Software Engineering Project Ideas

[NEWSBOT3]

(Should this be on GD ? im not sure)

As some of you may know, im in my final year of BEng Software Engineering, but i need a project idea to do, and im not exactly full of them.

Basically, i can do anything. Anything, relating to software engineering that is.

Last years list (which is supposedly inspirational, but im not finding any) is here

Having read it about five times now, i can safely say its got nothing i like in it :\\

A rough idea of what i've done in the last few years, which i figure is a good basis :

• Java
• Socket Programming
• OO Programming
• Software Design
• (will be doing) Image Processing
• (will be doing) Computer Games
• (will be doing) Computer Networks
• (will be doing) Advanced OO Programming
• (my free time) PHP
• SQL

Area's that i think experience in would be useful for me :

• Linux/Unix (i have a basic knowledge, but ive never really done much with either)
• Windows C++ (i can do console based, just windows C++ i've never managed to do)
• Oracle DB Admining (have used it, just never admined it)
• ASP (i did a little bit one year)

So i was hoping to use something combining any of that lot, i'm just not sure what :/

I have one vague idea, which was email server software. I think the current pop3/smtp system is far to easy to abuse (see, sending fake mails, various formats of emails in use which make parsing mail a sod), and is lacking a few features i'd like.

I've also written email clients in the past, so i have some knowledge in the area already, im just not certain if there is enough 'substance' in this idea for it to be useable :/

Your input is most appreciated

[Supernova9]

Am I the only one thinking this project:

An independent music Web site to discover different kinds of music and bands.

is rather simplistic a project for the concluding project of a 3 year degree course?

[edit] Never mind, the same guy's resit project is lower down the page

[Structural Integrity]

Whatever you choose, don't aim too high!! Simple as that..

But for as an idea:

- Chat application or Instant messaging thingy
You use sockets
You can do it in C++ and with Windows programming
You can use a database connection for logging
You can make it OO

We did the simple chat thing last year for a project in Java. It worked OK and wasn't that difficult at all.

[NEWSBOT3]

Originally posted by Structural Integrity

Yeah, im liking sockets work. I did an email client in java for a module, and i've done irc clients in it before, which were both quite easy.

Windows C++ would be a considerable challenge, since i've never quite managed to get my head round it, and im not sure if i want to force myself to do that and my project as well.

Having had to explain it a bit (we can stick ideas etc on the uni project site thing) , im liking my email idea more, since it'd involve a complete redesign of the protocol, then implementing a server for it. The scope of that seems to fit quite well with my degree - taking a system, analysing it, find the bad points, plan a new system, design it, implement it.

I think i'm gonna play with ideas till monday or so, then start chasing folks to get it sorted.

[Structural Integrity]

Since this is a Sofware Engineering project I don't think the emphasis will be put on the type of project and the actual implementation, or how inventive you were.
Our software engineering puts the emphasis on project planning and tracking, design stages, design models and cooperating with others. The actual result is not that important, how you got the result is more important.

I'd ask first where the emphasis lies and what they expect to receive from you in the end.
If they want pretty documents from you, and care little about the implementation (if you even finish it), then I'd put in a few more technologies that you are NOT familiar with. If you can't get it together with the new technologies then it won't be such a disaster.
If they really want a properly working product in the end, stick to things you know. The engineering part consumes quite a bit of time and you don't want to take any risk with new technologies then.

[NEWSBOT3]

from
the uni site

On completion, you should be able to demonstrate these learning outcomes - which will be assessed by:
an ability to use a systems approach to develop an effective software solution to meet an agreed specification - the development of a functional software product to a defined specification

a demonstration of methodical rigour in choosing and applying relevant methods, tools and techniques - an evaluation of a range of possible methods and/or techniques and the selection and application of such methods and techniques as may be suitable for a given application

mathematical and modelling ability in formulating software specifications and solutions - the production of a specification for a software solution or product which demonstrates formality and rigour

an ability to demonstrate relevant software project management skills - a project plan which shows an ability to manage a software project effectively with regards to quality, functionality and resources

which i think means its focused very heavily on the design side of things. Pretty documents for the most part then, with little regard for the implementation.

but then i read this bit.

The emphasis in a Software Engineering project will be on the production of a systems or applications software product using a rigorous approach and a recognised development methodology. An analysis (e.g. of user requirements) and a formal design is expected, with a sophisticated implementation evaluated against a balance of technical and functional criteria

what ?
:/

That seems to indicate its 50-50 either way :/

[Structural Integrity]

Don't rely on indications... ask your teacher!
At my school they usually link 20% of the grade to the actual implementation, 50% to the design and effort and 30% to the essay and presentation.

You could ask your teacher what their norm (correct word?) is.

[Raging.Retard]

My final year was balanced as so (BSc Soft Eng):

40% Project Writeup (Design etc)
40% Actual Implementation
9% Project Demonstation
7% Project Seminar
4% Initial Plan

But as has already been stated, yours probably wont be the same, so go find out what it is.

[NEWSBOT3]

Originally posted by Structural Integrity
Don't rely on indications... ask your teacher!
At my school they usually link 20% of the grade to the actual implementation, 50% to the design and effort and 30% to the essay and presentation.

You could ask your teacher what their norm (correct word?) is.

yeah, but i cant really do this until AFTER i have chosen a project :\

[djbass]

Funny enough but I was thinking about exactly this the other day and think it would be a great idea.

One feature I would like to see implemented in a revised SMTP protocol is to make it impossible to feign the return address so that the originater of the message can always be contactable (therfore accountable for the message). Even at its lowest level having the recieving server verify the return address is valid even if it isn't the genuine return address.

=[DJ Bass]=

[Raging.Retard]

Originally posted by djbass
One feature I would like to see implemented in a revised SMTP protocol is to make it impossible to feign the return address so that the originater of the message can always be contactable (therfore accountable for the message). Even at its lowest level having the recieving server verify the return address is valid even if it isn't the genuine return address.

AMTP

[NEWSBOT3]

Originally posted by djbass
Funny enough but I was thinking about exactly this the other day and think it would be a great idea.

One feature I would like to see implemented in a revised SMTP protocol is to make it impossible to feign the return address so that the originater of the message can always be contactable (therfore accountable for the message). Even at its lowest level having the recieving server verify the return address is valid even if it isn't the genuine return address.

=[DJ Bass]=

Well, i was thinking less of a single protocol, and more of a mail server basis. From talking to a guy i know doing a project on cheating in online games, i've pretty much concluded that nothing the client sends can be trusted (which, makes sending mail fun), so a revised protocol wouldnt really be my answer.

I *think* i'd be looking more at a revised mailserver (and/or new mail protocols) but im not certain on that.

RR - interesting link, ty

[wakey]

I just finsihed my degree and while I wasnt quite doing the same course (BSc Computing) most of our class did take the software route for the final project. Of all the ones I saw by far the most impressive was a "3D model of the campas". Basically it involved using direct3d to model the campas which could be navigated around to allow new students to learn the layout. Also had differnt moses such as Challenge mode where you have to get to a certain room in a given timescale, timetable where you entered your timetable and had to get around the campus as if you were going through your usual day or free roaming.

It really was impressive.

As for your email idea one of the projects we looked at from the year before was on a simerlar idea but tackling simerlar issues with newsgroups rather than email. You can certainly produce a very good project on such an area

[NEWSBOT3]

hmm, thanks wakey

ok, 2 current ideas i have , sumarised :

Title: A Web Based Project Management System
Objectives: To design and implement a web-based solution for team projects. The software will enable teams with web browsers to effectively communicate between each other, as well as providing a database of contact details, and a system for managing various areas of projects that people with the teams will be working on.

Title: A Design and Implementation of Improved Email Protocols
Objectives: To examine the current email implementation (pop3/smtp/[imap?]) and to design and implement a system with significant improvements in both security, and ease of tracking sent and received emails.
The server will not trust data such as 'from' addresses sent by the client, and will instead have the from address for an email account stored on the server.

[Raging.Retard]

Originally posted by NEWSBOT3
...The server will not trust data such as 'from' addresses sent by the client, and will instead have the from address for an email account stored on the server.

I dont see how that is feasible. If that was the case you could only ever send emails from the server you were registered with (which is bad), and whats to stop someone making a server that pretends the address is valid anyway and then forwards it to its destination... or even skipping the whole process completely, and pretending to be a server forwarding to the destination.

Granted I know you havent even decided to do a mail solution yet, but I felt like asking anyway.

[(°>]

Yeah ... I can't get my head around the email idea.

It sounds to me like hotmail users emailing each other and no one else \o/.

Tho would be nice if you come up with a solution.

[NEWSBOT3]

Originally posted by Raging.Retard
I dont see how that is feasible. If that was the case you could only ever send emails from the server you were registered with (which is bad),

I'm pretty sure thats the case now. Why should you ever want to send emails from servers you ARENT registered with ?

whats to stop someone making a server that pretends the address is valid anyway and then forwards it to its destination... or even skipping the whole process completely, and pretending to be a server forwarding to the destination.

nothing at all, however since i'd be considering this in both sending AND receiving then i'd also be verifying delivered email too.

[NEWSBOT3]

Originally posted by (°>
Yeah ... I can't get my head around the email idea.

It sounds to me like hotmail users emailing each other and no one else \o/.

Tho would be nice if you come up with a solution.

well, yes and no.

the idea is that the server does the sending, not the client.

ofc there is nothing to stop the client from pretending to be a server and sending mail if it wishes.