Perl is indeed like slitting the throat of a live chicken and smearing its blood and entrails on an effigy of your enemy.
66540 items (51620 unread) in 130 feeds
Linux
(11669 unread)
Gentoo
(516 unread)
Debian
(6571 unread)
Fedora
(5754 unread)
Kde
(2685 unread)
Gnome
(3577 unread)
Mandriva
(396 unread)
MySQL
(3238 unread)
Apache
(3436 unread)
Redhat
(90 unread)
Perl is indeed like slitting the throat of a live chicken and smearing its blood and entrails on an effigy of your enemy.
It seems there's a lot of people who think that images in Wikipedia can be used however they like. In fact, unless they've been released into the public domain, they have licenses that require some fairly minor, but still important, actions if you want to use them.
Below is an email I've just sent to the owner of Sydney Architecture Images: Newtown Gallery, who seems to have taken all the photos from the Wikipedia article on Newtown, without any attribution, and slapped some ads on it. I'm willing to give the person the benefit of the doubt.
Your Newtown Gallery is breaching the copyright of, it seems, all the
images from Wikipedia's Newtown article:
[www.sydneyarchitecture.com]
Specifically, and what made me notice, is that you are using my photo:
[en.wikipedia.org]
Images on Wikipedia are not free for you to use without restriction, unless specifically released into the public domain (for example, the Trocadero image you've used). Most of the images on Wikipedia are available under quite liberal licenses, but they do impose some requirements on you.
In the case of my photo, I have licensed it under the GNU Free Documentation License. This license has quite simple requirements, basically that you must attribute the author and license your copy of the work under the same license.
This image that you're using:
[en.wikipedia.org]
is licensed under a Creative Commons license, which has pretty similar
requirements.
So in summary, you are free to use these images, but only when you follow the requirements of the license. Please amend your site accordingly.
Wow Lindsay, that was bloody quick! Yesterday I asked about doing a geo mashup of the NSW government's name and shame list of dodgy food joints, and today Lindsay has done one. Impressive!
Now I wanna know how you did it, and how you managed it so fast!
A site screaming out for a geo-mashup is the NSW Food Authority's Penalty Notices list, which shows the companies that have received penalty notices for poor hygiene and the like. Problem is, the data isn't in a very human-usable form, but would be very amenable to a mashup and map overlay!
I think it's time for Australia to get something like PlanningAlerts.com, which does this for local authority (council) planning lists. So, for example, it could fetch the list of development applications for my local council, and overlay them on a map. More importantly, you could register with the site and have it email you if there's a DA near your house.
The code for PlanningAlerts.com is open, but it's kinda UK-centric. Any ideas where to start doing mashups in Australia?
Just bought this super cheap (AUD$88) black and white laser and spent a while going around in circles. FInally worked out that it works just fine with the Xerox P8e driver.
So to install, System > Administration > Printing. Click [New] and work out what port you've put it on (LPT or USB, with modern hardware the name will show up), click [Forward]. Now select Xerox and P8e. You're done!
Works a treat! Nice printer at a very competitive price. Even works just fine via my JetDirect which is nice. It's a bit noisy, but it'll be switched off most of the time anyway.
I bought a digital photo frame yesterday. On the face of it, this should be a pretty straightforward product. You stick a memory card in it that has photos on it, and it randomly cycles through them. But of course, the manufacturers of China can't stick with something so straightforward. Oh no, in the electronics manufacturing world you have to cram in a whole bunch of features that nobody needs into a simple product. Then crown it with a user interface that would make a 1980s VCR manufacturer blush.
So this thing, instead of just showing photos, also plays back video and sound files (thus requiring a speaker) and, of course, has a clock and calendar function. Because a clock is a mandatory feature for any electronics device. You wouldn't want to spend the next change to or from daylight savings sitting around the house with nothing to do, would you?
So what's my problem with all these extra features I'll never use? I can just ignore them, right? Well the problem is that instead of spending time making sure the thing works well for the one thing people buy it for, they've blown it on this stupid crap. I discover, after quite a bit of troubleshooting, that it doesn't support progressive JPEG files, something that's been supported in browsers since 1995 at least. This is the default format used by Smugmug for their resized photos. So I had to convert all those photos.
Then of course the UI is appalling. You can't seem to get back to the menu from the Settings page. Instead I had to turn it off, then on again to get back to "view photo" mode. There's no randomize feature, it seems, so you can get a random next photo. WTF? Surely this would be a base feature? And the UI crashes randomly, because they spent more time working on video decoding than fixing bugs in the stuff that counts. Grrrrr!
I also don't see why these things are so expensive. I'd expect the price to be where it is if it did something truly cool, like have a wireless connection and it just connects to your photo store and displays photos. But it seems instead an 8 inch 800x600 LCD with some basic electronics (and a bunch of crap you don't need) is $200 or so?
Yesterday's Crikey had a letter from Rachel Dixon that rather eloquently said what I've said before about the National Broadband Network projects aim of connecting 98% of the population to "metro-equivalent" broadband, or 12 megabits-per-second.
The crux of Rachel's argument is that she has a sister-in-law in Nundle, a town of 289 souls out towards Tamworth. It has excellent air quality and no aircraft noise, but isn't blessed with particularly good access to broadband. Rachel, on the other hand, has poor air quality and loud, regular planes flying overhead in inner-city Sydney, but access to very fast broadband. So if the residents of Nundle are to get the excellent broadband, why can't we demand country-equivalent air quality and serenity?
This is the big problem with setting the bar to 98% of the population. We haven't comitted to providing 98% of the population with sealed roads, reticulated water and gas, or town sewage. I bet a big chunk of the final 5% don't even have grid electricity! What about guaranteeing them access to metro-equivalent hospital services? I'm within walking distance of a fair number of hospitals, so shouldn't the good people of Nundle also have the same? Why is broadband so different?
If a business or family chooses to locate itself in a remote area, is it really the government's job to provide the services they would get if they located themselves somewhere closer to the population centres? Shouldn't there be some social and economic cost to locating yourself in a remote area? There are certainly costs associated with locating yourself in a big city: air quality, peace and quiet, land costs.
So apparently Lindsay Tanner reckons the government is keen to use web 2.0 buzzword compliant tools. The first lesson they need to learn is that we talk to each other, compare notes and compare responses we receive. Second thing about all this web 2.0 crap is that it's social, and the one thing that stands out more than anything else is robotically-created canned replies.
I sent Albo the following response:
Dear Anthony,
Thank you for your recent form letter. It's refreshing to see that my comments and queries are completely ignored, and instead of either considering or responding to them, you've sent all people contacting you the exact same message. I'll be sure to return the favour at the next election when someone tries to hand me an ALP how-to-vote.
In case you'd like to actually read, and respond to, my email, I'm including it again.
I'm not holding out much hope of a rational response.
![]()
It's US presidential election time again, and the big story as always in America is long queues at polling places. I don't get it. They have the simplest voting system going, where all you get to do is tick a box for your preferred candidate. Yet it takes ages, despite (or is it because of) the use of the latest voting technology.
Australian elections use a relatively complex system of tabulating votes, yet a clear winner is almost always known within a few hours of voting closing. The most I've had to queue to vote (outside of voting at Australia House in London, which often has queues around the block) in an Australian election is about two minutes.
![]()
So what is it that makes the voting process so slow in American elections? Why are there over 30 minute queues as standard?
So without even going into the serious problems with voting machines, it seems they don't actually solve any actual problems, and I suspect cost a lot more to operate than having a bunch of casual vote counting staff on the evening of the vote. In Australian elections I've been involved in (as a party volunteer) the same staff who operate the polling place all day stay back for the vote count in the evening.
The big question is, given this is likely to be the first election of a non-Republican since Bush stole the vote from Gore in 2000 (assuming they don't pull the same trick again), will the Democrats look at reforming the voting systems to prevent these state-based tricks?
Quote of the moment from the boss of the boss, holding his shiny new iPhone:
"I don't know how to make a call!"
JavaScript bit me again, with its concatenation operator being the same as its addition operator. So instead of the total amount of an order, I ended up with a string of all the prices in the order. Grrr.
I must remember what Crockford says in Appendix A: Awful Parts:
The + operator can add or concatenate. Which one is does depends on the types of the parameters. If either operand is an empty string, it produces the other operand converted to a string. If both operands are numbers, it produces the sum. Otherwise, it converts both operands to strings and concatenates them. This complicated behaviour is a common source of bugs. If you intend + to add, make sure that both operands are numbers.
Lots of people are writing to their MPs to explain why the proposed Internet censorship regime, as currently being pushed by Senator Conroy, won't work and is a draconian imposition on the rights of citizens. The biggest issue is that the system won't actually work, as measured against its purported goals, but will have a dramatic impact on the performance of the network and the availability of perfectly-legal material.
So get over to OpenAustralia to find out who your local member is, and get writing! The only language they understand is votes. Here's my letter to Anthony Albanese, who happens to also be Minister for Infrastructure, Leader of the House and a big shaker and mover in the ALP.
Dear Mr Albanese,
I'm writing to express my concern about the proposed Internet censorship regime being promoted by Senator Conroy. He would instead characterise it as a "filtering" regime, but as has recently emerged it will not be optional and so constitutes censorship, in the Great Firewall of China model, used to stifle political speech in China.
Two aspects of the proposed censorship regime concern me most, but they aren't the only problems.
First of all is the fact that it won't work. I'm an Internet technical type, and I've looked through the lab trial report and can't see how any technical approach could possibly work effectively. Regardless of whether or not it will slow down peoples' Internet connections--and it most certainly will--it will not actually be effective in blocking only the targetted material, and nothing else.
Any filtering system will either need to break the security mechanisms used for online commerce, trade and secure communications (like Internet banking and stock trading) or will be trivially easy to bypass. So illegal material will continue to be available to the people who want it. Technically-savvy children will also be able to bypass it, with their parents given a false sense of security that the government is doing their job for them.
A friend of mine lived in China for six months and was upset that he couldn't view sites like CNN, BBC or our own ABC to see what was happening in the world. In about fifteen minutes I set up a method for him to get around the Great Firewall of China and view these sites. This method would be used to get around the Great Firewall of Australia.
My second major concern is the way Conroy is stifling public discourse about his proposal. His office attempted to stop a respected network engineer from criticising the project, by placing pressure on the Internet Industry Association. He has also repeatedly characterised any critics of the proposal as being pro child pornorgraphy. This cheap rhetorical trick conflates the two aspects of the project: one is to "protect children from bad stuff on the Internet" and the other is to "censor illegal material".
The current system of voluntary Internet filtering software, made available for free to Australian households, is flawed but is probably the best approach we have. It means, at least, that parents have control over the level of censorship they want to impose on their children. Otherwise we end up with every special interest group banning their own hobby horses for everyone. That means Steven Fielding wanting to ban the human reproductive system and Nick Xenophon wanting all gambling made illegal.
Education is the key to handling the diverse range of content on the Internet. Families need the tools to make decisions for themselves, rather than having Big Nanny make decisions for them. One of the easiest approaches I've heard of is placing the family computer in a public location such as the dining room, so that everyone can be observed using the Internet.
I'd be happy to meet with you and explain the technical details of the proposal, and why it won't work. I also have some ideas about a face-saving fallback approach that does the maximum to prevent genuinely illegal content (child porn, terrorist material) that can be done by government without imposing draconian restrictions on citizens.
Simon Rumble
Marrickville NSW 2204
It seems Internet Explorer is quite strict about having multiple IDs in a document with the same selector. This, combined with the way I read the Prototype documentation, had me walking around in circles for ages completely unable to identify the bug. How annoying!
So I've documented it here, with a test to demonstrate it. I'll submit a ticket for the Prototype document shortly, as it does give the impression this will work.
I'm about to cut the agency out of the equation with my long-term contract (2 years so far) and reduce the amount they're taking from my wages. I don't want to avoid tax, don't particularly want (or really have) deductions and don't have any other significant income sources, so the company formation route isn't a good idea. I just want someone to do my payroll, insurance and super.
So can anyone recommend such an agency in Australia?
JavaScript's same origin policy security model means that a script can't (directly... there are sneaky ways aroun dit) request data from another site. It's quite strict. foo.site.com is different to site.com.
One thing I'd always assumed, but now I see how it wasn't a sensible assumption, was that the piece of code calling the data had to be served by the same domain name it was going to be calling. I've just done a little test of this, and discovered that's not true.
So if you're dead keen to see it, have a look at this test which loads the script from stout.rumble.net and actually loads Prototype from somewhere completely different. It has no trouble pulling data from www.rumble.net, but won't let me drag data from stout.rumble.net.
Already had a few responses about wireless sensors. Des suggests looking at SquidBee which is an open implementation. Quite interesting.
Rich recommends a book, Making Things Talk: Practical Methods for Connecting Physical Objects, which I've ordered. Looks very interesting. He also suggested Little Bird Electronics' workshops on the Arduino. There's one next weekend and I'm very tempted...
Thanks for the tips guys!
I've been watching the space of low-power wireless networking for some time, as it has some cool possibilities for sensor networks. ZigBee is one such standard that seems to be gaining support. It's a lot simpler and lower power than other protocols like Bluetooth.
Make magazine had a really interesting article recently about using ZigBee inertial sensors to track the crashes and hits taken by a roller derby team, and use it to trigger "zap", "pow", "boom" sound effects. Very cool, and it all seemed quite simple!
I'd like to use the ZigBee temperature sensors in my garden. The long-term plan is to monitor the weather and supply tank water through an irrigation system only when it's needed.
So has anyone used this ZigBee stuff and know any good resources to start out?
Geek resourcefulness at its best. This just got posted to the MythTV users list:
Subject: Cats and MythTV...
Recently my cat has decided that the top of my MyhTV box is a great place to have a nap... This is a situation I am not happy about, in spite of the positives (like noise reduction), I am concerned about fur shed and heat trapped. Moving the MythTV box to a place the cat wouldn't be able to climb/jump on top of isn't an option. I have tried putting a large plastic toy (a piggy bank) on top of the MythTV box, but the cat has found she can push that aside...
So, anything that hurts the cat is NOT an option, on the other hand, stopping the cat from napping on the top of that PC case is a a priority.
Suggestions?
Geeks being great problem solvers, the responses are well worth a read.
I can't remember who it was, but someone on the Planet Linux Australia was talking about using cheap USB flash drives to teach RAID. I thought it was Russell Coker but can't find it now.
Anyway, just wanted to point out that 1GB USB flash drives are $9 at Dick Smith. Insanity! I remember paying UK£80 for my 1GB flash drive. Not all that long ago.

So far this is my favourite little UI hack that I've found within Google Chrome. Kind of appropriate that a search engine company would get this right, I suppose.
If you're familiar with Firefox's search feature, Ctrl-F to enter a search term finds the results within the page progressively as you type in the little search box. Chrome moves the search box to the top-right and highlights the answers as you type. Even cooler, it highlights the location of matches in the current page along the side inside the scroll bar, so you can see the frequency of the search term.
That's so obvious now you wonder why nobody thought of it before, particularly given this is the same presentation used by graphical diff tools that developers use every day. Very cool.
Update: Julien Goodwin points out that this feature has been in Opera for over a year. I don't use Opera though, so I've never seen it.