Cyclekart progress, Nov 1, 2009

The weather has finally cooled off enough that I was able to work outside comfortably, and so I’m back on the cyclekart project. This weekend, I built half the front suspension.

First I had to fabricate a spring shackle:

Above, the raw steel parts used for the spring shackle. I welded the three parts together, then drilled holes on each end to accept the bolts (grade 5 steel) that connect the frame frail to the leaf spring.

Above, you see the spring shackle assembled. Note the world’s shittiest welding job. Yeah, I did that.
Next, I had to cut a hole in the bottom of the frame rail to accept the spring shackle:

It’s ugly, but that’s ok, because I welded some metal slabs over the side to (I hope) make it stronger where the cut has weakened the frame rail, and also to make the wear on the bolt that connects the frame rail to the spring shackle less severe.

Above, you see all the suspension parts disassembled.

And here they are assembled. Putting it together is a bit harder than it looks because all the holes don’t exactly line up perfectly, and the spring has to be stretched a wee bit to get it to all fit together.

Above, you can see how the shackle connects to the frame rail through the hole, and to the spring. And, you can see some more evidence of my super-excellent welding skills.

And, one final shot of the assembled frame rail. I was surprised how bouncy it was. As stiff as that spring was, I wasn’t sure the spring was even going to move at all, but it seems to work pretty well.

The scale of things small

Via metafilter, I find this: Cell size and scale, which is a neat little web app (flash, I’m guessing by the way it steals the mouse focus, though I haven’t checked). It allows you to zoom smoothly, viewing things as large as a coffee bean or as small as a carbon atom. The smooth zooming really gives you a much better feel for the relative sizes of things than similar things I’ve seen in the past which jump in 10x increments. One perceptive metafilter commenter asked how come it was that the X chromosome shown seemed too large to fit into the sperm cell which was also shown. Another few commenters clarified this mystery. Apparently there are special “packing” proteins in the sperm cell that kind of squash the chromosomes down. Neato.
Another metafilter commenter mentioned universcale which I’d seen before, but forgotten about. It i s kind of similar (though a bit more annoying with the music and intrusive interface, but still neat, and well worth checking out.)

Ok, Obama, that’s more like it.

Previously, I wrote about the Obama administration’s disappointing response to the U.N. resolution regarding blasphemy laws. It seems they may have come around.
according to this story on MSNBC

U.S. opposes bid to bar religious defamation Hillary Clinton says move by Islamic nations would restrict free speech
WASHINGTON – The Obama administration on Monday came out strongly against efforts by Islamic nations to bar the defamation of religions, saying the moves would restrict free speech.

WTF, Obama?

President Obama, please listen to Pat Condell.

Also, read
Jonathan Turley’s piece, “Just Say No to Blasphemy Laws in USA Today (Mon, Oct 19, 2009).

The Far Left Side

OMG! I have a new favorite comic. I was going to say web comic, but, what other kind is there these days? Newspaper comics? Who reads those?

The Far Left Side

Holy shit. It’s like the Far Side, but, extra-liberal!

Found this one via Dr. Jim’s Thinking Shop and Tea Room, which itself looks to be a blog well worth adding to your RSS feed subscription list.

Some love for… no wait, not love, for C++, and some art

So, awhile back, I bought a book, Coders at Work by Peter Seibel, and while reading it, I was gratified (my confirmation bias was tickled) to see the opinions various uber-programmers expressed about C++. Via reddit I came across this internet posting, “C++ in Coders at Work,” in which the author of this book discusses the opinions of the uber-programmers he interviewed regarding C++. Well worth a read, in my biased opinion, BTW.

Probably the most damning bit is what Ken Thompson has to say: “… But by and large I think it’s a bad language. … ” It reminds me of a joke I invented some time ago:

I don’t think the guy who invented C++ knows the difference between “increment” and “excrement.”

Jamie Zawinski doesn’t mince words either:

C++ is just an abomination. Everything is wrong with it in every way. So I really tried to avoid using that as much as I could and do everything in C at Netscape.

(If you don’t know who Jamie Zawinski is, he did a lot of coding on Netscape, Emacs (he gets an honorary high score in my vi vs. emacs game Word War vi for that), Xscreensaver, and probably lots of other stuff.)
And he has a few words about Coders at Work

Anyway, in scanning that reddit thread, someone (of course) linked to the C++ FQA lite, FQA standing for “Frequently Questioned Answers,” an amazingly thorough and well thought out (and well known) of response to the C++ FAQ lite.

But, in this latest visit of mine to the FQA lite, I discovered that its author, Yossi Kreinen is a bit of an artist. And he’s got an art blog:
http://www.yosefk.com/images/

Somehow I’d never noticed that before. As one programmer/C++ dis-aficionado/artist to another, I say, “pretty cool, man, pretty cool.”

CGI trees

Tonight, I got inspired by a post on reddit about the canvas widget that modern browsers support and also, by some stuff I’ve seen over the last year or so at Cambrian games (here is the cool canvas javascript app: Clicky. It’s pretty neat, check it out.)

Anyway, I decided I should try to make a “tree generating algorithm.” The usual gdk line drawing suspects were a bit too weak for this, so I started looking into cairo, and ended up writing a little GTK + cairo app that generated the image you see above.

Here’s how the app works: I created this notion of “cells.” Each cell is a square, which has a position (x, y coordinate), direction (angle) size, “turning,” and age. In the beginning of the program, there are just three cells, with size = 1, direction = up, (in radians), age = 0. With each iteration of the simulation, I scan each cell. Thee is some probability that each cell will spawn a new cell, with some limits based on the cells age (cells too young, or too old can’t make new cells.) The new cells inherit the parent cell’s direction with some small modification, partly random, partly the parent’s “turning.” The coloration is based on the cells age, so younger cells are lighter and older cells are darker. As cells age, they become bigger, to some limit.

That’s pretty much the whole algorithm, and that alone is enough to generate a picture like you see above.

This is my favorite kind of programming project. Small enough to do in one night, but big enough to get some really cool looking output. Damn computers are so fucking cool sometimes. This is one of those times.

Feed the world

Sarah Silverman has an idea about how to feed the world:

NRA push poll

I just got a phone call from an organization doing a push poll for the NRA, in which they had me listen to some NRA guy blathering on, then finally the question, something like: “Do you think third world dictators and Hillary Clinton should decide your gun rights?”

I told the pollster that “I think the Republican party and the NRA are a bunch of crazy, ignorant morons, and that this push poll is a pile of shit.” and hung up. That question is so insane it beggars belief. I can just see the pollsters dreaming up the ultimate question to drive the right wing insane, “hmm, we should threaten to take their guns away, and mention something about third world dictators, and uh, let’s see, oh yeah, Hillary Clinton. And the U.N.” Good grief, what blatant idiocy.

A couple of Daily Kos entries about this poll: 1, 2.
google search for “nra push poll “third world dictators” “hillary clinton”"

And now for something completely different…

popout
by youtube user cyriak via the comments over at jwz’s blog.

Told you it was different.