A grab-bag of annoyance

I've been meaning to write more, or indeed at all, on this blog. In the interests of making this easier, I'll try to ease my way in with a bit of a gripe post. Always easier! So this will be deliberately more extreme than my actual views. With that said...

The concept of "definition"

Why does it seem that education academics find this so difficult? This morning's 5500 lecture featured a slide titled "Definition of School 1.0 & Web 2.0" with the text "School 1.0" and "Web 2.0" linked. Both of these links went to images, and the "definition" was derived by asking us "what sort of words do you think describe the pedagogies & teacher-student-knowledge relationships inspired by these". These are not definitions, damnit, and it seems that Tony doesn't really get it. Later in the lecture a student asked "I don't really get what you mean by 'technological determinism' and 'social determinism', could you explain?", and it didn't seem that Tony understood the question - he certainly didn't reply with a definition of what he meant by "technological determinism". From the context of the lecture, it seems to me that he could very well have said "technological determinism is the idea that technology is awesome, and so if you use it in some task it will make that task more awesome". However, I don't know if that's how Tony understands it, or even if he could make his understanding explicit at all. I find, myself, that being unable to explicitly state what I mean is indicative of my poor understanding.

This is by no means isolated to Tony. One of our readings was an excerpt from Professor Ewing's book, the first chapter of which was titled "Towards some Definitions". In this chapter, she surveys the wide range of definitions of "curriculum" in the literature - ranging from "the list of dot-points a teacher wants to cover in the year", to "the whole sum of experience adults would like a child to receive over the course of their life". No where does she suggest what she means by the word "curriculum", something which I feel would be useful in a book about the subject!

On the choice of labels

Dear academics: when choosing the label you'd like to use for your particular assessment technique, please be aware that calling it authentic assessment will make you look like an arrogant know-it-all. Thank you.

The Will to Macros

The excellent MapTool full of useful features. One of which is the ability to associate macros with tokens - particularly useful it 4E D&D, since the maximum number of different attacks a character can have is less than 10.

Sadly, the documentation is somewhat sparse. Let's remedy that, with a worked example: writing Graham Tom's basic attack macro.

So, at its simplest, a basic attack is d20 + Str modifier + 1/2 level + proficiency vs AC, with [W] + Str mod damage. Graham Tom wields a longsword (weapon die: d8) and has Str 16 (modifier +3). This gives us:

Attack [d20 + 3 + 1 + 3]
Damage [d8 + 3]

We can parametrise this; MapTool allows you to define the attributes of a token, so the parametrised attack looks like:

Attack [d20 + floor((Strength - 10)/2) + floor(Level/2) + 3]
[d8 + floor((Strength - 10)/2)]

Again, this is sub-optimal: a natural 20 on the to-hit is a critical, which does maximum damage. We can get this behaviour using MapTool's ability to assign variables, and the fact that eq(var, 20) = 1 iff var = 20 and 0 otherwise.

I'm not sure how to suppress output from these macros, and MapTool seems to evaluate only the first full expression in [ ]. Feel free to remedy these flaws in the macro, which ends up as:

Attack: Roll [v=d20] + modifiers = [v + floor((Strength-10)/2) + floor(Level/2) + 3]
Damage: [(eq(v,20)*8 + ne(v,20)*d8) + floor((Strength-10)/2)]

So, we can see the unmodified roll, the final result, and this correctly calculates the damage in the event of a critical.

Adventures in future upstream nightmares

I clearly need to move "Write a 'How to be a good upstream' Ubuntu wiki page" closer to the top of my TODO list.

This piece of wrongness seen in #ubuntu-motu:

(20.24.31| screennam)) folks, I am sent here 'cos I have a bit of software to release under a modified gpl

(20.24.53| screennam)) and I guess I've not done this before so I'll need some advice on making it publishable

(20.26.23| wgrant)) Isn't the GPL immutable?

(20.26.44| screennam)) yes, so what?

(20.27.31| wgrant)) Releasing something under a mutated version of an immutable license seems unwise.

(20.28.05| screennam)) not if I call it something different

(20.28.23| screennam)) If I call it 'custom licence' no-one will complain

Remember, kids: releasing your software under a modified GPL will earn you the eternal enmity of all right-thinking packagers. Smart people with legal degrees argued over the wording of the GPL. Is your reworking going to be as sound?

Copyright is the most annoying and time consuming parts of a lot of Debian packaging. Please, make it slightly easier for packagers to get your work into Debian and Ubuntu - use one of the wide variety of common licenses for your code!

Agricultre, Neptune, Commerce, Cyclops

My friend Bice came up to Sydney for the first time at the beginning of this week in order to burn off some of his annual leave. He's so lazy that he hadn't bothered organising any leave for the three years he's been with his company. Habits are easy to fall into and hard to break - I doubt I'd do much differently.

Speaking of which, I have a habit of accumulating TODO items and insufficient time management to do anything about it. Currently on my plate, in rough order of priority:

• Finish a paper on rational inversive geometry
• Upload a fixed specto package to Debian
• Add some better checks & configuration to the Ubuntu Xgl package
• Reverse-engineer LDVS-on-nVidia to make nouveau's XRandR 1.2 branch work on my lappy.
  
• Write a config system for specto/notifrenzy
• Write a xscreensaver hack theme editor for gnome-screensaver
• Package the Tao CIL OpenGL bindings
• Update the compiz CIL plugin loader to work with current compiz
• Do some work on Playtools
  
• Hack on Joybot
  

... the list goes on. I wish I were a chronomancer like Saint Germain.

My heart is elsewhere

Sam's gone off to a materials science conference, where she will be presenting some of her work. This means that she's not here. Somehow, a couple of years ago this wouldn't have mattered. Wow.

Before she left, I tried to update her Windows XP laptop. I'd forgotten how strange windows is. I needed to install three install programs before I could actually install any updates.

The importance of micro-optimisations

Sam and I bought some wardrobes and a desk from a post-doc who's moving off to England to take up a position as lecturer. This is good: no longer will our clothes horse have to do double duty as our entire clothes storage space. The bad is, of course, that we needed to get the furniture home ourselves.

Now, getting a removalist/furniture taxi would've cost about $140. Hiring a ute for the day, cost $69. Plus $16 for insurance, which seemed prudent since I've never driven in Sydney, or driven a ute, and the last time I drove was over a year ago. Plus 1.5% stamp duty, for some reason. Plus petrol.

So all up it cost around $100, an important, necessary, saving of $40, and now we have to lug some heavy furniture up a couple of flights of stairs.

I commit this note to self to the boundless memory of the intertron: Next time just pay someone to do it!

So, on Friday Sam and I went to see the new Transformers movie with SpockSoc. And it was good. Who'd have thought that battles between huge, city destroying robots could be so cinematic?

I'm learning dvorak, now that I've got a lappy that I can move the keys around on. Except, curiously, for the 'b' key, which is different to every other key on the keyboard. I'm now at that awkward phase where my fingers kind of know where they're meant to go, just as long as I don't think too hard about it :)

Things I've recently learnt:

• Inversive geometry is really about the pole-polar relationship in a projective space.
  
• The time taken to pack up a flat and move is dwarfed by the time taken to unpack.
  
• My laptop bag is sufficiently waterproof to not kill my laptop in the rain. Yay!


• It's not too hard to to Test Driven Development in C with the check package


• In related news, I'm a much less proficient C programmer than I was five years ago.


• Two-fingered tapping on my touchpad generates a Mouse2 click, and 3 fingered tapping generates a Mouse3 :).