With the demise and pending demise of newspapers around the world, Straits Times’ Andy Ho’s frustration in his article “Two sides of the same coin” on 21 May 2009 is understandable.

In the article, he repeated Singleton’s rant against Google News for stealing news. This is a red herring. Google News does not even have advertisements on its pages. What it has are short snippets and links to news articles. Readers interested in the articles can click through directly to the pages on the underlying sites. In essence, Google News provides a directory or index to these news sites. It is difficult to see how Google is stealing contents and not providing a free and valuable service to both the readers and the news sites.

Mr Ho also erred in citing scientific publication as a model for the news industry. Little, if any, of the revenue generated from scientific journal subscriptions are used to support the research that generated materials for these journals. The scientific and medical journal publishers, including the largest one, Elsevier, had gotten rich off the backs of researchers who both provide papers for publications, and then pay exorbitantly for the privilege of reading the same papers. This lucrative business model did not prevent Elsevier from engaging in questionable business practices of abuse and misrepresentation of promotional material as peer-reviewed papers or journals. There are already movements to establish mechanisms for freer (in both freedom and cost) exchange of peer-reviewed papers. It is my expectation and hope that this would soon replace the existing scientific publication companies.

This is not to say that content should “droppeth as the gentle rain from heaven”, or that we should not reward serious journalism or good contents. However, we need to recognize that this is not a matter of fairness, but one of what would work in the new environment.

Robert Murdoch’s idea of locking up valuable contents for paying customers is unlikely to work. There are alternate sources of good contents from the masses (blogs, twitter, friendfeed, etc) and (this is the shocking bit) not necessarily from the established and sanctified news organizations or their reporters. Contents, being largely fixed-cost items, are expensive for a few but extremely cheap if amortized across many. Restricting contents’ distribution would diminish the contents’ total value, not increase it. As the music industry is discovering, creating artificial toll gates and charging entrance fees is not a viable long-term business.

History might provide some lessons as to what might work. Broadcast radio and television had been free for a number of decades, and are able to generate enough revenue to keep running. Scientific publication and quality research exist (even without companies like Elsevier) because it is widely recognized as a public good and so is funded publicly or through private donations. I do not mean that advertisements or public or private donations would definitely work for news, but these examples shows us that there are business models that could work.

Also frequently missing from similar discussions are data points on the costs for news organizations. How much of the costs are for news collection and content generation, and how much are for printing and distribution? How much of these costs can be effectively eliminated or reduced?

For all our sakes, I hope news organizations can find and adapt to good viable business models before they get overwhelmed by rising costs and reducing revenues. But it would be self-defeating if these organizations see the threat as only “other people” stealing their contents or otherwise refusing to pay for their services. Instead, they would do well to let go of the past and their longings for how-the-world-should-be, and deeply and fundamentally re-examine and restructure their business models.

In case you cannot read the warning hidden on the page, it say “Your browser does not support script” and “this site requires Flash Player Plugin installed and javascrpt enabled. Please enable your browser javascript and download Flash Player here.” Page display and navigation also not work properly.

1. Can anyone tell me why Flash is even needed?
2. Would it be so difficult to degrade gracefully when javascript is disabled?

I really expected better

 gem install activeresource --source http://gems.rubyonrails.org

First, he extrapolates Microsoft’s seriously flawed WGA outage on 24 August to the general vulnerability of all SaaS. Then, he spouts one of my favourite quotes all time on Google Docs:

“Easier to share files? So how hard is it to attach a doc file to an email anyway? Cripes.”

How can such a major technical writer be so clueless?

Pulling up the iFixit guide to MacBook Pro, I was ready to go:

1. Pick up hard disk. Being the greeedy sort, I opted for the Hitachi TravelStar 7K200, featuring 7200RPM (I like them fast) and 200GB (actually more like 200,000,000,000 bytes. Which is quite a bit smaller than 200GB). Sets me back S$390 at Funan.
2. Bought the Torx T6 screwdriver. You might think you already have some Torx screwdrivers but this is a really small one. S$15.70 at Sim Lim Square for a whole set, though I needed only one. And this is only for 2 screws on the MacBook Pro.
3. Back up disk using SuperDuper! to an external disk image. Made sure you eject the backup disk image before shutting off the external disk. When I forgot to do this, the whole image was not usable. Took me a few days to reconstruct my disk.
4. Followed the iFixit instructions, pretty straight forward except for the last part of jiggling the top cover off. Need to be patient with lots of jiggling to get it to loosen.
5. Reverse the process.

Except for the top cover, the whole process is quite doable. Needless to say, i) make sure you take out all the screws and done skip a step, ii) arrange the screws nicely so putting it back would be smooth.

Now the MBP is more responsive and the extra space is very nice. Well worth the upgrade.

One question though: what’s the best way to restore the hard drive? I ended up reinstalling the system, and copying the files from the mounted backup disk image.

– another advantage of Fat Models Thin Controllers.

Consider

class Box < ActiveRecord::Base
  has_many :marbles
end

class BoxesController < ActionController::Base
  def show
    @box = Box.find(params[:box_id])
    @box.marbles.find(params[:id)])
  end
end

To test this using rspec, you write something like

describe BoxesController, "handling GET /boxes/1/marbles/2"
  def do_get
    get :show, :box_id => 1, :id => 2
  end
  it "should get the right marble" do
    @marble = mock_model(Marble)
    @marbles = mock_model(Array)
    @marbles.should_receive(:find).with("2").and_return(@marble)
    @box = mock_model(Box, :marbles => @marbles)
    Box.should_receive(:find).with("1").and_return(@box)
    do_get
  end
end

I think the @marbles part in the controller specs is misplaced and ugly. This behavior should be pushed into the model specs. So lets refactor by combining the collection (@marbles) and find method into find_marble:

class Box < ActiveRecord::Base
  has_many :marbles
  def find_marble(id)
    marbles.find(id)
  end
end

class BoxesController < ActionController::Base
  def show
    @box = Box.find(params[:box_id])
    @box.find_marble(params[:id)]
  end
end

This makes the controller specs:

describe BoxesController, "handling GET /boxes/1/marbles/2"
  def do_get
    get :show, :box_id => 1, :id => 2
  end
  it "should get the right marble" do
    @marble = mock_model(Marble)
    @box = mock_model(Box)
    @box.should_receive(:find_marble).with("2").and_return(@marble)
    Box.should_receive(:find).with("1").and_return(@box)
    do_get
  end
end

and model specs code:

describe Box do
  it "should find_marble" do
    @marble = mock_model(Marble)
    @marbles = mock_model(Array)
    @marbles.should_receive(:find).with("1").and_return(@marble)
    @box = mock_model(Box, :marbles => @marbles)
    @box.find_marble("1").should == @marble
  end
end

I feel that this is much cleaner. What do you think?

The first program I’m playing with is Brain Age. This originates from Japan and works on a Nintendo DS. There is a test which scores your brain with an age. The closer you score to 20 years old, the better your brain is supposed to be functioning. It comes with a set of increasingly challenging brain training exercises. The program maintains interest by being slightly quirky and helping you track results and usage. My score had been up and down depending on which tests I draw, but it’s been quite fun trying to do better.

The second is Lumosity, a somewhat more serious looking site with various flash-based brain training programs. I’d been a beta tester, and when the site went live, I signed up for the subscription. The programs are categorised into Attention, Memory, Processing Speed and Cognitive Control, apparently each will improve your brain in the specified area.

Do these work? I can really test, but being a puzzle junkie, this is as much entertainment as training. They seem to be backed by real researches tho, so probably can’t hurt. Both are highly recommended.