Starting your project

Written by: Jonathan Briggs

October 14, 2005 [5262 views]

Here are some general comments about things you should do to make sure your project starts in the right way.

Make sure that you know what you are investigating

By now you should have filled in a form and had it signed setting out what general area your project will be in. It is almost certainly too broad and probably too simple at this stage (both at the same time).
I am regularly asked to supervise projects “to build an ecommerce site for X”. This is rarely a good project because the client underestimates the work involved and the student ends up designing a few screens and implementing them in ASP or similar. Don’t do this!

Instead choose a piece of ecommerce that is really interesting such as improving the customer experience, helping the customer choose, content/service syndication, price comparison etc and focus your project around this. I want you to get deeply involved with something rather than skin across too wide an area.

Understand the difference between an ok project and a good one

Make a list of deliverables and outcomes that you would need to achieve to be awarded a pass and agree this with me as soon as possible. Make a second list of the additional work that you think you would need to achieve to get a really good grade. The extra work is never “more of the same” but will involve tackling some harder problems, demonstrating greater insight or managing the complex politics of your organisation or client.

Brainstorm the area in which your project sits

Draw a rich picture with your project at the centre and mark on all the technologies, stakeholders, information sources etc that will influence it. Read around some of the technologies and see what other people are doing. Have you included some of the hot trends in ecommerce such as web services, accessibility, Ajax, open source, fraud prevention, scalability, content management etc?

Come and discuss the picture. We need to work out what you need to research, design, analyse, build and evaluate as soon as possible.

Set yourself implementation goals

Make sure that you have decided what you will prototype or implement. This is a key part of any computing, information systems or multimedia project. Learn new skills to make it possible.

Get hardware and software set up

Every project must have at least a prototype and while some of you have already shied away from programming your project is a good opportunity to teach yourself some.

You need to design an environment in which to try out ideas. I would recommend one that uses Open Source but Microsoft is good too.

If you have Internet access that is great.

I would recommend everyone having access to a Linux server running LAMP (Apache, MySQL, PHP) or an Apple Mac (of course). If you are definitely a Windows person then take a look at .NET. Most employers won’t me interested in someone who has only a tiny amount of ASP experience.

Do this now rather than 4 weeks before the deadline if you want to do better than a bare pass.

Imagine that your deadline has been reduced

Many students only start to get into their projects 4-6 weeks before the deadline. This is a highly dangerous strategy unless you want to risk failing and it also looks bad from a prospective employers point of view.

I am a believer in iterative development; build something simple, build something better and continue to improve. This is often described as Rapid Application Development and is linked to methodologies such as DSDM and Agile.

I recommend that you take this approach to every part of your project. This means “time boxing” activities such as building a first prototype and attempting to develop your ideas well before they are needed.

Many students listen to this advice and then ignore it. Don’t! Imagine that the project was only 4 weeks long and attempt a prototype of all of it in the first 4 weeks: analysis, research, design, prototype and documentation. By doing this you will start to understand your project and the second “re-factored” version will be much better.

Buy a few books

As part of building your prototype you need to equip yourself with additional programming or development skills. Buy a good book to give yourself a head start and work through it. Come and discuss relevant titles for your project.

Make notes about your research and software sources

The best (and surest) way of failing your project is to cheat and copy someone else’s. Unfortunately Kingston University is expert at detecting this and tends to throw people off courses without degrees. Waste of time!

A few people who don’t copy whole projects still copy parts of their programmes or their documentation. They still get caught and the consequences are the same.

It IS ok to base your work on Open Source code or to make reference to other people’s written work but it MUST be properly referenced. Every member of staff will tell you this but every year some people fail to do this properly.

Start as you mean to go on. Keep a note of every reference and every chunk of code that you have downloaded. Make sure that you discuss with me what is permitted and how it should be shown in your dissertation.

Start a blog!

This site is a blog and it helps me document my work with my students and my projects. Why not start your own blog to document your project? Its much more fun than keeping a notebook, it allows other people to comment and it will help you get a job if you can point at your blog and tell someone “this is how my project and my thinking developed”.

You can start a blog for free at http://www.blogger.com

Use your supervisor

If I am your supervisor then contact me regularly, ask lots of questions and take notice of the suggestions above. Remember that it is not my job to chase you.

If I am not feel free to ask questions here by adding comments. I am happy to help even if you are not on my course or even at Kingston.

Recent comments:

What do you think?

Add your comments