Printed Matters

What’s the difference between a blog post, a forum post, a news story, a comment, an event listing, and a business listing? Not much, as far as your database and your app are concerned. They all share similar attributes, like title, body, author, time published, time edited etc. and so it might make sense to put them in a database together. I’m thinking, of course, about sites that might have all those types of things, like a community newspaper site.

If you did that then you could easily build new types of applications within your site just by defining a new category and (optionally) creating a couple of new templates.

Advantages: simplifies search, speeds coding time (since one type of query would work for all content areas).

Default template: display title, body (if any) and then list any children in a dl (title, snippet) format. From there, you can define special list, list item, and single item formats that each child member can inherit.

Wordpress does something like this, and I was pretty close to doing exactly this on ParrySound.com.

Ok, back to coding.

Ok, I tried Disqus and didn’t like it because it wouldn’t let me change my display name, and wouldn’t let me change my comment system without losing all my (myriads of) comments, and didn’t do trackbacks. So I’m trying Intense Debate, which looks a little slicker. Not a lot of blogs supporting it yet, and they don’t appear to handle trackbacks either, but it’s better than a system that you can’t personalize and that you’re stuck with.

One thing I really don’t like is that all your comments are in JavaScript, which means that links back to your own blog won’t count in Google. Hmmm.

A couple of recent articles in EditorsWebLog have me thinking about advertising again. Newspaper execs are having a tough time figuring out how to monetize their websites, while advertisers are concerned about how to measure audience on websites.

And I think interstitials deserve another look.

Interstitials are ads that take up the full screen as you move from one page to another on a website. Here’s the Wikipedia entry. And here’s a better explanation, complete with an example.

I think we all know the problems with standard banner advertising, those flashing, blinking rectangles of annoyance that occupy the margins and header areas of webpages. People have trained themselves not to look at them. There’s evidence that people don’t even see them. Sufficiently irritated people can get programs or browser plug-ins or mini-scripts to block them entirely. Click-through rates, about the only measurement we have to determine whether people are seeing the ads, are dismally low and at any rate can be gamed.

Pop-ups/unders/overs have similar problems and are even more annoying. Modern browsers come stock ready to block these.

Yet, a content provider has to make money from the content it provides somehow. And when it comes right down to it, there are really only two ways to make money from content: the consumer pays to get content (e.g. subscriptions, box sales, pay-per-view) or the content is supported by advertisers (e.g. sponsorships, banner ads, flyers, etc).

I think, and the consensus today among publishers is, that advertising is the way to go on the web. Pay walls go against the spirit of the web and reduce its utility. Besides, big media outlets like the New York Times have found that the money they can make from advertising far outweighs the money they can make from charging for their content, prompting them to take down their pay walls. The traffic that comes into their archives from search engines like Google is massive, and also, nobody’s going to pay for content (news, in this case) that they can get elsewhere for free.

But if advertising is the way to go, and traditional web advertising performs poorly, where does that leave content providers?

I’d like to make a case for interstitials. Cookie-based, server-side interstitials. Why? They can’t be skipped unless the user turns off cookies. You can limit how often a user gets hit with an interstitial – say, every 10 minutes, as the user moves from page to page, he gets hit with an interstitial. He’s forced to look at it for, say, 10 seconds, before he’s allowed to move on to the content he wanted to get to.

If the user dumps his cookies, then he just gets hit faster…he moves to one page and then when he tries to move to the next page: an ad. Some users may turn off cookies. In that case, you can also turn off certain premium content, like videos. Those are expensive to make, and don’t do anything for Google traffic anyway, so if cookies are off (like a search engine spider) then the browser doesn’t get to see video. They also can’t log on, for example, so depending on how you have comments set up, they won’t be able to comment or participate on the site. I’m sure there are other ways that users could be gently punished for turning cookies off.

But why would the user do that? Surely the user can stand looking at an ad for 10 seconds every 10 minutes? There’s a much bigger ratio of ads to content on TV and yet people have watched TV under that regime for many years.

The advantage to this scheme is that you know for sure users are seeing the ads. At first, they may use their 10 seconds looking frantically for some way to skip the ad or avoid the horrible punishment of looking at the ad. After a while, they’d get used to it, and see it as an acceptable compromise, I think. Somehow the content has to get paid for, and this it is really not much to ask a user to contribute less than two per cent of his time to the cause. Acceptance levels would be higher, of course, if media companies got together and standardized the details.

The numbers can be adjusted, of course. Maybe the ratio should be one per cent or five per cent. Maybe five seconds is enough time to see the ad, and maybe it should be served every five minutes. Testing will sort all that out.

You can get ad-blockers that work by domain name, so that stuff from doubleclick (for example) can be filtered out. In that case, the user would be left staring at a blank screen. So maybe ads should be served from your own domain name. If you belong to a network, maybe they have to push the ads to your server somehow.

Note that the “one-hit wonders”, people who come in from Google to read your stuff and then are off again, don’t get hit with an ad. You know what? That’s OK. They probably don’t care about your message anyway.

I’d really like to hear some feedback and comments on this one.

I just want to draw some attention to what I think are some cool ideas being bandied about on news website design by some Dutch developers. I learned of it today via my feed from Poynter’s E-Media Tidbits. Take a look at www.en.nl. It’s in Dutch, but I was able to figure out what was going on with the help of Babelfish.

The first thing I noticed about the site was that there are no menus. The space normally reserved for a menu of categories – news, sports, entertainment, etc. – is instead dedicated to a list of stories. By default, when you arrive at the site, the list consists of all the stories that were posted on the site in the last hour. But you can see what happened in previous hours by mousing over the bar graph at the top of the page, which interaction designer Wilbert Baan calls a newsriver.

The front page has exactly one story on it – the last one. There is no package of news like on a traditional news site, where they give you the five top headlines, some sports, some entertainment, and 300 or so links to articles and features inside the site (which I whined about recently in my reviews of local Toronto newspaper sites). None of that. Just one story, the newsriver, a textbox where you can tag the story, a search box, a place where you can rate the story, a place where you can comment on the story, some social bookmarking tools, and some lists of the most popular stories by importance, comments, and views.

Missing entirely are old-school concepts of packaging. Gone are the news, sports, and entertainment sections. There’s just the flow of news and some (for lack of a better term) Web 2.0 ways of interacting with the news.

This strikes me as an embrace of the unbundling that Nick Carr was talking about on Britannica Blog. They’re not sorting the news into any kind of pre-ordained canisters that reflect the way newspapers break things into sections.

Instead, you tag stories yourself. You don’t even have to be logged in on the site to add tags to stories. I tagged a story about swimming as “swimming” (in English). Once the story is tagged, you can click the resulting link to find other stories that other people have tagged the same way.

Unfortunately, the search function was broken when I visited just now. But I’m sure it would search among the user-defined tags as well as the article texts to generate results.

Also interesting is that they let you see different versions of the same story. By default you see the latest version, but you can see the original version too as it came in (presumably from a wire feed or some such).

This represents a new way to think about news. No pre-packaging. Several different ways for readers to get involved (by ranking, commenting, tagging, sharing, and even, if I understand it correctly, changing the actual articles in Wiki fashion). Emphasis on the flow of news rather than on the editorial judgment of site operators. Emphasis on the stories themselves, and their metadata, instead of on branding and sections and categories.

I agree with Paul Bradshaw at Poynter: this is one to watch.

I just have two complaints, one minor, one more important.

The minor thing is that I don’t like the placement of the search box. I can’t see it down there under the tagging text box. If I were a regular user, I know I’d eventually tag an article with something I meant to search for. They should put it somewhere top right, where people are used to looking for it.

The other complaint is that I can’t easily see how to go back into archives. I can see what stories were posted in the last 24 hours….what about before that? Or am I still thinking old-school?

Update: I see in this comment by Wilbert Baan that the original articles are in fact coming in on a feed – the ANP, a Dutch press agency. I wondered what all those ANP logos were about

Over at Brittanica Blog they’re having a big forum this week called “Newspapers and the Net,” all about the state of newspapers in the digital age. This morning, Nick Carr and Clay Shirky weighed in, talking about the new economic model: the shift from scarcity to abundance, the unbundling of the traditional newspaper model, and the loss of distribution as a service worth paying for.

They’re interesting and well worth the read – but, I must say, a little too hysterical.

They’re both lamenting the loss of good (read: expensive) investigative journalism. How does that work? Because they think that the unbundling means that each story is a stand-alone business. As Carr puts it, “Each story becomes a separate product standing naked in the marketplace. It lives or dies on its own economic merits.”

A story on depression might do well as it attracts ads from drug companies, and a story on RRSPs might do well as it attracts ads from investment companies or banks. But a story on government corruption might not do so well, even if it attracts a large readership, since it won’t simultaneously attract expensive advertising.

Therefore, since investigative journalism must stand alone and won’t do very well by itself, there’s no money to pay investigative journalists and the trade will die off. And then governments, unfettered by their traditional watchdogs, will run amok. And that’s a bad thing, by any standard.

That’s the argument, in a nutshell. Various statistics are trotted out supporting the doom and gloom: newspaper readership has been shrinking year over year since 1984, and the shrinkage has dramatically accelerated in the last couple of years. Only half of American adults read a daily newspaper in 2006, while half of adults under 36 got their news online. The NAA reported recently that newspaper revenues took the sharpest decline in half a century last year. And so on.

Nobody doubts that the industry is in turmoil, that readers are moving online, and that online revenues are not robust enough to replace falling print revenues. Newspapers, and media companies in general, need to hustle to figure this out. But to argue for the eventual death of investigative journalism seems to me a little over the top.

That argument depends on this whole notion of unbundling, so that pieces of investigative journalism would have to stand on their own, “consuming costs and generating revenues in isolation.” So let’s take a closer look at this notion of unbundling.

Carr says that online, the bundle falls apart, because people don’t read online the same way they read the newspaper bundle. They usually won’t even come through the front page of your site – they come to a specific story page via Google, or their feed reader, or Digg, or Twitter, or what have you. Then they go away again, to a different news report or blog on that subject. They may not even be aware which news organization’s site they were on.

The idea of a news report standing completely on its own can be parodied by imagining that every news story was published under its own domain name – like HarperVisitsAuschwitz.com – with no branding, no outside links, except maybe to the reporter’s news organization, and some ads. Now the people responsible for selling those ads go to advertisers and say, look, we have this story on Harper visiting Auschwitz, you want to put an ad on it?

That is not how online advertising is sold, now or in the future. And it’s not how stories are put online, today or in the future. Advertising is sold onto sites, whole sites, and they do well or don’t do well depending on the overall traffic that site receives. The bundle is not completely gone.

Yes, some services from the traditional newspaper, like classifieds and real estate and TV listings have gone away to be aggregated in online-only services like Craigslist – where, quite frankly, it makes a hell of a lot more sense both from the advertiser’s and the consumer’s point of view. And news aggregators make it easier for readers to search for just the kinds of news they like to read, from whatever sources are available.

But the aggregators need to get their news from somewhere, and that somewhere is, and will likely remain, the websites of traditional news companies. Yes, the dreaded MSM. (UPDATE: this comment on Jeff Jarvis’s blog post on the same issue got me thinking: it won’t be the MSMs at all, rather, it will be from a handful of press agencies like CP, AP or Reuters)

And an online news story is not completely stand-alone in two other important ways. First, when the fickle reader arrives on a story page from Google, he can be enticed to look at other pages on the site. Related stories, Most Popular/Read/Emailed/Commented widgets, related multimedia, readers’ comments: all these serve as very effective advertising for other parts of a news org’s site. You CAN get people, even people coming in from their feed readers, to look at more pages. And then these page views get pooled into your traffic (also called readership) which you then sell to advertisers.

The second way is that you can develop community around the traditional content your site puts out. Community papers in particular are in an excellent position to dominate the local web space in their respective communities. I’ve talked about this before.

In summary, I think it’s a little early to be sounding the alarm on the death of investigative journalism, or of the media organizations capable of supporting investigative journalism. While I agree that now is, in fact, the time to panic – and to experiment relentlessly – we have no reason (yet) to believe that the robust companies that have developed around the gathering and distribution of news will not be able to adapt to the new economic models being forced on the industry by the Internet.

This is the last in a series of posts all this week taking a hard look at the websites of the four major Toronto dailies. Now we summarize.

Here’s a menu of the posts covering each individual paper:

1. The Toronto Star: D-
2. The Globe and Mail: C
3. The National Post: D
4. The Toronto Sun: F

None of the sites of the four major Toronto dailies are excellent, for various reasons. The Globe fares the best because it’s the only one that allows comments, but it doesn’t get an A because of its egregious pay wall, and the front page link bombardment.

It might be useful to summarize things in a chart:

This is the last in a series of posts all this week taking a hard look at the websites of the four major Toronto dailies. Today we look at canoe.ca’s flagship, the Toronto Sun.

The Toronto Sun’s website gets the Most Ugly award. Basically, it’s a set of rectangular blocks stacked one upon the other, but they don’t even fit together! The content area is perceptibly and irritatingly wider than the header and ad blocks above it. The whole thing gives the impression of a ziggurat.

And look at that image-based menu where one tab, the Services tab, looks like its raised (as in that’s the section you’re in) but you can’t even click on it. So 90’s.

And the huge area of white space near the bottom is appealing…not.

But, we’re not here to be design critics, which is a subjective art. On the good side, the Sun’s home page has the fewest links (120, which is almost in the realm of usable) of any of our four study subjects, and the second-fewest validation errors (only 163). At least they didn’t have the hubris to put a fancy XHTML DocType at the top, and then not bother following it. In fact they have no DocType at all.

Onward. The site is part of the canoe.ca network and appears to be running on a ColdFusion platform.

Full stop: no RSS feeds, and no story comments. Wrong century, guys. You fail.

Overall mark: F.

There are two nasty, ignorant issues taking shape around how you use the Internet.

The New York Times Bits blog yesterday reported on the progress of two companies who want to install devices at your ISP to record everything you do. The idea is to build a profile around your computer’s IP address so that they can show you targeted advertising. They say this is actually less privacy-trampling than what Google or Yahoo are doing because they don’t store any actual log files about your searches and which web pages you’re viewing, they simply use that information to build out your profile and then throw the rest away.

I’m the last guy to want to come down hard on companies who are trying to figure out how to do advertising and make money on the Internet. These things have to be figured out, and soon, before the painful transition from print to web starts actually killing newspapers, for example. But, boys and girls, read my lips: spying on me not the way to do it.

We’ve always had contextual advertising. It’s a sensible thing to do. If I’m watching hockey on TV, then you can guess that I’m a guy and you should try to sell me beer and running shoes. If I’m watching a gardening show, makes sense to try to sell me gardening tools and contracting services.

Let’s be clear: what Phorm and NebuAd are doing is not contextual advertising. It’s an extremely invasive form of profiling with associated targeted advertising. If it is done here (they’re starting trials in Britain and the States), then it will be done without your consent and quite likely without your knowledge. There will not be, as far as I can tell, a way to opt out.

They say that they will never associate the profiles with anyone’s real name and address. That’s not comforting. As soon as the ISP gets a subpoena, there goes that promise and any promise like it. And unhackable? Such a thing does not exist.

They say that people – you and me, the hapless surfers – will like it because the ads we get served will be more relevant and useful to us. The counterexamples to THAT theory are mind-boggling.

• Say you are a problem gambler who has difficulty staying away from online poker sites. Do you need an ad for BigPokerNet?
• An example from the comments at the Bits blog: say you’re having marriage troubles and you’ve been looking for divorce help. Do you need your kids being bombarded with ads for divorce lawyers?
• Suppose you’ve been researching a digital camera purchase, and have gone ahead and made a purchase. What do you need with more camera ads?

A good analogy to something in the physical world: suppose the people at 407ETR decided they would use that transponder thingy they give you to track where you enter and exit their highway to track which stores you drive to. Then they flash up targeted advertising on billboards as you drive their highway.

No thanks, get out of my face! Jesus, I didn’t even remember parking in front of the gay porn shop that day!

This is an appalling invasion of privacy and the mere fact that ISPs are teaming up with this nonsense shows that they care not one bit about their users’ privacy. I’d suggest, if they really think people will want this enhanced service, that they trying making it opt-in and see how many people actually opt in. Personally, I can’t imagine saying, “Yeah, I really want better ads…sign me up to trade in my privacy for that!”

Now today, we learn that smaller ISPs who depend on Bell for upstream bandwidth are taking Bell Canada to court to make the telecom giant stop “shaping” the traffic it sells to them. Bell, and the other giant ISP Rogers, have been using techniques which throttle certain types of traffic, namely P2P and torrent traffic used to transfer large files.

Smaller ISPs are saying that when they buy wholesale traffic from Bell, they should be allowed to use it and shape it the way they want, not the way Bell wants. They say it interferes with their business if they sell a customer 5Mbs of download speed and the customer only gets 3Mbs when he’s downloading movies and music and other big files. And, seriously, when else are you going to notice it?

The reason people get high-speed is to download movies and music and other big files. If you just wanted to surf web pages, and you didn’t want to tie up your phone, you’d just get one of the Lite offerings. You’d rarely notice the difference. No, you get high speed specifically for the types of services the ISPs are now throttling – downloading big files at ripping high speed. And you PAY EXTRA for that.

It amounts to a form of censorship. The way this throttling is done is through port numbers. So, for example, they could let all the traffic on the ports usually reserved for HTTP (normal web traffic), FTP, e-mail, and a few select others come through unfettered, and choke down everything else. So, if I’m downloading a big disk image from RedHat over HTTP, I’m good, but if I do it over P2P, I’m bad.

Who are they to tell me how to use the bandwidth I’ve paid for? Why don’t we just run everything on port 80 then?

Now of course they can’t let the sub-ISPs have unthrottled bandwidth, because then everyone would flock to Bell’s competition so they can have an unthrottled, non-judgmental service. That wouldn’t be good, if the people buying wholesale from Bell actually gave a better service than Bell.

The whole affair appears to be forcing Bell to introduce a pay-for-use model, so you might have to pay extra if you go over a certain amount of usage each month. That’s just like the 407 people charging you for how far you drove your car on their highway. That makes sense. Rogers has already introduced their plan.

It sure doesn’t make sense to pay extra for higher bandwidth and then be throttled down when I want to use it.

ISPs can’t just arbitrarily invade my privacy and censor my Internet usage. I won’t have it!

This is the third in a series of posts all this week taking a hard look at the websites of the major Toronto dailies. Today we look at canada.com’s flagship, the National Post.

There is really not much to be said here that hasn’t already been said. The Post has contextual menus, good story tools, “most popular” widgets, and excellent RSS feeds broken up very finely into categories. On the ugly side, it has the most front page links of any of the four papers we’re looking at (382), and also the most validation errors (317). That’s a mess.

The site has a couple of things that stand out for me: an awesome Flash-based video browser (that takes a bit too long to load) and an excellent (and prominent) set of blogs.

I’m starting to think a summary of the features and drawbacks of all four sites in chart form might be most useful.

Overall mark: D.

This is the second in a series of posts all this week taking a hard look at the websites of the major Toronto dailies. Today we look at Canada’s Gray Lady, the Globe and Mail.

The Globe’s site is built on a Java-based platform that is, as far as I can tell, a home-rolled solution. According to some digging in the Internet Archives, they started using this platform sometime between April 1999 and August 2000.

The Good

1. Nice big pictures – when you click a thumbnail embedded in a story on the Globe’s site, you pop up a window with a nice big version of the pic, usually 500px wide, but often as large as 800px wide. The photographs are very well done.
2. Contextual Menus – as the Star does, they have a main menu that stays the same on each page, and a secondary menu that changes as you enter each section. Again, works for me.
3. RSS Feeds – again, like the Star, they have categorized RSS feeds so you can subscribe to Technology feeds but not, let’s say, Sports.
4. Article Tools – nice complete set of things you can do with each article: print it, share it on social media sites, email it, and, importantly, comment on it!
5. The Day in Pictures – Damn, I love that little multimedia widget they have in the right sidebar. the actual photo galleries aren’t as snazzy as the Flash-driven galleries at CBC (here’s an example for Earth Hour) but the pictures are great and it does the job.
6. Comments – Ah, lovely comments. Check this story about Bob Rae, which has 402 comments right now and climbing. The Globe is the ONLY Toronto daily that allows comments on almost every story they put on the site.
7. The Popularity Widget – It shows up on every page, and it has three tabs – Most Viewed, Most E-mailed, and Most Discussed, which tell me at a glance what people are interested in and thinking about.

The Bad

1. Crazy URLs - The Globe uses some crazy-long URLs, which don’t let users use the URL like a UI, and which interfere somewhat with the web’s social navigation. Check this one, picked randomly from the home page: http://www.theglobeandmail.com/servlet/story/RTGAM.20080401.wfrance0401/BNStory/International/home. 99 chars, folks, enough to cause display issues just about anywhere.
2. Home page Link-fest – like the Star, the Globe seems to think it has to show you every single thing you can do on the site on the home page. 375 links on that sucker today. Where do you start? Beats me – probably just the top headlines and the menu is all you really need.
3. Irritating Ads – I need to get my mouse up to the menu…to do that I need to cross over that Rogers ad…oh, no, it’s opened up and covered up the thing I was trying to click on. Jesus, it won’t go away! Ok, sorry, sorry, I won’t try that again.

The Ugly

1. Pay Wall – Only one thing in the Globe’s ugly category, but it’s a doozie: the god damned pay wall. No, I will not get a subscription so I can see Jeffery Simpson’s column. Since I don’t get the paper, I have no way to know whether I want to see Jeffery Simpson’s columns. So stop putting them in my RSS feeds, then saying “Nope, Globe Insiders Only” when I click on it.

The Globe and Mail fares pretty well. The annoying pay wall only affects certain columnists and other material that they deem too good for us Internet users. Overall mark: C.