The Sun is five orders of magnitude larger than the Earth.
Eugene Volokh found that only about two thirds of lawyers understand what orders of magnitude means. As in, “The population of the US is greater than that of Iceland by three orders of magnitude”. Which means of course . . . by a factor of 1000.
And I have to sadly agree with the take-away lesson he gets from this: Even though most lawyers know what the phrase means — including, reassuringly, Chief Justice Roberts — one shouldn’t use it in legal writing. After all, the purpose is to communicate, right?
CivicApps and the Search for Ideas
I just found out that I won one of the two Best Idea awards given in the first annual CivicApps for Greater Portland contest; I was completely surprised and very honored. My idea is Community-Contributed Datasets.
CivicApps.org in a nutshell
Back in the old days, i.e. 2008, cities like Portland would create apps (interactive websites) for us like this:
The old way of making apps
They’d get a team together, look at the Datasets available to it, come up with ideas, and then build the apps. PortlandMaps, an awesome web app, was made this way. But there’s a new trend and Portland’s setting an example for other cities. First, the city adopts a policy of transparency, publishing all of its raw data in addition to the finished apps:
First, cities make their raw data available
Then, curious and entrepreneurial individuals, companies, and other organizations step in and independently create new apps. This frees up city resources and results in more apps and services coming online:
Then, 3rd parties can step in and contribute
This is the entire purpose of the CivicApps.org project and website: to enable this creative environment. By the way, this is the same idea behind the federal Data.gov project.
My idea: we can contribute data too
It occurred to me that many organizations have valuable data about the city. And some even have the expertise to publish the data in “open formats” just like the city of Portland is doing:
My idea: We can provide datasets too
For example, my project, OregonLaws.org, publishes a glossary of legal terms specific to Oregon. This is already up and running as a Dataset.
It will, though, take some work to implement my idea. Some questions to answer are; How will these 3rd party Datasets be integrated into the palette of “official” ones? How will the data be authenticated or verified?
The near future
The end result of all this is that the city could become a nexus of information. In addition to providing raw data about itself, it can act as a means of identification and authentication. There’s a precedent for the government acting in this role, e.g. the Secretary of State’s business registry performs a similar service.
Update
Oregon Business Magazine covered the CivicApps contest: Keeping PDX a Hot Spot for App Development
The court granted summary judgment for YouTube because
- The Safe Harbor provision sure does cover the YouTube site,
- YouTube kicks a** implementing it, and finally,
- Safe-Harbor-defeating “actual knowledge” of infringement means actual instances of infringement, not some vague knowledge of infringing activity, or whatever it was that Viacom was trying to assert.
Why “git” instead of “svn”?
As I learn git, I’m experiencing advantages that I haven’t seen described together. Also, I haven’t seen a description from the perspective of the real-world differences in your files, so here goes:
Background context
Subversion uses a single, central repository which is in a remote place. And in your project, it maintains .svn/ directories in every sub-directory. Git on the other hand, gives each developer their own full-fledged repository. And then, instead of relying on many hidden directories in your project, it creates one that lives only in the top-level directory; in .git/. So again: that’s the only place it stores info; nowhere else in your code.
Interesting, huh?
The advantages I’ve found
So now that I’ve begun to use git, it’s already been a nicer experience than subversion:
Deleting a subdirectory doesn’t screw with the versioning like it does in svn. (Because no version info is stored there. In svn, you need to repair things; re-create the empty directory so you can delete it, etc. I do this to myself periodically.)
It’s totally painless to put existing stuff into git — it stays right where it is. E.g., a website that’s in use. You can initialize the repository and commit the current version of the files while they’re in use. This isn’t possible with svn: you have to check it in, then check it out.
It’s easy to work on a new idea: it’s really easy to create a branch to work on it, and then switch back and forth. In svn, this is enough of a pain that I don’t do it in actual practice. Git finally makes it easy for me to use revision control like I’ve always wanted: be able to check out our group’s code, test ideas out, and rollback in my own little world without disturbing the others, and finally commit only the changes I want to.
Git assumes you meant to delete that file. That’s been nice and convenient: I deleted a file without going and telling git first about it. Git adds a ‘delete’ operation to your list of things to check in. Awesome. Svn does it the other way around: You can’t just do normal file operations; you have to also tell subversion that yes, you did mean to do that.
Git has the GitHub killer app. It’s the Netscape of the WWW. I haven’t found an equivalent for Subversion — web-based or not.
So that’s what I’ve come up so far; I’ll add more when I see them.
I’m releasing a new open source library, HTTP-Assertions, that I use behind the scenes at OregonLaws.org. It helps me keep the server running smoothly and make sure that my changes to the config files haven’t introduced new bugs.
It introduces these new assert methods to the Rails testing environment:
assert_200 assert_forbidden assert_temp_redirect_to assert_perm_redirect_to
Here’s an example:
assert_200('https://www.oregonlaws.org/ors/161.360')
See the README on github for more details.
The End of (Visual) Orphan Works?
Orphan work: SpilledCoffee.jpg
Ironically, I was about to write a blog post, A New World of Orphan Works. I had wanted a picture of spilled coffee in an ad for OregonLaws.org: I had found this one online, perfect for my project, but without any attribution. Contacting the web page author didn’t help.
Enter Tin Eye, a “reverse image search”. Upload an image, and it instantly finds many places on the web where it appears.
It may turn into a really useful tool for several purposes. Here, it found unauthorized re-uses of a product logo belonging to Panic Software, a local Portland software shop:
Here, a more typical search result showing where one’s work ends up:
Finally, Tin Eye can be used to track down the origin of an “orphan work”. In my case, I discovered that I already had the rights to use the image (stock photo in a collection I own).
From a website for kids / marketing vehicle made by a “naturopathic doctor and mother”.
This is how you use the Belladonna one . . . you hold it upside down . . . then you twist . . . then you pop it in your mouth! . . . It tastes just like sugar.
Update: Kids & Homeopathy has been taken off-line. However, I had been careful to save a copy. Here are a couple of stills, edited to protect the child’s privacy:
“...you hold it upside down...”
“...then you pop it in your mouth!”
How many ethical issues does this raise? In my opinion, the video steps over the line by…
- Marketing medicine to children,
- Using the pills’ similarity to candy to sell them,
- Blurring the distinction between medicine and candy,
- Promoting a culture of pill-popping and keeping a medicine cabinet full of “remedies” at all times to children,
- Teaching children how to circumvent a child-safety device,
- Teaching children to self-diagnose instead of turning to a parent or guardian,
- Blurring the meaning of dangerous poisons such as Belladonna,
- Focussing on the symptoms instead of the underlying problems: Why does she have headaches frequently enough that she knows what “medication” she prefers for them? Is there some underlying problem that’s going ignored while she self-diagnoses and self-medicates?
- Marketing a disproven therapy (homeopathy) to children,
- Harming this particular child through the inculcation of these values.
Did I miss anything?
Update
- The video’s creator responds.
- Another problem with the video’s concept occurred to me: As can be seen in the screenshot above, the girl takes medication on camera even though she’s not experiencing any health problems. This is both questionable (further teaching that taking meds is a lighthearted & fun activity), and odd (evidence that homeopathic medications maybe really aren’t.)
Malcolm Gladwell is Dead to Me
I’ve been a fan of his for a long time — I loved The Tipping Point, and saw him speak at Powell’s downtown Portland. I knew he had his critics who challenged his conclusions or methods, but I saw that many of the attacks were ideological, and so didn’t think much of them.
But I recently stumbled onto his blog post, Pit Bulls…. It’s short; just two paragraphs long. The first is decent; he describes how, even though a pit bull attack caused no serious physical harm,
. . . the attack could have been much more serious. If everyone hadn’t run to the defense of the boy so quickly, and if the boy’s mother hadn’t done exactly the right thing (lying down, with her body covering the boy) the boy could easily have been badly injured. The attack was horrifying, even if no one was seriously hurt . . .
But then the concluding paragraph is wrong on its face:
. . . part of the rhetorical arsenal of those who get hysterical about Pit Bulls is to pretend that every dog bite is a medical catastrophe. . . . [But] more people are admitted every year for non-dog bites than dog-bites—which is to say that when you see a Pit Bull, you should worry as much about being bitten by the person holding the leash than the dog on the other end.
[Emphasis mine] I mean, really? He really believes (and thinks he’s proved) that people, or at least owners of pit bulls, are more likely to bite you than their dogs are? My friend Morgan pointed out that this goes against our own experience: We could each remember numerous times we’ve been bitten by dogs, but we’ve never been bitten by a person.
Where Gladwell goes wrong
His logic violates conditional probability which is based on this central idea: It can be the case that if A happens, B occurs too. Yet this does not imply the reverse: That if B happens, then A will occur.
Here, Gladwell’s premise is, when people land in the hospital due to a bite (A), it’s probably due to a non-dog animal (B). He then draws the erroneous reverse conclusion: When encountering a human and a dog (A), it’s more probable that the human, rather than the dog, will bite you, sending you to the hospital (B).
And another thing
I have to also say that his “those who get hysterical about Pit Bulls” characterization is B.S. In my experience, pro-Pit Bull people (they exist) and Anti-BSL advocates often get hysterical. However, I haven’t seen this in advocates for BSL.
Laws that restrict dog breeds which are thought to be dangerous are called Breed Specific Legislation, or “BSL”. I haven’t made up my mind about them, but I lean in favor of thinking they’re a good thing: after all, (1) we already ban the ownership of many kinds of dangerous animals. And (2) the cost to society and individual freedom isn’t great, e.g.:
“No, you can’t keep that coyote you saw in Powell Butte. And if you want a pet, there are plenty of Labs waiting to be adopted at the Humane Society.”
So I conclude that if there are overly dangerous dog breeds, then it’s reasonable to restrict them in the same way.
A cute but questionable “IQ Test”
But there are many people and groups who advocate against BSL — and in favor of their being able to keep and sell these breeds. Ok, fair enough. And one of them, “Paul Glassner SF/SPCA”, has written a breed ban IQ test that begins,
1. If you were the sheriff in your town and you learned that Toyotas were disproportionally involved in more auto accidents than any other model, would you:
(a) ban Toyotas and confiscate the Toyota of anyone caught driving one
(b) arrest the drivers responsible for those accidents?
And that’s it — those are the only choices. The rest of the test builds on this question and makes the points one might guess it would. And so I found the “test” pretty shoddy—essentially propaganda. Because what about,
(c) look at car ownership to see if a disproportionate number of people own Toyotas as well. If not, investigate the Toyotas to see if they’re flawed by design.
…which then leads to,
(d) potentially all of the above.
Because that’s what we do in the real world: If a car type is too dangerous by design, then we ban it. If a model’s occupants would die from a single low-speed collision, we ban it. Ditto if a model’s brakes are fundamentally flawed.
The car metaphor works in favor of BSL
Now that I think about it, this does apply to dangerous dogs: The overly dangerous dog is one who’s likely to bite more often (a car that causes more accidents), and when it does bite, more often maims or kills its victim (a car that does the same to its occupants). In other words, the “overly dangerous dog” is one that’d be responsible for higher incidence and mortality rates. If there were a car like this — and there have been — you can bet we’d take action against them.
I love NPR and Fresh Air. But right now, I’m very disappointed.
Terri Gross is chairing a a discussion about health care solutions with two people. One is Paul Krugman. I knew the name, but not too much about him, and so here’s what I found:
- Professor of economics and international affairs at Princeton,
- Centenary professor at London School of Economics,
- 2008 Nobel Memorial Prize winner in economics,
- Author of “25 books and 40 scholarly articles” (Wikipedia).
I thought to myself, wow, awesome — I want to hear what he has to say. And the other person? “Stuart Butler, [some position at] conservative thinktank, the Heritage Foundation”. Ok, so red flags go up for me. Who is this guy? Here’s the best I can find:
- Author of a series of “WebMemos” and “Heritage Lectures” on the Heritage Foundation website,
- Adjunct faculty at Georgetown,
- Identified as a very influential person by a D.C.-insider periodical. (per his resume)
For cryin’ out loud — why why why? How in the world does a discussion between these two people serve listeners? How is Butler any different from Creationists who write “research” papers that they publish in their own personal “scientific journals”?
Here’s just one of the many problems with this:
As a consumer of news and content, I’m shortchanged.
If there is a well-represented competing current of research and solutions in the field, then I’d like to hear from someone who would disagree with Krugman on some issues. But please, this person should have similarly solid credentials, and should represent a similarly broad stream of research in the subject area.
