Well, I can spot some confusion
Computing == Computing Science
ICT /= Computing Science
The former is about theoretical aspects of computations, algorithms it's like maths.
The later is about practical(and rather shallow) use of computers and is more like sociology.
And now back to OU, I've done http://www.bcs.org/content/ConWebDoc/12057 BCS IT Cert which is equivalent to year one of university, and as it happen I had to retake one paper :D - I've done it as a benchmark(only the exams) but the OU is nowhere near this standard (maybe some T-modules would give you some preparation) but none of M- ones... no wonder that Computing will be gone... and only ICT would stay.
Must admit I feel a sinking feeling when I occasionally look at Higher Physics or Higher maths papers, all sorts of stuff which would have been deemed essential for my A level (many many years ago) is just not there. Also the level of problem solving skills just doesn't seem to be developed even in the advanced Higher syllabus.
I think the problem came when the exam boards stopped being run by universities. I heard this morning that the exam boards this year set some impossible questions for the A/S levels you wonder if this was a deliberate ploy to inflate the grades.
I think that the problem is that there's a lot of 'computers are important therefore we have to teach computing'. I think that this is wrong.
When I was at school we weren't taught computing. That was the era of the Sinclair Spectrum when British programmers virtually invented computer-gaming [ok a wee bit sweeping ].
I think that you've both identified parts of the problem.
If we're going to teach computing then we have to do it properly. Which means that we'd have to greatly restrict the syllabus and tackle things in depth
Because we're dancing to a political agenda we don't challenge our children properly, we give them masses of data and reward those who can remember it best
An example of this was one question in the exam about calculating the size of the system bus. I mean what!? What possible use is that to a child!
And now they've taken it a stage further and started asking impossible questions.
Teaching techniques not tools would be sufficient. As to write an algorithm solving a given problem it's not sufficient to know syntax of any particular language. Just have a look at http://en.wikipedia.org/wiki/The_Art_of_Computer_Programming
And say how ready to read it you are :D
Knuth is a genius. Much of spur for my OU stuff is down to wanting to understand things like that.
May I ask you a question? Chris and I are kinda locked in to the last chance saloon—we attend the same tutorials and we've both signed up for Topology and Groups and Geometry next year. I seem to recall you saying, somewhere, that Number Theory was the best programming course ever. True? Even if not, is is a good course?
For me M263 just whetted my appetite, is Number Theory a follow on?
Well you are obviously looking at it from a computer science perspective and I've no doubt Haskell is rigorous. I'm more interested in logic for it's own sake especially the implications of Godel's theorem for philosophy.
At work I've had endless debates with people urging me to learn the latest fad OOP, C++, C# etc and I suppose eventually Haskell will feature.
However most scientists/engineers really just want to solve the problem (usually some form of Numerical Integration or differential equations) in as quick and easy way as possible. Also dare I say it the underlying equations are more important than whatever language you choose to implement them.
For that purpose I find FORTRAN and MATLAB sufficient for my purposes. Indeed I would argue that FORTRAN still handles numerical calculations better.
Comments
New comment
Well, I can spot some confusionNew comment
Must admit I feel a sinking feeling when I occasionally look at Higher Physics or Higher maths papers, all sorts of stuff which would have been deemed essential for my A level (many many years ago) is just not there. Also the level of problem solving skills just doesn't seem to be developed even in the advanced Higher syllabus.
I think the problem came when the exam boards stopped being run by universities. I heard this morning that the exam boards this year set some impossible questions for the A/S levels you wonder if this was a deliberate ploy to inflate the grades.
New comment
Hi Marcin/Chris
I think that the problem is that there's a lot of 'computers are important therefore we have to teach computing'. I think that this is wrong.
When I was at school we weren't taught computing. That was the era of the Sinclair Spectrum when British programmers virtually invented computer-gaming [ok a wee bit sweeping
].
I think that you've both identified parts of the problem.
An example of this was one question in the exam about calculating the size of the system bus. I mean what!? What possible use is that to a child!
And now they've taken it a stage further and started asking impossible questions.
New comment
Teaching techniques not tools would be sufficient. As to write an algorithm solving a given problem it's not sufficient to know syntax of any particular language. Just have a look at http://en.wikipedia.org/wiki/The_Art_of_Computer_Programming And say how ready to read it you are :DNew comment
Marcin
Knuth is a genius. Much of spur for my OU stuff is down to wanting to understand things like that.
May I ask you a question? Chris and I are kinda locked in to the last chance saloon—we attend the same tutorials and we've both signed up for Topology and Groups and Geometry next year. I seem to recall you saying, somewhere, that Number Theory was the best programming course ever. True? Even if not, is is a good course?
For me M263 just whetted my appetite, is Number Theory a follow on?
arb
neil
New comment
Well you are obviously looking at it from a computer science perspective and I've no doubt Haskell is rigorous. I'm more interested in logic for it's own sake especially the implications of Godel's theorem for philosophy.
At work I've had endless debates with people urging me to learn the latest fad OOP, C++, C# etc and I suppose eventually Haskell will feature.
However most scientists/engineers really just want to solve the problem (usually some form of Numerical Integration or differential equations) in as quick and easy way as possible. Also dare I say it the underlying equations are more important than whatever language you choose to implement them.
For that purpose I find FORTRAN and MATLAB sufficient for my purposes. Indeed I would argue that FORTRAN still handles numerical calculations better.
New comment
I can agree, I've only managed to write some Monte Carlo approximation, but if I had a mainframe haskell could use its CPU... Must see for concurrency: http://blip.tv/file/get/OSCON-OSCON2007SimonPeytonJones914.mov slides: http://research.microsoft.com/~simonpj/papers/stm/STM-OSCON.pdf or: http://media.libsyn.com/media/seradio/seradio-episode108-simonPeytonJonesOnFunctionalProgramming.mp3 --- some haskell related stuff: Maths: http://haskell.org/haskellwiki/Books_and_tutorials/Mathematics to see some research: http://research.microsoft.com/en-us/people/simonpj/ fun comparison to fortran: http://bitcheese.net/wiki/haskell-vs-fortran speed comparison with fortran: http://shootout.alioth.debian.org/u64q/benchmark.php?test=all&lang=ghc&lang2=ifc