What I needed was something like Merb routing in Rails…

What I found was literally Merb routing in Rails.

It was a perfect fit and an easy win because Merb’s routing just works more like routing should work.  In rough benchmarks it was just as fast, too.

I don’t brag on projects very often, but this was a damn nice piece of work.  It comes as a Rails plugin.  You have your choice of forks on GitHub.  I started with drogus because it came Rails 2.3.x ready.

Most existing Rails routes can be rewritten in short order.  You need to rewrite them because Rails routes and Merb routes don’t coexist.

To install and use:

1. download and install as a plugin
2. create config/router.rb

Things you can do (put the following in router.rb):

Merb::Router.prepare do |r|

  # yeah yeah
  r.match("/posts").
    to(:controller => "posts", :action => "index")

  # the simple stuff
  r.match("/posts/:id").
    to(:controller => "posts", :action => "show")

  # regex FTW...
  # matches "/foo-bar/sys/bang/12-4-6" and gives us:
  # {"sys_id"=>"foo-bar", "knob_ids"=>"1246",
  # "action"=>"frobnicate", "controller"=>"systems"}
  r.match(%r[^/([a-z-]+)/sys/bang/(\d+)-(\d+)-(\d+)$]).
    to(:controller => "systems", :action => "frobnicate",
      :knob_ids => "[2][3][4]", :sys_id => "[1]")

  # matches "/this-is-the-slug/12-3-45" and kindly gives us:
  # {"a"=>"12", "b"=>"3", "c"=>"45", "action"=>"show",
  # "controller"=>"categories", "id"=>"this-is-the-slug"}
  r.match("/:id/:a-:b-:c").
    to(:controller => "categories", :action => "show")

  # does what you'd expect
  r.match("/:controller/:action/:id").
    to()

end

Props to the originators. They literally made my day.

While writing the original Magic Sac, the time came for me to code the keyboard & mouse driver. This was really hard, and I struggled over the design for a long time. Problem was, on the Mac, the mouse was run directly by the 68000 via interrupts, through the SCC chip. On the ST, the microprocessor, which periodically interrupts the 68000 to tell it where the mouse moved to. Also, on both units, the keyboard was run by a standalone microprocessor, and the codes returned were totally different. It was all totally different.

I dreaded doing it. I saw problems on top of the problems. Interrupt conflicts, mouse resolution, keyboard mapping… Finally, in desperation, I reached far back into my bag of programmer’s tricks for the old axiom: no pain, no gain.

I extended this to: If pain, gain. Well, it’s logical, right?

So I looked through my record collection for something that would be true pain. And I found it: Neil Young, live in concert. Off key, trembly voice, songs whose lyrics would depress a hyena.

I put on the record player on automatic repeat, turned it on, and set to coding. Six hours and 1,500 lines of code later, I was done. I tried assembling it for the first time. No errors.

So I ran it. The keyboard worked; I typed “a”, and the Mac knew I typed “a”. I was really impressed. so I got ready to debug the mouse code (which I was really dreading); I moved the mouse to make room for a legal pad to take notes on… and the mouse pointer onscreen tracked with the mouse as I moved it.

I played with the mouse in utter amazement for a few minutes. It worked.

This is the first piece of programming I have ever written that worked the first time. As any programmer will tell you, this doesn’t ever happen; it did for me, just this once.

I attribute it all to Neil. So, these days, I keep in reserve a cassette of Live Rust (no kidding, that’s the name), and when the programming gets tough, Neil gets going.

- Dave Small

Yahoo’s announcement that it will be closing GeoCities raises concerns about the permanence and persistence of the information we entrust to others.  Their announcement follows other closings–Digital Railroad and Pownce, to mention a couple.

GeoCities has been around forever (figuratively, but nearly true in Internet years) and its product is web-pages–a base pair in the DNA of the Internet/Web.  GeoCities’s disappearance later this year begs the question, what else could be at risk once enough time goes by and a company decides it can no longer afford to foot the bill for your data…

Google’s sequential quarters of declining revenue coupled with the observation that e-mail is not the communication medium of choice for Millennials and younger, make me wonder if/when Google will become responsible for stewarding over a huge chunk of data with very low utility.

Twitter still hasn’t figure out how to make money and they’re $35M into investor’s money.  A big bailout in the form of acquisition is almost certainly their strategy now; but Google’s down, Microsoft’s down, Yahoo’s down (clearly), and Facebook has valuation problems of its own.  Old-timers (early adopters) have years of their history stored in the Twitter data cloud.

MySpace hasn’t played out as expected for News Corporation and their deal with Google ends soon.  What happens when Rupert Murdoch decides to jettison and underperforming property?

What about Flickr, YouTube, Mint…?

Luckily the bit bucket is linked to /dev/null (never gets full).

Then we all started worrying about (in rough chronological order) tweets per day, number of followers, ratio of number of followers to number of friends, sufficient @ replies, retweetability, tweet reach. All we need to do is measure follower turnover rate, avatar attractiveness, and penis length to round out the metrics suite (oh, yeah… we’ve now got penis length).

Twitter pundits (AKA media whores) now write articles on how to grow your Twitter follower base (as if the fast follow/unfollow, rinse, repeat wasn’t all you need).

My little status updates (”dodging raindrops..“) look pretty paltry in comparison to all of this innovation.

Twitter is interesting.  But more interesting than Twitter is the vast ongoing social experiment on the use and abuse of Twitter. Psychologists and anthropologists should be having a field day.

If anyone is interested, I’m starting a Google group on Twitter social metric hacking…

- Todd

Not anymore, anyway. A good friend of mine in the advertising industry pointed this out to me recently.

Today, your brand is the aggregation of a thousand points of view. With regard to consumer goods and services this has long been the case, but until recently it wasn’t possible to really sense these points of view, so we confuse brand with artifacts: full-page advertisements, Super Bowl commercials, and billboards. After all, in its original use, a brand was a symbol literally burned into an object.

So, a while back I started an experiment. On all of the forms and profiles that I complete online, I’m supplying a Google query as my home page. Rather than attempt to tell you who I am by pointing you at a carefully crafted web-site, I’m using Google’s page rank, and the links you’ve created to my artifacts, to tell you who these thousand points of view think I am.

I realize, I open myself up to “miserable failure” style manipulation of the system, but then, isn’t that in its own way reflective of the brand?

Todd AKA http://www.google.com/search?q=todd+sundsted

Specifically, I’m tracking tweets that mention “coworking”.  The last few weeks have been incredible.  Take a look at the numbers (by week and by day).

I use Pwnserver to gather data from Twitter and a few handcrafted tools to digest the data. Data updates happen once an hour.

…how we live our lives can’t be separated from what we make. It is all in there…

Pwnserver isn’t a brute force, bulk download tool.  It you want to hammer Twitter and suck down every tweet as fast as you can, you are better off rolling your own tool.  Pwnserver is designed to digest bite-sized chunks of data over a span of time, to avoid drawing attention to itself.  It’s also designed to reliably keep on top of new data while simultaneously working backward in time to earlier data.

Get it here…

Control over unique, hard-to-recreate data sources that get richer as more people use them…

[ Tim O'Reilly, Core Competencies of Web 2.0 Companies, What Is Web 2.0 ]

Tim made a fortune publishing books–many of them about Free or Open Source Software.

He also articulated a set of characteristics (core competencies) common to some of the most successful technology companies since the dot com bust–the so-called Web 2.0 companies.  Think of Facebook.  Think of Google, for that matter.

In early 2008, Wired Magazine published an article by Chris Anderson titled Free! Why $0.00 Is the Future of Business.  Chris used “free” in the sense of “free beer” (no monetary cost).  Free products and services are often paid for by third parties–parties who are willing to pay to participate in an exchange taking place between two other partners.  Advertising is a specific case of this three-way marketplace.

Of course, these third parties are only willing to pay if there’s no other way to play.  This is why control is important.  Control creates scarcity.  Scarcity creates a perception of value.  Commerce ensues.

In the last few years, we entrusted a tremendous amount of personal data to Web 2.0 companies.  True, in many cases it cost us nothing to post/store/share this data.  However, thanks to Tim, at the heart of every Web 2.0 business plan and every Web 2.0 private equity investment is the idea that control of a hard-to-recreate data source is valuable.

In 1985, Richard Stallman published the GNU Manifesto.  Like Chris, Richard believed that free was important; however Richard was concerned about “free” in the sense of “free speech” (unfettered or unrestrained).  The GNU Manifesto put forth the radical idea that software should be free.  Users should be free to read, modify, and share software source code.

This kind of freedom is powerful–I’m writing this post on a laptop that runs entirely on Free Software–but it’s incompatible with the kind of control embodied and embraced by the Web 2.0 world.  This is ironic, because a great amount of Web 2.0 software is powered by Free Software.  It is also disappointing, because Tim has to know that sharing, not control, is at the core of Free Software movement.

It’s time to learn a lesson from history, to look back on the incredible value that was created by and with Free Software over the last fifteen years, and to free our data so that we can see an explosion of innovation down the road.

I would be willing to pay to use my favorite Web 2.0 applications, but give me an API and easy access to my data.  I’m unwilling to trade one kind of free (freedom) for another kind of free (monetary).

The price is too great.