M359 Relational Database: Theory and Practice

In 2007 I studied M359 Relational Database: Theory and Practice. 2007 was the first presentation of the course and I generally try not to do 1st presentations, as they tend, from my experience, to be a bit buggy.

M359 was no exception, the main gripe being that in TMA3 we were unable to use SQLAnywhere to answer one of the questions (as originally specified) and had to treat the question as a ‘paper exercise’ due to a bug in the software! Rather strange this wasn’t spotted when the question was proofed?

Another thing I found about the course was some over complicated explanations, I could work out normal forms before doing this course, but they way they teach it confused the hell out of me. I ended up explaining normalisation in 'plain English' at a tutorial, after the tutors expiation, using ‘course speak’ left everyone scratching their heads!

As for the actual course, I found the bulk of the material OK, if a little heavy going, having some basic knowledge of databases & the SQL language did help considerably.

The course is delivered in 5 blocks

1 Databases in context

A rather easy start, looking a data and the history of data storage, before going into Conceptual Data Models and the Entity Relationship Model in particular

2 Introducing Relational Theory

Looking at relational representations, manipulating them using select, project, join operators etc, constraints & normal forms.

3 The database language SQL

Speaks for itself

4 Database life cycle

I quite enjoyed this, which involved creating ERMs, relation representations and then the actual database from some textual descriptions & sample documentation. This was rounded off with a discussion on denormalisation, data warehousing (which was rather too brief for my liking, especially in view of the TMA question we were asked) and a quick look at distributed data management. However some of the text was rather heavy going, I sure there are simpler ways of explaining this stuff.

5 Java & XML

I found this part of the course the most frustrating. These bits were ‘tacked’ on to the end of the course. The java aspect was a very brief look at java ‘because java is the most likely language in which you will find SQL’. I’m never likely to ever want to get involved in Java and I think the overview was too brief to be of any benefit.

Likewise the XML part was all to brief to be of any practical use.

Block 5 makes up 1 TMA question and 1 Exam short question (XML being either the exam question and Java the TMA question or vice versa)

The TMAs

In previous courses I've done, their was generally one TMA per block. This gives a good sense that you are moving on and that you have put the block to bed. On this course however the TMAs were spread over blocks, ie TMA1 covered Block 1 and part of Block 2, TMA4 covered part of Block 4 and part of Block 5. This did leave me with the impression that the course was never ending...

The Exam

The Sample Exam Paper also had Errata, and in my opinion was not particularly helpful when compared to the real exam – especially putting the data needed for the SQL questions in an appendix printed at the end of the exam paper (the SEP had these next to the actual questions!) – cue lots of paper rustling mid way thru the exam! Also Q15 was a rather flowery essay type question in the exam, whereas in the SEP it was about drawing schema diagrams.

The exam itself is in 2 parts - 12 short answer question each worth 5 points. These are fairly basic stuff - one question where we had to 'fill in the blanks', 3 questions on SQL, 1 on normalisation, 1 on data warehouses etc.

Then in part 2 you need to answer 2 of 3 questions, one question being based on each of Blocks 2, 3 & 4 with some bits from the other blocks thrown in to the mix to round things off. Each of the long questions being worth 20 marks.

In conclusion

I enjoyed the course, mostly. Bit annoyed that my revision was a bit skewed due to the SEP, going forward students taking the exam in future years will be helped by being able to procure the past papers from OUSA. Hopefully the course text has now been fully debugged and their will be fewer errata in future presentations.

A number of students thought that this course should be 60 rather than 30 points, but I thought it was about right. It was also the first course I've studied where other students seemed to be always complaining about certain aspects of the course, the documentation, how things were explained etc. Hopefully due to the volume of complaints from students, future presentations will be better!

Final thought; never do an OU course on its first presentation.

Final Errata

The final errata on my presentation of this course was only spotted on 13th December 2007, when the helpful notice on my student homepage was changed from 'Your course result should be available by Friday 14 December 2007' to Your course result should be available by 'Friday 21 December 2007'.

Ho Hum.