The Japanese space program is doing some cool shit in the way of interplanetary spacecraft, and I have to hand it to them. I’m primarily impressed by their sheer ambition, launching bleeding edge missions that the more conservative NASA would tend to work their way up to. With previous interplanetary experience consisting of a couple probes to Halley’s Comet in the 80s and the Nozomi Mars orbiter in the 90s, JAXA launched Hayabusa seven years ago, on a sample collection mission to asteroid Itokawa.

This was the first attempted asteroid sample return mission by any nation. It was also powered by electric ion propulsion, a still fairly new technology that provides a long (years) but light push with little fuel mass compared to chemical rockets, and could autonomously navigate. After reaching the asteroid in 2005, it commenced trying to capture a sample, but here is where things started going wrong.

There were a couple brief landings but it didn’t seem that the pellet gun fired to cause debris to be captured. Also, a separate mini-lander called MINERVA was released too far from the surface. (Again – trying a lot of new things on one mission. And the mini-lander seems straight out of some Anime plot) Due to some malfunctions, getting the probe back to Earth has taken 5 years of Apollo 13 style engineering improvisation. Failing attitude control gyros were compensated for by angling the solar panels to surf the solar wind. Intermittent functioning of the 4 ion engines required constant navigational changes and (somehow) re-routing of components from two of the engines to allow operation of one.

All this resulted in the successful re-entry of the sample return capsule on Sunday. In the great video below, the capsule is at lower right, while the rest of the spacecraft is disintegrating behind it. Now we wait to find out if there was actually any sample dust inside.

As if that we’re impressive enough, JAXA just launched a solar sail demonstrator mission called IKAROS which has unfurled it’s sail successfully. This is the first functioning solar sail to be deployed – which relies on the radiation pressure of sunlight to provide propulsion. Solar panels are integrated into the sail, as well as variable reflectance patches for steering (this is crazy). And if that’s not enough for you, IKAROS released a small camera module to grab the below image of itself, relayed back to IKAROS wirelessly.

Really Japan – now you’re just showing off.

Let me take you back to an era before computer control of all aspects of spaceflight was considered necessary. It’s the early 70s, and although additional flights after Apollo 17 were eventually canceled, there were at the time plans afoot for longer duration stays on the lunar surface. However, a longer stay entailed an increase risk that the LEM ascent engine would not ignite when the time came to return to orbit.

What would the two surface bound astronauts do? Would they wait for a rescue mission? No. They would unstow a wire-frame with small thrusters and collapsible fuel bladders from the LEM.

They would then transfer the ascent stage fuel to these bladders, and climb onto a perch on top, with life support supplied only by their space suits. They would then ignite their small rockets, and arc into the sky, guided to a rendevous with the command module only by an attitude indicator, a clock, and a list of desired pitches and times.


Once the pitch and time sequence was complete and they were in orbit, they would sit tight and pray they matched the checklist close enough that the CSM could find them before they ran out of oxygen.

It seems impossible that such a guts-only scenario would come up in the future. Imagine riding from the surface to an orbital rendezvous on essentially a jetpack, holding a joystick and a stopwatch.
I love this.


Apollo Lunar Escape System at Wikipedia

I came across an entry on Google Books, and realized they have a map with all locations mentioned in the book tagged on the map. This is a neat feature. Check out the map for Around the World in 80 Days. Certainly an interesting way to access data in a book.

I was recently reading how St. Jerome was partial to plagiarizing the hell out of Origen and other sources, in some cases most of entire biblical commentaries. The author makes the point that plagarizing was quite a bit easier in the ancient world, as there weren’t that many copies of the books floating around when each copy had to be written out by hand. Printing was one paradigm jump in text availability, and we’re now going through another one.

I hear often how the internet makes it a lot easier for people to cheat and plagarize, but I think the increased accessibility of our accumulated texts actually makes the opposite true. We’re just coming through a period where people aren’t yet submitting papers electronically, and these papers aren’t yet routinely run through a programmatic comparison with the database of all books to check for plagarism. (This won’t address paying someone to write your paper). Think of how accessible human knowledge is now compared to even 10 years ago. Doesn’t it seem astronomically harder to find plagarism if you have to look through a physical book for the copied text?

From Wired:
“… at precisely 3:31:30 p.m. Pacific time on February 13, 2009, the 10-digit “epoch time” clock used by most Unix computers will display all ten decimal digits in sequence.”
1234567890 seconds since the epoch start.

If you have python installed you can see this by opening the idle shell (in start menu) and typing this:

1
2

import time
int(time.time())  #will display seconds since epoch

Or you could have it print a message at the appropriate moment so you don’t miss it (if that’s important to you :) :

1
2
3
4
5
6
7
8
9

def check_time(thetime):
   now = int(time.time())
   if now == thetime:
       print 50*'-'
       print 'time is',now,'!'
       print 50*'-'
 
while True:
   check_time(1234567890)

Hallelujah, Picasa for Mac has been released. I never could get used to iPhoto and it’s foibles, always wishing there was a Picasa release for the Mac. This also allows you to standardize on one photo organizer for all platforms. Fantastic!

ArbCamp08 was great, thanks to all the organizers, who had to scramble at the last minute to handle a huge turnout by changing the venue.  With more than 160 attendees, there was a wide variety of interesting breakout sessions and impromptu discussions.

I enjoyed the parallel programming session led by Jon Cohen – I hadn’t realized that graphics processors were up in the hundreds of cores, and that they can be utilized for scientific computing with the help of Nvidia’s Cuda compiler.  Users of this technology were first hand, including Eric Janikowski Jankowski, who agreed that this technology has revolutionized molecular simulation work.

Other sessions of note included a plan to set up a co-working space in the Arcadian Antiques location on Main Street, and a loose discussion of DIY photography.  Apparently the disposable digital cameras available at CVS are a good source of image sensors for DIY photography projects like kite photography.  Someone has even mounted one inside a volleyball to take pictures on each impact.

The event was, more than anything, a great opportunity to meet some of the interesting people in Ann Arbor, and for people in different disciplines to cross-polinate.

I’ve settled on the killer combination for easy backups to a NAS on OS X, and it involves installing no software.  It’s a combination of rsync, and scheduling with launchd.

The first step is getting your rsync command working.  I like it because it allows you to maintain a mirror of your files on the remote drive, and only copies over files that have changed when you run it.  Being a built-in command, it’s also a lot sleeker than installing a GUIed app of questionable speed. I mirror various key folders like this:

rsync -aSv --delete --progress ~/Pictures/ /Volumes/MACSHARE/rsync/Pictures > ~/scripts/logs/pictures.txt

There’s a number of options you can look at through the rsync man pages. I store my rsync commands in a file called backup.shIn order to Automate this process, you can set up a launchd task. Launchd replaced a lot of unix apps like cron on more recent OS X releases. You need to set up a plist file that points to your backup script and tells it how often to run, and store it in ~/Library/LaunchAgents/. Here is the file I use (you can call it something like com.backup.plist):

<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd">
<plist version="1.0">
<dict>
        <key>Label</key>
        <string>ryan.backup.script</string>
        <key>LowPriorityIO</key>
        <true/>
	<key>ProgramArguments</key>
		<array>
			<string>/Users/ryan/Scripts/backup.sh</string>
		</array>
        <key>StartCalendarInterval</key>
        <dict>
        	<key>Hour</key>
		<integer>2</integer>
		<key>Minute</key>
		<integer>0</integer>
        </dict>
</dict>
</plist>

You can look at the Apple man pages for launchd to determine what other arguments you can use. This one is set to run at 2am. You can also do some cool things such as running a script when a watched directory changes.Once you have your plist file, you have to register it with launchd.  In the LaunchAgents directory at the terminal, you can type launchctl load com.backup.plist A trick I used to make sure the scheduling was working, is to point to an alternative script (test.sh) containing say hello. I set the schedule a few minutes ahead and made sure the speakers were on to hear it say “hello”.

I’ve been following the Long Now Foundation and their quest to build a 10,000 year clock for several years now.  This episode of BoingBoingTV showcases some of the physical artifacts the foundation has machined, including an orrery, and a chime system developed in part by Brian Eno.

There are a lot of challenges associate with building a 10,000 year clock, from durable materials, to non-sticking bearings, to power sources, some of which are touched on in the video.  The Foundation’s overriding purpose is to encourage more long term responsibility.

I need to explore the anecdote told by Alexander Rose in the video, where when it was time to rebuild the roof of the hall in New College Oxford after 500 years, it turned out that nearby they had planted the oak trees that would be needed when the building was originally constructed.

via Machine Thinking

The user interface design snafu that bothers me the most is the “hit cancel for credit” routine while swiping your card at a local retailer.  Clearly, it would be ideal to present a choice between credit and debit with separate buttons on the touchscreen, as opposed to defaulting to debit and requiring the user to take the obscure step of cancelling out and then going to credit.  The interface is software, it can be changed, I don’t understand why this horrible user experience has persisted for so long.

One Laptop Per Child is an effort to produce a $100 laptop that can be distributed en masse to the school children of the world.  The first pass attempt at the hardware was going for $200, featured a 400 MHz CPU, flash for storage and wifi. And it’s powered by winding.

That’s great, but the second generation machine looks very interesting, both the screen and keyboard portions, as we’d conventionally describe them, appear to be multi-touch displays, offering a number of ways to use the device. You have to check out these pictures to see what I mean.

Nicholas Negroponte’s TED talk describes the concept:

I’m not convinced this is the educational silver bullet it’s sold as, but I’m interested to see how it develops over time.