UML Anyone else finding it's more fun than expected?

I'm busy with assignment 2, so this will probably only be of interest to those who're also obsessive about removing assignment pressure as fast as possible, so as to be able to then get on with the business of actually getting to understand things (nice and slow and peaceful like).

My assessment of the theoretical side. Well there's not a great deal to it, is there? You can read the book in about a week, and have a fair idea of what it's all about. (That's not to say you can finish studying it in a week - but then try to read a Calculus textbook in a week and see how far you get ...). There just isn't a great deal of theoretical meat ... but Ah! Watch Out! The practical side can be quite a bugger.

What do you think? (Until assignments are past, the only applications I could discuss are assignment-related - which is fine as long as none of us crosses that line between collusion/ plagiarism whaddever. Feels a bit like cheating asking for assignment related discussion, so I don't ask. Telling's OK, but asking's not).

Really what I'm blindly groping toward is some way we could work out to make this forum something that helps with our studies. So far it doesn't look like a particularly fruitful academic resource for anyone (apart from COS407 - which I went and looked at on a tip --- but there are big kids in there, and they speak a weird language. I got scared, started crying, and ran away as fast as I could).

Iv done the module last year so if you need any help and if i can help id be glad to assist. Dont be too complacent about the fact that theres not much theory. I suggest you know the different diagrams well and how to perform the steps in analysis and design to show it on the diagrams. Not as easy as it looks. They could also ask alot of theory about the diagrams like they did in the supp. The module is alot of fun but dont think its way too easy either. Study the two running case studies. Study the different relationships like association, aggregation...Its not alot of theory but you should practice the diagrams.Oh and know your assignments. Good luck.

I also enjoyed this module when i did it 2 yrs ago.I have to agree with Amy it looks like a lot of theory (i thought so too in the beginning ) but as i progressed some concepts i just could understand by reading ..even if i read it a couple of times ... i had to draw diagrams and apply it in my code as well ( c++ , inheritance, aggregation, etc ) .... then only the light started to go on ... well it's still not burning brightly enough( for me that is ! ) ... have fun with this module and enjoy !!

Thanks muchly. Ja, it does look like the way to the theory is through the practice here. Glad to know it does come in useful. Can sort-of see that, now a bit, actually. The Java course has some UML in the first section (as you probably both know), and after the phase of thinking "Oh hell, so this is what life's going to be like" at the start of modelling the behaviour of an ATM, after some pain a picture starts to develop, and it starts to feel like one understands the machine. There that goes to that state; that should go there; should I move that round like that? It's nice to have picture in front of you.

As you say, Amy, it's not as easy as it looks, once you start actually scratching head about what relationship this one is, whether that thing is an object or an attribute, etc. Yes, I must get to the running case studies. At the moment I'm trying to just clear away assignment pressure by means fair or foul, but when the air clears the thing to do is to run through the case studies, and discuss those when things get foggy. (Or when things suddenly appear luminous).

ai ai, I hate UML with a passion. Let me begin by saying that UML is by no means easy/simple. The text book for INF207 only just begins to touch the tip of the iceberg wrt UML. For example, for assignment 2, in whatever the relevant chapter was, the guide to make class diagrams did not even mention operations . So at that point a class diagram only has attributes . I was unsure whether or not I was supposed to add operations in my answer, I decided not to since the text book mentioned nothing of the sort.
When I was studying bsc computer engineering, we had various software engineering courses, which had a textbook the same size (maybe a little thicker) which is composed entirely of UML. The confusion got compounded recursively with each new chapter. Good 'ol Satzinger UML is just beginners stuff.

The idea behind UML is that code can be generated from a decent enough diagram. I believe MS Visio has this function, generating code in VB or VB .net I think. In order to produce a sound piece of software, the diagrams would have to be extremely detailed. Given you would need to review the generated code and tweak it a bit.