Examples from our department include:
Co-operative Web based Project Planner
Requirments management in XML
An XML database.
etc..

I suppose some of these could be useful to a RISC OS market. XML perhaps.

Also final year projects are not known for the stability! I would not want my final year project to be released as a product! :-)

So what should they write? Ideas anyone?

I just remembered that I recently asked for a database module that developers could use freely or for a small charge to build their software upon. Maybe that's something that could be written but I do agree on the need for a movie player that will play some "standard" format.

On my computer science degree at Leeds, and probably many others around the country, we have to do a final year project.

This generally consists of finding a problem, and solving it.

A problem with the RISC OS market is that there is no software for it. A solution could be found by writing some.

Be VERY careful here. Looking at that web page then what you're proposing probably wouldn't be sufficient for a final year project.

Some points which struck me

Although software engineering projects often lead to the implementation of a system, the final grade is based solely upon the report.

This suggests you don't actually need an implementation at all. This seems a little odd to me but certainly at Lancaster the message always was that the report is far more important than the code (although at Lancaster you MUST have a "working" solution)

The report must make it clear why anything needed to be implemented at all,

Simply saying "It doesn't exist on RISC OS" is not likely to be a good enough reason. The description suggests that any solution would be bespoke, and therefore hardware and software would be evaluated as appropriate. If a Windows based solution exists then writing a version for RISC OS would be considered a waste of time and money unless it demonstrated significant advantages.

Understand the problem
In order to demonstrate that there is a problem to solve, it is necessary to consider whether or not a solution already exists, and if so, what is wrong with it?

This just backs up my previous point.

I have to start thinking about what I could do for my project - do people think writing RISC OS software would be a good idea?

As someone who has been there, done that, brought the t-shirt I'd say that basing your project around RISC OS, unless you wish to work for a company such as Pace, could prove very damaging no matter how good it is. A good employer will look at what you did for your dissertation and try to infer what skills you gained from it. If you write it on an "unknown" system I guarantee that, in 90% of cases, you will be overlooked in favour of someone who has done something similar on a more "commercially acceptable" platform.

I like to see new RISC OS software as much as anyone else but think very carefully before going down this route. Your dissertation could prove vital to your first step into a career and it's not worth sacrificing this just to boost the RISC OS market by one more application.

However let's just consider some of the ideas people have put forward.

Development tools. Yes RISC OS is leagues behind other platforms. OTOH RISC OS' current main market is in the STB arena. Outside of Pace there is little need for these tools so it would be hard to justify.

In contrast video playback IS important to the STB arena. Assuming future STBs will include XScales or ARM10s then software playback may be possible. Given that the majority of the work would be in justifying why video playback is important and actually learning the video algorthms then there is potential here - although you could fall foul of the "why don't you just port a player?" argument.

However neither of these really fall under the software engineering remit. Ideally you should be looking for a specific problem where the use of IT can help. Development tools and video playback don't really sit here.

I don't wish to put you off I'm just trying to make you aware of the wider picture. Anyway you may find your project comes back to help in unexpected ways. My dissertation looked at network communications (multicast in particular). It effectively paved the way for the last three years of my life (research, and then into a job) BUT it also allowed me to help VOTI with their networking problems - something I couldn't have done before.

BTW before anyone reads too deep into this, that is NOT a statement saying that VOTI have a working solution, merely that the project is ongoing.

This project will investigate the practicalities of maintaining access to digital materials for the long term using software emulation. This will use the CAMiLEON concept of developing emulators in a form which is itself easy to preserve and migrate to future unknown computing platforms. The project could focus primarily on the development and specification of this preservation language, or on the engineering of an emulator (for the Apple II or a Z80 based machine for example). A focus on language development would require knowlege of some of the concepts covered by the final year "Compiler Design" module. Further information on the CAMiLEON project's idea's on emulation can be found in "Emulation, Preservation and Abstraction"
at http://www.leeds.ac.uk/camileon

--------------------------------

CAMiLEON (Creative Archiving at Michigan and Leeds Emulating the Old on the New) is based at the University of Leeds and the University of Michigan. The project is investigating the use of emulation as a tool for enabling long term digital preservation. In particular it aims to develop practical strategies for preserving digital materials, and test these strategies with practical emulation work (involving 1980's software like the BBC Micro Domesday project) and performing user testing of emulation and
alternative strategies.

Infact today I got the chance to see a Domesday in action... Very impressed.

This generally consists of finding a problem, and solving it.

A problem with the RISC OS market is that there is no software for it. A solution could be found by writing some.

http://www.comp.leeds.ac.uk/fyproj/softwareeng.html

I have to start thinking about what I could do for my project - do people think writing RISC OS software would be a good idea?

Of course, the market would still need to exist by the time of submission a year later...

We could get, ooh, at least a couple of students writing stuff, and the software would have to be delivered - otherwise they'd fail their degree!

On my computer science degree at Leeds, and probably many others around the country, we have to do a final year project.

This generally consists of finding a problem, and solving it.

A problem with the RISC OS market is that there is no software for it. A solution could be found by writing some.

I doubt that this would qualify as a final year project

http://www.comp.leeds.ac.uk/fyproj/softwareeng.html

I have to start thinking about what I could do for my project - do people think writing RISC OS software would be a good idea?

For RISC OS it might be a good idea, depending on the project.
For you it could be a bad idea in terms of available development tools. You'd probably be required to use object oriented design and development and given the lack of decent object oriented development tools you might run into trouble. If this is not the case than it might be a good idea.
Maybe you could write a development tool for RISC OS? An up to date C++ compiler and a WIMP class library are seriously lacking for RISC OS

Of course, the market would still need to exist by the time of submission a year later...

It's hardly going to disappear in one year.

We could get, ooh, at least a couple of students writing stuff, and the software would have to be delivered - otherwise they'd fail their degree!

So what should they write? Ideas anyone?

Small world init like