Proof of the Power of Permalinks


What do avid bicyclists and handgun owners have in common?


Probably more in common than not. Two things are a need to know the law, and a desire to discuss it.

Bicycle traffic law discussion on
Bicycle law discussion on

So I conclude that the ability to find, read, and share the law online is a good thing. It might even be the beginning of a more transparent and democratic era. These screenshots show people participating in exactly this way, who were helped out by this website. (Awesome!)

Forum discussion about concealed handgun license application procedures
Concealed handgun license discussion

This all might look easy, e.g., “So just make a website, and people can link to it … what’s the big deal?” In reality, though, it’s difficult to pull off, and so the status quo is, it doesn’t happen.

Engineering a working permalink system can be tricky
Engineering a working permalink system can be tricky

Just look around the web: most pages don’t have permalinks (“Persistent URLs”, “PURLs”) that people can use. And often, even if a web page does have a URL that seems good for linking, there are other show-stoppers: it may not stay working. Or if it does stay working, then it may not contain the same content in a few months’ time.

Discussion about landlord/tenant law on Reddit.
Discussion about landlord/tenant law on Reddit

Some great people have taken the story further, writing lighthearted web pages (Tim Berners-Lee), and creating videos (

Family law discussion on Reddit
Family law discussion on Reddit

USPS: ZIP Codes are “commercially sensitive . . . proprietary” information

In a Nutshell: No ZIP Codes for You


The U.S. Postal Service denied our FOIA request for a list of current ZIP Codes, because the information is “proprietary business information” and its release would not be considered a “good business practice”. This followed their refusal to license us their online ZIP Code API because we were not planning on using it for USPS shipping. They suggested that we purchase a $3,000 per year “data product” containing the information. This offering, however, has so many technical and licensing hurdles that it’s essentially unusable.

Our Hunt for ZIP Codes Started When We Became Interested in Mapping Restaurant Health Inspections


We create apps which increase access to laws and government info. This is mostly a labor of love, earning enough money to cover the costs of running the Internet servers.

There isn’t yet an app or website for easily browsing restaurant health inspections in Multnomah County, Oregon. The County does have a web app, but it has a lot of issues, and won’t be upgraded until late next year.

Critically, while sorting through the County’s data for each restaurant, we found that the city is missing, although the ZIP Code is there. And so to place these on a map we needed a current list of ZIP Codes and corresponding city names.

Our search for this simple information culminated in this rejected FOIA request to the U.S. Postal Service. This is because our “open source” open-government use of the data is not an acceptable reason to either 1) provide us online access to the USPS API, or 2) send us a current list of ZIP Codes and cities.

Texas statutes are online, updated daily

The date in this footnote is correct — we really did check this section earlier today. In addition to rolling out the Texas statutes, we’ve made Texas the first state in our new “Cloud”-based statute processing pipeline. This new platform retrieves, scans, and publishes changes to statutes on a daily basis. It finishes up by creating links to the sources so readers can “trust but verify”.

Over the next few weeks, we’ll be migrating California, New York, and Oregon over to this new system. One note: although Texas is online, we have a lot more work to do for it to meet our standards for a usable, modern reference site. Coming soon are internal hyperlinks, an error checking review, “semantic” searching, and more.

Oldstyle figures for better statute number legibility

I’ve been re-evaluating fonts for, and one issue that caught my eye is the style of the numerals. In running text, these proportional oldstyle numbers (font: Buenard) are perfect: they visually flow with the text. The wide variations in figure height and positioning help the reader unambiguously read the number.

But in a vertical navigation bar, I’ve chosen Georgia for its monospaced oldstyle numbers: the monospacing enables the reader to easily compare numbers while scanning vertically. Best of all, the font is already installed on all platforms.

It’s interesting that although the navbar numerals are in an entirely different font (Georgia vs. Buenard), the contrast is not jarring due to the sizes and positioning.

Research Trail

 Related Posts

Reference text web fonts: ITC vs. URW Baskerville vs. Buenard vs. Caslon

I spent a couple of hours today looking at web fonts for legible reference text. After reading the New York Times experiment finding that Baskerville is “the king of fonts”, I thought I’d give it a try, comparing it with a couple of others for legibility.

I first researched web fonts that I could test inexpensively. Then I tested them on a real web page. Here’s what I found. Currently, text (and nearly everything else) is Helvetica Neue. (All screenshots from a Mac. I’ll post IE tests soon.)

The sidebar is Trebuchet because I found that it’s more legible for small numbers. I like the way this looks, simply visually. But legibility-wise, I believe the text (“The presidential…”) is harder to scan with the eye than it ought to be.

Font of the Baskervilles?

While looking for Baskerville web fonts, I found Buenard via Google Web Fonts. It’s very close to Baskerville, and I find it amazingly legible:

Finally, all the fonts for comparison:

Helvetica Neue
ITC New Baskerville
URW++ Baskerville
Caslon 540

My take: In Buenard, the words hold together the best. The letter spacing is tight and the font is heavy. I feel like it’s super-easy to read. In comparison, in Caslon and Helvetica Neue, the words don’t hold together as well. The fonts look good, but for web text to be read on a screen, I think that Buenard is the best here.

Related Posts

Counting to 10 in Californian

I’m working on getting the California Codes online, focussing on the user’s experience: reading, searching, and accessing. This weekend I thought I had found an error in the numbering in the state’s online version, and so I headed to the law library to open a real book and see what’s going on. I was surprised to see that the printed text is the same. (West’s annotated and non-annotated editions.)

Here are three examples from the Code of Civil Procedure. The first, sections 676 et. seq. are the way I’d expect, in so-called natural sort order:

But here, a mathematical decimal ordering is in use:

Weirder still, both kinds are used in this group of codes:

If Only California Statutes had Names

I’ve run into the oddest problem as I add the California Codes to The Sections, the actual statutes themselves, are not given any kind of name as they are in other states. Here are the basic burglary statutes of Oregon and California:

Oregon: ORS 164.215, Burglary in the Second Degree.
California: Penal Code Section 459.

+1 for Oregon.

Oregon Burglary Statutes

These screenshots show one problem this creates. For starters, there is no reliable way to provide a good table of contents for a group of California Code sections.

Now, in the legislative business, these names are called leadlines, and they’re usually not part of the actual controlling law. But their usefulness should be obvious: leadlines help everyone write about, talk about, and research the law. And it turns out that these names are important enough that publishers like Westlaw and LexisNexis have created their own for their customers’ use.

California Burglary Statutes — would you have guessed?

From an economic perspective, California is not performing this part of the legislative process, delegating it instead to private companies. And so, the only people with access to this part of the code are those who pay for it. In other words, instead of the cost (of naming their statutes) being distributed across all tax payers, it’s paid by those acutely needing access to the law.

There’s an additional cost to this scheme that all Californians bear, however: inefficiencies resulting from a lack of standardized names. I mentioned that Westlaw and LexisNexis have leadlines for the code. They do differ, of course.

Caveat: I’m new to California law. Is there something I’m missing here? Let me know.

Making the California Laws Easy to Read

Screen Shot 2012-06-24 at 10.41.08 PMI’ve been doing a lot of work to get the California Codes online, and I now have something to show for it. This is a screenshot of Business & Professions Code Section 22947.4, an anti-spyware statute. (Which is pretty cool; I wonder if Oregon has something similar. Anyone know?)

For comparison, here’s the original code as its presented on the state’s website. I’ve been spending a lot of time on the fonts, white-space, outlining, and navigation. Ahead on my roadmap:

  • Print feature
  • Citations to sources
  • Legal news
  • Interlinked & related statutes
  • Smart search

This is Why We Need Universal Legal Citation

Courtney Minick nails it:

. . . if you want to cite to judicial law, you must pay to access the Reporter’s opinions.

Sounds crazy, but it’s true.

As a law student I wasn’t aware of the problem. But it frequently holds me back, now that I’m building online information systems. E.g., say some text on contains a citation to an opinion, such as

PGE v. Bureau of Labor and Industries, 317 Or 606, 859 P2d 1143 (1993)

. . . one of the most important Oregon cases; and of course law, owned by all of us.

To make the website more valuable to my visitors, I want to grab the text of that opinion and display it. Or, failing that, I’d like to simply hyperlink to it, online.

But neither of these are possible. The unique identifiers we’re provided to this public law: “317 Or 606”, “859 P2d 1143” — point to resources accessible only behind a pay-wall. If you’re skeptical, try googling for “859 P2d 1143” or “317 Or 606“. Zilch.