adrian holovaty

September 9, 2002, 11:57 PM ET

Site review: ocregister.com

A few thoughts on ocregister.com, the Web site of the Orange County Register in California, which, as the splash page will tell you, is "NEW":

• Yes, there's a splash page, and it's kind of jarring. When it comes to important site announcements, I favor the approach recently taken by digmo.com and washingtonpost.com -- that is, putting a small notice on the home page that links to more information. (An example of the polar opposite is the Philippines' Freeman Online, which is offering no content at all during its redesign. I can't think of a better way to lose users.) Fortunately, in ocregister.com's case, you only have to see the splash page once; subsequent visits to the home page bypass the splash welcome, as long as you have cookies enabled.
• Articles aren't dated. This is a huge oversight. A news article without a date is like a road map that doesn't designate which city it portrays. In the site's defense, breaking news stories on the home page are given a timestamp.
• The 113-by-70-pixel boxes on the right of the home page look a lot like advertising, and for that reason I think a lot of users will ignore them. (When I viewed the home page this evening, they were links to "Humor" and "Entertainment".) From my experience, Web designers, particularly those with a print graphic design background, just plain love making these sorts of things. I suppose they're kind of pretty, but in reality I believe straight text would reach more people. Proof: Jakob Nielsen, Stanford-Poynter EyeTracking Online News study, various industry studies.
• A red "MORE" icon is scattered across the home page, in a seemingly random fashion. The "NATION AND WORLD" header has it, as do "LOCAL" and "SPORTS", but it's missing on some of the other headers -- most notably "TOP NEWS" -- which do have "more" pages but lack a "MORE" icon for some reason. If you're going to use a navigational helper like this, make it consistent. Otherwise, it's confusing. For example, a user might assume there's no more top news, because other sections have the icon but "TOP NEWS" doesn't.
• The customer service page has a bunch of links randomly scattered in a table. That thing is very difficult to scan. If you're looking for something in particular, you pretty much have to read every link; there's no hierarchy to help.
• Web pages like this one are what immediately came to my mind when I read today's Editor and Publisher article that pointed out how many sites still "shovel" content from the print edition online. The page in question, an event listing, is just a static file. Stick that stuff in a database! It's easy to do.
• Featuring a "Browse days" link right under the search box on every page is a very smart idea. The same users who will want to perform article searches will likely find some use in browsing previous issues. It's great to display these together.
• The search engine does some strange things. Most notably, it identifies the content-type of each search result -- for example, "text/html". It doesn't do users much good to know that an article is "text/html"; most Web pages are, anyway.
• Also, the search result list displays a snippet of each article found, but, for all the searches I did, the snippet was exactly the same thing: a bunch of navigation stuff like "Jobs Cars Homes". It's obvious this is completely useless and unhelpful, not to mention it's a waste of bandwidth.
• Also, the search results do not explain what the numbers in the brackets mean. I assume they reflect how many of the search terms were matched in each particular entry, but there's no clear explanation. And a minor nitpick that's probably just a matter of their programmers making a quick change: Each entry displays Y-m-d h:i:s instead of the actual timestamp.

Comments (4) / Permalink

September 9, 2002, 1:18 PM ET

Extra BBC syndication feeds won't work

A gentleman from the BBC e-mailed me regarding the "hidden" BBC syndication feeds I found last week. (That news I "broke" was blogged all over the place.) It turns out those feeds weren't intended for the public, won't be available for much longer, and weren't being updated anyway. Here's a snippet of the e-mail:

NewsOnline's public RSS service was only intended to include the following feeds

[front_page]
[technology]
[uk]
[world]

Your list includes a number which we had decided not to offer but had neglected to clean up properly from our website. None of these 'extra' feeds have been updated since Aug 22nd and we will be removing these from our website in the near future.

That's a pity; the extra feeds were great. But there's an interesting lesson here: Anything you put online, regardless of whether its URL is public, is fair game. Anybody can type a URL in and try to guess at patterns and conventions -- that's what I did with the BBC feeds. Site managers, save yourself some trouble and keep sensitive stuff to yourself in the first place.

I think this is something news sites in particular should bear in mind, knowing how common it is for sites to put advanced versions of stories online -- like Scripps Howard foolishly did with its Ronald Reagan obituary, which is still available via the WayBack Machine. In creating or testing anything that's not intended for public view, it's best to use a non-public server.

Comments (1) / Permalink

September 9, 2002, 12:53 PM ET

Monday's recommended links

Editor and Publisher: As 2003 Nears, Where Is Online Newspaper Biz? In the gutter, according to an anonymous source quoted in the article: "Most media have stopped talking big talk and launching big projects and are making do, for the most part, with just slapping their print content online."

Jared Spool: Evolution Trumps Usability Guidelines -- A practical alternative to following Web usability guidelines, such as "Always put your search box in the upper-right corner", blindly. (Link from Small Initiatives)

wannaBrowser lets you see what happens when you visit a site with a user-agent string of your choosing. (Your user-agent string is an internal description of the browser you're using; sites you visit sometimes will perform certain actions based on this information.) Some sites ban particular user-agent strings, such as those of harmful spam robots. For example, WebmasterWorld bans the user-agent string EmailSiphon, which searches Web pages for e-mail addresses. On wannaBrowser, you can see this for yourself by entering http://www.webmasterworld.com for the location and EmailSiphon for the user-agent. This is useful for testing, if you're setting up your server to ban particular user-agents (as discussed lately on WebmasterWorld).

Comments (1) / Permalink