Remember last time I set up a windows machine? Well I found myself doing it again and doing it in a hurry. I’m pretty psyched I got something up fast that’s pretty usable. Major change: Keybreeze.
Cygwin: install everything, add cygwin/bin (etc) to your path, point “my documents” to ~
Dexpot: It gives you Expose, multiple desktops and an improved alt-tab. Can teach you German
Keybreeze: This one is big. Hands down, this beats all the other Windows launchers I’ve seen. I’m still waiting for the deal-breaker failing that all the Quicksilver wannabes have. If, like me, you are scorning your Windows machine for lack of QS, there is hope.
Windows Registry: remap caps lock to ctrl
Python: I found it a lot easier to just to download the newest python, scipy and numpy for Windows rather than compiling scipy and numpy in cygwin.
I think I’m going to bind Windows-c -x and -v to copy, cut and paste to make the OS switch less jarring. I’d do this as a Dexpot macro rather than figure out how to do this in the registry
Expandrive just came out and Jon had a hand in it. It lets you mount ssh servers are local drives. It’s neat, you should try it. It’s $29 to keep—soon to be $39 no doubt. But when Cheetah comes out, Apple should have bought it and it will be included. Apple needs this. “Connect to server” works like shit.
Being atheist, nihilistic and all that, kinda puts a person on bad footing in terms of holidays. They’re all sort of arbitrary. Like the whole calendar. Random month length is really stupid, just like putting the leap day not at the end of the year. I know, December already has 31 days and February was made the bitch of a couple too many Roman emperors. At this point in human history, changing the calendar to something sensible would be harder than convincing the United States to switch to metric.
Have you guys heard about this?! Basically, there is a bus full of poison that is going to fall out of the sky and we have essentially no idea where. So we’re going to say “to hell with orbital debris” and try to blow it up.
This year, I think I’m going to put my IRA contribution into bonds. At least, I don’t think I’m going to keep my 90/10 spread. “Why,” you might ask, “when Vangaurd target 2045 is so safe and consistent?” The answer is because my brother-in-law, Brian, told me to.
The “guns kill” is where you get to show how good of a pilot you are. Missiles… eh. The guns kill is what you are really going for. Not only that, the system is recording video so if you can hold the guns on the other guys helmet for 8 seconds, you can show him later. The problem is that the bullets are supersonic and you can’t see them. So to help aim, you’ve got these tracers on your HUD and it’s just like in a video game.
I just retired my venerable Motorola E815. It was pretty beat up. Utilizing 1) the “new after two”, 2) online discounts and 3) selling my soul for another two years, I got the LG enV VX9900 for free. I’m hoping it lasts until someone (e.g. Google) has a phone standard (i.e. OS/API) and Verizon lets such phones on their network.
Looking out over a snowy Cambridge right now, I see total gridlock. A few cities, most notably London (and New York to some extent) are trying to fix the problem by implementing a “congestion tax”. I think it’s a great idea. I think it should be dynamically priced and implemented anywhere where it could be made to work logistically (e.g. any highway that already has a toll system).
If you don’t know about the Darpa Grand Challenge you should read about it a little first. You can probably find some good videos if you look. MIT entered for their first time this year and did remarkably well: 4th. 40 started, 11 qualified, 6 finished.
I was minding my business in the lab when a Macbook fell in my lap. It’s not like I’m going to throw it out. So I’m putting all the sweet stuff on there. Except I don’t know what all the sweet stuff is. I know Quicksilver and Texshop, but that’s about it. Could you guys (after the obligatory jokes/sarcasm) tell me what the other indispensable apps are? I’ve got the Leopard, I think… Whatever the new one is called. Well, I’ll have it in “about and hour and 52 minutes”.
In particular, I’m hoping there is something like Cygwin’s install-the-kitchen-sink option that makes the command line into something that could pass for a real unix prompt. Actually, I’m hoping for something like apt. But barring that, a kitchen-sink option would do.
Is the subject of the most recent “Speaking of Dartmouth” email. After reading it, all I know is that 1) it has something to do with the board of trustees and 2) I’m supposed to be mad about the lawsuit.
The lack of substance was pretty troubling. It reminded me of all the constitution mailings I got. i.e. a power grab. I can see Joran’s point that such power grabs are not necessarily bad for the college, I just don’t like when people try to manipulate me.
So with a little effort you can find out that the issue is the following. The alumni elect half the board of trustees. The board votes to add more seats to the board. Should half of those new seats be elected by the trustees or not? Jim thinks no, most of the alumni think yes.
Does anyone know how this problem came to be? If half of the board is responsive to the alumni, how could the other half manage to push such a thing through? And if there is dispute within the board, it should certainly be resolved within the current membership as opposed to after one side has granted themselves extra votes. So it would seem, I too advocate an injunction. I imagine that such would not severely “harm” the college or if it did, I hope it wouldn’t be “immeasurable” nor be too much of a “distraction … to the students faculty and staff”, or be “wildly expensive”.
Geocaching is the “sport” where you go “online:http:/geocaching.com” and look up some GPS coordinates, and then go to them looking for the hidden treasure. Yeah it’s dorky, but it’s my type of dorky. I’m into it.
I like to explore and people generally hide the stuff in interesting places, many of which I wouldn’t have found on my own. Most are relatively tame, but some are pretty adventurous. I think I might start hiding a few of my own.
But I don’t have a GPS.
This isn’t as much of a problem since I found Mapfinder. Mapfinder will overlay the relevant USGS topo map on top of Google Earth. Since you can also map the geocache on GE, all you’ve got to do is print it. If there are decent terrain cues, with a USGS quad (7.5’) I can orienteer with nearly identical precision to a cheap GPS—better if there is heavy tree cover.
Here are my complaints about Geocaching so far:
They use DMS and,
They hide stuff too well since…
They don’t appreciate the limits on GPS’s precision
Universal Transverse Mercator, is a locally linear coordinate system1, where you can simply read off the meters to target to the east and north. As in I’m at
0270400E, 4939200N and I want to get to
0270450E, 4929150N, then I should just walk
50m E and 50m S.
On a map, UTM makes a square grid that equally easy to use. The trouble with UTM is that it is local and approximate. Around here (from about Hanover to the tip of Maine and from the coast to past Quebec city), our reference surface is 19T. It is difficult to calculate distances between places with reference surfaces of different longitude (but not the same longitude1). There is an average length distortion of like .05%, but near the edges it is more like .5%.
1 The linearization is made by projecting the ground or “datum” (e.g. the WGS84 ellipsoid, the NAD27) to a reference surface. This surface is a segment of a longitudinal “cylinder”, parallel to the ground at the center of the region. It’s radius is chosen to minimize the average longitudinal distortion. Cylinder is in quotes because it’s cross section is the same as the cross section of the datum which is slightly elliptical. This fact means that it is surprisingly complicated to recover latitude from a UTM position.
Degrees, Minutes, Seconds, on the other hand is globaly exact. You have the angles made by you, the center of the earth, the axis of rotation and Greenwich England. You are standing exactly above the datum at point they are telling you. But to figure out how far away anything is, that is going to take trigonometry… Trigonometry and a book about datums… And Matlab would help.
The fact that the geocaching community uses DMS is indicative of that most of them just set a waypoint and follow the arrow. It is more fun if you also have some idea of where you are in addition to the GPS.
People think that because of all those significant figures in their GPS that they can bury a film canister and expect someone else to find it. You’re out in the middle of the woods with a $100 device receiving signals at like -160 to -180dBW! From space! You’re going to have to wait a long, long time before you can have centimeter accuracy.
What they could do is give a reference so that you can make an ad-hoc differential signal correction. For instance if there is something obvious nearby, like a stake, and they gave their measurement of both the stake and the cache, you could measure the stake, compute the difference and then add that to their measurement of the cache to get a much better estimate of where you will find the cache. And if they gave the measurement for the stake and cache in UTM, someone like me could probably just do it with a compass.
I built a zipline the other day. I thought the internet might want to know about it. Other people might want to do something similar.
I went out and bought 125’ of the strongest cable that they had at Home Depot which was 1/4” steel with a plastic coating. The first time I did it, the cable coating melted and plugged up the wood block. This actually was a more gentle stop than it became after I stripped the coating off the stopping region. (NB: Don’t use coated cable!). I would have used 5/16” if they had it but they didn’t, so my “Factor Of Safety” might not be up to aerospace standards. This also meant I couldn’t have as taut a line as I had originally planned1. As it happens, it turns out that the more slack line is a better ride anyway since it is closer to the ideal brachistochrone.
To attach the cable to the trees I got three u-shaped wire clamps for each end, end loops, a handful of (5/16”) steel quick-links (sometimes called rapides) and 2×6’ of 5/16” chain. I stripped the ends before I put the clamps and loop on.
For stopping, I got 75’ of 3/16” coated cable and 2×75’ of 1/4” and 50’ of 3/8” bungee cord (to be used as 6×25’,2×25’ resp). I put a lot of calculation into getting these lengths right.
First of all, you want it so that at maximum elongation, you don’t hit the lower tree.
Secondly, you want to keep the stress tolerable in the bungees.
Third, you want to make the force on the participant tolerable.
Fourth, you want the thing as fun as possible which means stopping as quickly as possible while satisfying one through three.
I decided that (3) would be satisfied by a peak acceleration of about 3.5G. Our target audience weighs about 140lb with gear, this plus (1), (plus that you would like to go nearly all the way t the lower tree) specifies a length and spring constant for the bungees. Then I dictated a maximum weight of 200lb and checked to make sure that 1 is not violated. This design made the top speed about 30-35mph.
This is made much harder by the fact that it is impossible to find specs on the bungees. What I ended up doing was experimenting for half an hour in REI with a yardstick and a quart of water. Based on reading in the internet, I think 2 is not being violated—I would like to have put a little more area of less bouncy bungee in there, but that’s what they sold.
I got one more 50’ length of bungee so that I could have the rest position of the arrest bungees be out of the way. I’ll make a diagram.
Then I got some scrap 2×4 and notched a groove into each so that I could bolt them together onto the wire. The arrest bungees are attached to this block.
I put the thing together almost completely by myself. I had Courtney belay me for setting up an anchor and threading a pull cord one day and that’s it. After that, I just used ascenders. To get the cable up, I used the trucker’s hitch and an auto come-along. I also had some turnbuckles but I didn’t need them. In fact I tried using one, but broke it. You should read (and believe) the spec on the hardware store turnbuckles: they are weaker then they look.
What else? Let’s see, there’s getting up and getting down. Initially I had planned on climbing a knotted rope (with an ascender for self-belay) but between weakness and the difficulty of finding an outdoor-durable, non-slippery rope material, that got scrapped. Instead we have a rope ascent. This is done with an ascender and grigri. Yes a grigri — it is awesome.
Once you get up there, you put your pulley (trolley actually) on the cable, put a backup beiner on there too and take the ride.
Once you stop, there is a rope and a sling attached to the arrest block. You put the rope through your grigri, stand up on the sling and unweight the pulley. Take it (and the beiner) off the cable and then you can lower yourself.
From beginning to end the whole process is about 4min if you’re comfortable with all the gear. Here’s a video. I set the camera up on a branch so the camera-work is on the minimalist side (do you guys know that I’m a cinematographer now?). The beginning and end are sped up, but the ride itself is at regular speed. I’ll try to get some better video later.
1 You can take a rope and tie one end to something fixed and the other to something heavy. Then by pushing perpendicularly at the midpoint, you can exert a much large force on the heavy object. Clearly this is a simple machine, but it defies easy catagorization as an inclined plane, pulley or lever. Any thoughts?
I’ve always said that the primary flaw in Jon’s theoretical economics is that agents never have perfect information, and typically have very bad information. Ad’s, as annoying as they are, can only contribute to the information.
So I went to do my taxes yesterday—online, of course—easier for me, easier for the government, more accurate, faster return. So I went with TurboTax. Although I would have loved to access the pages that defined the fucking terminology, those cost money. So I avoided those. I made it all the way to my 1099 (for consulting) at which point it said that I needed to fork over $30 to continue. Since I wasn’t about to pay that fee, I had to either punt a 30min investment or lie and file that income as something else. Looking at TurboTax’s competitors doesn’t make me feel any better. I’m going to head down to the post office and do the thing by mail. I am not pleased by this development.
Why the hell doesn’t the IRS have their own site? The only reasons I can think of are dirty, i.e., that some fraction of those fees are going to the IRS but they are going as bribes.
