Running for Parliament? Better quit your job.

Anyone who works for the Government has to quit their job if they want to run for election. Bugger.

Commonwealth public servants are lucky when it comes to standing for Parliament. They’re generally covered by a policy to hand them their old job back if they don’t win the election.

It’s a hassle for everybody else though.

And the basis for the rule is what? So that members of Parliament, whilst in power, don’t have a conflict of interest. Jeremy’s suggestion on Joshua Gans’ blog, of letting such candidates quit their jobs after winning the election is quite a good solution. The persons would not have any time being under such a conflict of interest.

The Constitutions’s section 44, which prohibits persons under an office of profit from entering Parliament, is the blocker. It disqualifies persons from being “chosen or sitting as” a member. Maybe if our High Court feels like bending the rules a bit (i.e. any day of the week), we could have them “interpret” that to mean that a public servant can quit just as (or just before) being announced as the winner in their seat.

Alternatively, “interpret” that a teacher on unpaid leave is not in an office of profit at that particular time.

How else can we sort out this technicality that just makes things hard for people?

Zhanzuo v 56.com v za warudo

Zhanzuo.com have such a silly drop in traffic in late August? Well, I’m no expert on Chinese culture or web surfing patterns, but I can draw a few inferences from an Alexa graph. 56.com is another website funded by Sequoia Capital, which is the capital behind Zhanzuo. Look at its drop:

Custom-built applications are now a good idea

In fact, they’re the only sensible option.

Some people say that custom-built computer systems involve too much cost, time and risk. They say that it is better to take pre-existing, off-the-shelf products from software companies, and then adapt your business to make use of them. I disagree. In this current technological environment, custom-built applications are not just feasible, but they are more beneficial than ever.

Research research plagiarism. Plagiarism plagiarism? I hope so.

Apparently these days they do research into research. Fair enough. In Australia, we charge tax on taxes, victimise victims, and lots of other amusingly repetitive-sounding fun.

Not only that, but some bastard has been plagiarising! Naughty boy! It’s because of filthy cheats like him that I have to wade through 15 pages of “Plagiarism Policy” at work, uni and home. (“Oh wow, you would not believe the dream I had last night.” “Pat before you tell me, remember the plagiarism policy. You’re not allowed to have dreams that are the same as TV shows.”)

So plagiarism is apparently a huge problem, understated by all accounts. In fact, plagiarism is so prevalent that someone will plagiarise an article about plagiarism! What will we do about it?

Coffee is awesome: babies don’t lie.

Not only does caffiene have scientific memory and productivity benefits for adults, it also helps prem babies remember to breathe!

Searching for code injection

Oh wow, code injection is fun. Especially in the Encyclopedia Britannica!

Look at this for a hyperlink URL: [link].

If you just look at the first part, you see http://www.britannica.com/. Great, fine, dandy, everybody feels safe following a link to the Encyclopedia Britannica site.

Uh-uh. Nope. It’s actually abusing the Britannica search page to inject JavaScript code into the search results page. The moment you land on the Britannica site, you’ll be wooshed off to some dodgy pharmaceuticals vendor!

I hadn’t seen this kind of shenanigans before. It reminds me of a few years ago, when every two out of three sites running SQL databases would succumb to ‘SQL injection’ attacks. Those were cooler, because they injected code into the server and could earn the attacker money, while the present method just annoys people.

But then think about a website like Lifehacker. When you post links in the comments section on one of their articles, the link is abbreviated to just the website name. So, the link above would be shortened to www.britannica.com. Most folks wouldn’t know they were visiting some dodgy third party site until it was already happening!

EDIT: This blog has ‘Snapshots’ enabled, so when you hover your mouse over a hyperlink, it gives an image preview of the site you will visit next. When you mouseover the magic link above, Snapshots still shows you Encyclopedia Britannica. Yet when a human actually clicks on it…

Compare this behaviour to what happens when you search on Wikipedia: [link]. Wikipedia just turns around and says, “ask a silly question, get a silly answer”. No redirection for you.

It just goes to show that the Encyclopedia Britannica is poor by comparison to Wikipedia.

+1 Net-SNMP

I tried out Net-SNMP today. It’s a great set of command-line utilities for monitoring your SNMP devices.

In my case, I used it to monitor bandwidth usage on my Linksys AG241 broadband modem. If you like the command-line, as I do, then you’ll enjoy the no-nonsense, functional approach of Net-SNMP’s tools.

Plus there’s a handy wiki full of tutorials, how-tos and other information to get you up and running!

Start with snmpwalk, which lets you query a device for all the information it can give you. You can use that to find the names and numbers of network interfaces, for example.

Once you have the object identifiers (“OID” in SNMP-speak) for the data you want, use snmpget whenever you’d like to grab that info. Use snmpwalk and snmpget with grep, awk and perl to periodically poll the data you need, and:

• record it in a database,
• create a graph,
• alert you when bad things happen (e.g. too much downloading in one day), or
• anything else you could possibly want!

SNMP is great, and Net-SNMP is a great implementation.

Analyzing the word ‘analyse’

Tantalizing fact: -ize is the actual English way to spell!