This is the fourth time I’ve seen Bill Bailey at the State Theatre (Tinselworm, Dandelion Mind, Qualmpeddlar and today, Limboland). Each time hilarious, intelligent and tonight; featured Miley Cirus in the Style of Kraftwerk.
Posted by: Ian on 29 October 2014
Posted by: Ian on 25 October 2014
A couple of days ago (presumably on a slow news day), the Herald Sun ran an article stating that the Police were calling for a ban on cyclists using Heidelberg-Kinglake Road in Victoria ‘for their own safety’.
“It’s the most dangerous road to ride on”
“I would love to see a ban on cyclists going up that stretch of the road as my concern is for their safety.”
“A car might hit them and the car driver is going to be fine, but the cyclist is going to be badly hurt or pushed off the embankment.”Sergeant Lindsay Dixon (Diamond Creek)
I was pretty sure that information on accidents would be available online, and since my day job is mapping, I thought I would see just how dangerous this section of road was.
Firstly I hit up data.vic.gov.au for crash statistics, and was lucky to find a data extract covering January 1st 2006 to June 30th 2013. A quick skim through the download showed it to be a collection of CSV files that were keyed to a unique ID for each accident – easy to join as needed using GIS. The biggest issue with displaying the crash locations was trying to work out the coordinate system so I could display it with some underlying roads. The coordinate tags are labelled amg_x and amg_y so I was looking at the various Australian Map Grid options, but none of those were close. Finally I came across a hint that I was looking for ‘Pseudo-AMG’ so that gave me some routes to search along.
Fortunately, Nyall Dawson had run into the same problem last year and posted the proj4 string which I could use to create a custom Coordinate Reference System in QGIS.
+proj=tmerc +lat_0=0 +lon_0=145 +k=1 +x_0=500000 +y_0=10000000 +ellps=WGS84
+towgs84=-117.808,-51.536,137.784,0.303,0.446,0.234,-0.29 +units=m +no_defs
If you have access to ArcGIS you can also create your own custom coordinate system using the following data:
Central Meridian: 145 E
Central Scale Fact: 1
False Easting: 500,000 m
False Northing: 10,000,000 m
Geodetic Datum: AGD66
Having a custom coordinate system really sucks in general; there is rarely any need for one. However, I can see the reason here: most, if not all the standard Australian systems have a boundary that goes right through the middle of Melbourne.
By comparison, the basemap was easy; I’d previously downloaded an OSM file covering all of Australia, so it was just case of clipping out with the bit that I wanted with the command line and styling it up. It comes in in WGS84 (Lat-Long) so its easy to manipulate.
Ian$ ogr2ogr -f "SQLite" -dsco SPATIALITE="YES" -spat 145.07034302 -37.71098578 145.48095703 -37.47921744 Kinglake_OSM.sqlite australia-latest.osm.pbf -progress
First up, we have the data for all non-bicycle accidents on Heidelberg-Kinglake Road between Kinglake and St Andrews.
Second, we have the data for all bicycle accidents over the same time period1.
The statistics from VicRoads show that on this section of road, the ratio of injurious motor vehicles accidents to that of cyclists is approximately 25:1, far from ‘the most dangerous road’ as claimed.
Following the devastating fires of 7th February 2009 (Black Saturday), cyclists were some of the first people coming back, buying coffee and helping support the community. A good cafe is something cyclists look for, willing to ride out of their way for a favourite pre or post ride coffee every week. A hoard of hungry riders make a good and welcoming cafe a lot of money before most people are even out of bed at the weekend while an unwelcoming one will soon fine itself ‘blacklisted’ by cyclists. A sudden ban on that road will immediately cut down on the number of riders in the area, and much less income for the cafes.
‘Ban cyclists for their own safety’ seems to have currently replaced the constant bleating for registration and insurance, but it is no less flawed and, as far as I’m concerned, a reasonable indication that nothing said by that person is worthy of consideration.
Posted by: Ian on 13 October 2014
Just before heading out for this evening’s run, I remembered that the weather forecast had mentioned storms coming through, and since it had been sunny all day, they had to come sometime.
Fortunately, we have a Doppler radar covering Sydney, which is great for seeing what is on the way. Expecting a few storm-cells, I loaded the page and was confronted with a red and black snake stretching the height of the map, and with the animated frames, it was apparent that it was going to hit in the next 15 minutes.
Now there is running in the rain, and there is insanity, so I decided the only running was going to do was up onto the rooftop, then down again to get a tripod, then back to the roof again.
As I set up, I could see the last few planes scrambling to get in and out of Sydney Airport ahead of the storm. Fortunately the front came through so fast that delays for travellers shouldn’t have been too long.
An iPhone 6 with aGorilla Pod and Grip Mount makes a remarkably good setup for quick deployment. I was using it to hold the iPhone steady while I tried out the Hyperlapse app. I did add an umbrella till the storm got closer, when I got slightly uncomfortable being on the top of a tall building with rapidly approaching lightning.
Added 2014-10-15I have a Python script that I have been messing with that captures the data from a Bureau of Meteorology, processes it, and then graphs some of the values with a nod to XKCD. Sometime I want to develop it further with a wind rose as well.
This storm event is represented by the smaller of the rain spikes, for all the mess on the radar, the actually quantity of rain was rather small. You can see the sudden drop in air pressure that preceded the front.
Posted by: Ian on 10 October 2014
Feeling pretty tired today and didn’t fancy too many hills; the 80km round trip to Cape Solander has about 400m of climbing in total.
Posted by: Ian on 8 October 2014
White wine with red meat is probably a hanging offence, but it works for me. A very nice meal out at our local Mediterranean restaurant.
Posted by: Ian on 5 October 2014
Fresh rock oysters after a morning climbing the gorges north of Sydney.
Posted by: Ian on 19 September 2014
Five years after getting my 3GS, it was finally getting past its best (and held together with sellotape). I managed to get get in on the pre-order and TNT dropped off a shiny new toy yesterday.
Posted by: Ian on 5 September 2014
From now one, time before 11:47am today shall be knowing as BC – Before Cronut.
Posted by: Ian on 26 July 2014
Note: In this post, I use ‘Kickstarter’ as the generic term for crowd-funding websites, the same way that all vacuum cleaners in the UK are referred to as ‘hoovers’.Crowd funding platforms have been around for a few years now – allowing someone to propose a project, and for the general public to ‘back’ these projects with up-front cash in exchange (typically) for reduced price products if and when they are backed and developed. I don’t have a problem with the majority of projects, and have backed several myself, including the Mio Alpha, Misfit Shine, LIFX, See.Sense, Kensei Fantasy Creatures and the Scanadu Scout. There are however plenty of Kickstarter products I have issues with, or more accurately, i have issues with how they are using these crowd funding platforms.
For Market ResearchI see plenty of products on Kickstarter which, frankly, seem to be solutions desperately looking for a problem. This one is for some lights that you fit to your helmet and depending on which way you tilt your head, they flash as turn-indicators. With 10 days left, it had raised 0.4% of its goal. So rather than risk spending your own time doing market research, you can put it on Kickstarter. If it gets the funding you probably had a good idea and you are on your way, if it sinks like a stone then your idea was probably rubbish, but you saved yourself money with free market research and you can get on the next project.
For Risk AvoidanceSomeone on Twitter linked this project today. I was amazed to see they have made 10x their goal, for an alloy framed belt drive bike. Really. The tagline is “Maintenance-Free Bicycles that Make Cycling Easy”, so I looked carefully. Does it have bearings made of ceramic or some other advanced material? No. It has a belt drive, like some bikes have had for years. Also, looking at the traditional (rather than A-head) headset, that is one part that will be anything but maintenance free, especially when you consider the complete bike is ~$4001. So, why is this bike on KickStarter? It offers exactly nothing innovative or new. Its a cheap bike with a belt drive. Rather than risk their own cash on the venture, Kickstarter allows the project creator to get someone else’s money to fund the project, even if the bank won’t give you a loan. Why risk $30,000 of your own when you can get it funded with backers money.
One thing to remember – these are often projects that venture capitalists have passed on.
As a Business ModelRevolights were originally funded on Kickstarter in September 2011, getting roughly 5x their target funding. The first Revolights were a kit that you could fit to your bike wheels so the front half of the front wheel was white, and the back half of the rear wheel was red.
They were back on Kickstarter again in May 2013 with ‘Revolight wheels’ – wheels with permanently installed lights and with which they achieved 6x their goal, and once more in April of this year with ‘Revolight Arc’ – a light-up mudguard, which failed to get its $100,000 goal.
Given that the idea of Kickstarter is to get an idea or business off the ground, you would think that going back for a second and third round with new products was a bit much. If the first product was funded, surely they should be using profits from those to fund new projects. It is worth noting that Revolights raised $215,000 with the first round of funding and $85,000 for the second – surely enough to get a company running, if you have a functioning business model.
Essentially, Revolights model seems to be that rather than risk revenue raised from existing projects, running another crowdfunding round allows them to get free market research and offset the risk of launching a product that might fail.
To Pay for a HolidayThe Printing Bike Project is trying to raise funds so they can build a cargo bike with a printing press on it, which they will then ride from Bristol in the UK to Mainz in Germany, where Johannes Gutenberg invented the printing press. Apparently they are going to print postcards on the way and do some outreach when they get back, but to me it looks like they want someone else to pay for a holiday in the name of ‘art’. They are looking for £8,000 and I’d just like to point out that you can get a Bullitt Cargo Bike for around £1,500 and the press for £600. I guess the leftover £6,000 covers alot of holiday.
I do like Kickstater and the other crowd-funding platforms – they certainly allow small startups to have global exposure for little cost (much like iTunes and small artists), and the natural selection process helps the good ones and doesn’t harshly punish the less good, they just don’t get funded.
Of course, once a project is actually funded, there is a whole world of issues from then onwards in terms of timely delivery and promised functionality, but those are project management issues which can apply to almost any project.
- To put $400 into perspective, a Chris King headset (generally considered the best), retails for around $250.?
Posted by: Ian on 25 July 2014
This years L’Etape du Tour was run on the classic Pyrenean stage route from Pau to the ski station of Hautacam, a total distance of 145km.
Before the half-way point, the route starts to ascend the fearsome Col du Tormalet, crossing the 2,115m summit with 50km of the stage remaining. A 35km descent into Ayros Arbouix leaves just the 15km climb back up to 1,520m.
In the ‘amateur’ race, Frenchman Loic Herbreteau crossed the line four minutes ahead of second place, with a time of 4 hours 47 minutes and 29 seconds. Vincenzo Nibali won Stage 18 of the Tour in 4 hours 4 minutes and 17 seconds, almost three-quarters of an hour faster.
The Tour organisers, gave todays stage a coefficient of 5 as a ‘very difficult short stage’, which, given Nibali’s average speed allows riders an additional 17% of the winners time to avoid elimination. So for todays stage, the elimination time was 4 hours 43 minutes and 29 seconds – this year, none of the Etape riders would have survived the ‘broom wagon’.
Lampre-Merida rider Davide Cimolai was the last TdF rider to cross the line, safely inside the elimination time by around 9 minutes.
Posted by: Ian on 18 July 2014
Picnic in the park today; fresh bread, good cheese and a Spice Weasel.