able to carry out tasks via the web, the phone, SMS and other technologies. Unfortunately, the resourcing of such services has not grown at the same level.
In order to continue to be a useful, robust and economically viable service to the Monash community, the my.monash portal’s development has followed a path of continual engagement with the stakeholders and a process of iterative delivery, “growing” services organically in line with demand.
In the majority of cases, this approach has worked well – with usage of my.monash growing consistently and with an increasing number of stakeholders seeking us out to place services on my.monash.
The ‘creative chaos’ of the approach is great for delivering useful services on time, however, occasionally robustness issues are encountered in the production environment as a result.
This poster reviews how techniques from eXtreme Programming (XP) help us to address the challenge between rapid development and robust quality.
Monash University is Australia’s largest learning institution comprising approximately 45 000 students and 10 000 staff at 7 locations in five countries.
The my.monash portal [HREF1] is an enterprise portal, enabling secure and personalised access to useful and relevant resources for all members of the Monash University community.
The overall goal of my.monash is to provide a ‘one-stop’ shop for all Monash students and staff to enable them to interact with the University, regardless of location.
Starting as a prototype in 1998, the my.monash portal is now (May 2003) a critical production service seeing on average 12 000 unique staff and student users per day. The usage profile applied across 25 services delivered through my.monash makes it imperative that the portal is sufficiently robust and highly available.
Because of Monash’s global scope, high availability cannot be constrained to “normal” business hours – Monash business hours include London, South Africa, Prato and Malaysia. With such a broad scope of services and time, there is no convenient maintenance window in which to execute changes.
These demands require an extremely agile development cycle that can deliver incremental and continual change that incorporates usability principles and business needs.
These changes must also be executed within an environment of constricting resources through the higher education sector, where Universities are being asked to do more with the same (if not reduced) resources.
The Flexible Learning and Teaching (FLT) Program within the Information Technology Services Division at Monash University is following a development approach that, to a large extent, works within this constrained environment.
The success of my.monash to date has resulted from two main factors:
As outlined in Figure One (below), the development of the my.monash portal is reliant upon constant feedback loops with stakeholders of projects and cohorts of end users.
Sourced from Kennedy, D. M., Webster, L., Benson, R., James, D., and Bailey, N. (2002). [HREF2]
Large scale projects require significant buy-in from senior management which can be difficult to obtain without clear business cases. Often it is difficult to effectively communicate the business need for an emerging technology without demonstrating its usefulness in the current environment.
Our model focuses on the principle of building on small successes. Rather than endeavouring to deliver complex projects over long time frames, we work with leading edge adopters to deliver prototype services that meet their specific needs. These become the “case studies” that provide the basis to grow the prototypes into full services.
Input from stakeholders and users from the very inception of each development project goes a long way towards ensuring that the resultant product is one that will solve the business problem and be usable by the eventual end user.
We have also found that as the stakeholder learns and experiences what technology can deliver, it often sparks a change in direction about the existing development, and/or initiates new ideas.
Rapid prototyping assists us here. Rapid prototyping is used to get a ‘working model’ of a concept to a point where the stakeholder can visualize (if not actually click through) what the development will enable users to do.
Advantages of rapid prototyping include the ability for the stakeholder to ‘see’ the development at a very early point in the process, enabling them (and the development team!) to painlessly make amendments to the specification and confers the ability to release a draft of the interface to a cohort of end users for the purposes of usability testing.
In order to respond to the challenge of changing requirements throughout the development cycle we employ an iterative delivery model that focuses on managing scope, rather than time or resources. This ensures that something is delivered in a reasonable timeframe, and something that is still useful and wanted by the stakeholders, compared to longer development cycles where the end product may be delivered, but stakeholders needs may have changed in that time.
An iterative delivery model is one in which a full project may (or may not) be scoped and specified, however its development and release is staged into small increments of the larger whole.
Consciously staging the development and delivery of projects allows for:
To illustrate this by way of an example, the my.monash portal hosts a My Library service. The business objective of the My Library project is for any student, regardless of location or degree program, to have access to all relevant library resources in a personalised manner and in a single location.
The first release of the My Library service included a listing of which books a patron has out on loan, which were overdue, which were accumulating penalties and the like; a comprehensive search tool to be able to find almost any online Monash Library resource; and a ‘sleeping’ announcement box through which the Library could communicate to all patrons about recent or upcoming events.
Future releases of the My Library service are planned to include additional services such as searchable discipline-specific databases and the ability to search and enroll in Library run courses.
In terms of enacting iterative delivery and development, in essence we identify those aspects of the service that will result in the ‘biggest bang for our buck’ and focus on those first, unless of course the stakeholders have other ideas :). Developments that will confer significant benefit to the end users, but require a degree of integration/ planning/ development are typically those given the longer development timeframe.
The biggest challenges we have experienced via this development and delivery model lies in capturing changes to the prototype and undertaking the subsequent refactoring required in order to be able to effectively run the development in production.
In order to respond to these issues, we have drawn on the techniques of eXtreme Programming (XP).
Extreme Programming (XP) is a ‘lightweight’ software development methodology [HREF3]. What distinguishes XP from more traditional ‘heavyweight’ SDLC methodologies is that the XP philosophy empowers programmers to “deliver the software your customer needs when it is needed” [HREF4] yet remain focused and productive.
XP has four key tenets: communication (with the stakeholders and end users), simplicity (in terms of design and code), feedback (from users at all points in the development process) and courage (empowering the team to respond quickly and appropriately to changing requirements). In addition the XP model confers a degree of process upon the often ‘creative chaos’ that is the FLT model!
In particular XP recommends refactoring code regularly to maintain the level of simplicity, and to unit test the new development as it is being developed to simulate its impact on the existing production environment.
The application of these process steps is helping to establish the critical balance between agile and flexible incremental change and robust, highly available service delivery. Our current development model is proving more response and less volatile as we draw on these principles.
Kennedy, D. M., Webster, L., Benson, R., James, D., and Bailey, N. (2002). My.monash: Supporting students and staff in teaching, learning and administration. Australian Journal of Educational Technology, 18(1), 24-39.[HREF2]