How early online newspaper production tools led the industry down the wrong path

Wisdom is the ability to see your life and career not simply as a line going forward from wherever point you are, but as an arc that extends from the past into the future. That’s why I believe it is important to teach online journalism students about the history and development of the Internet and for online news professionals to remember the early days of their craft. (It’s also why I find books like Malcolm Gladwell’s “Outliers” so interesting.)

I’ve written about how legal precedents shaped the thinking of early online news managers. Today, I’d like to suggest that early online publishing technology affected industry thinking in profound, and, ultimately, tragic, ways as well.

For those of you who weren’t working on a newspaper website around 1996, let me take you on a trip into the pensieve (or, down memory lane, for those of you overdosed on Harry Potter references this week). I started on the Rocky Mountain News website in November 1996, and was the only person at the paper updating and maintaining the news side of the website. Every morning, I came in around 5 am, selected a couple dozen stories from the newspaper, then called them up on the paper’s ATEX terminals. One by one, I sent a copy of each story to a queue we’d created which interfaced with the Pantheon Bridge program on a Windows NT box in the paper’s computer room.

Pantheon was a set of programs used by many newspapers at the time to port copy from the papers’ publishing systems into flat HTML files. One by one, I’d call up each story in Pantheon, to make sure that it had come over and then to assign the story to the appropriate section in which it would appear on the website.

Pantheon built index pages for each section, in addition to create an HTML page for each article. To do those things, it had to read the head, deck, byline, publishing date and story copy into fields in an Access (!) database, from which it would push each article into page templates. (Heaven forbid that anyone on the copy desk had decided to use a different way to mark up a head or byline, because that would break the parsing process.)

I pulled whatever photos I needed from the paper’s photo server (on a Mac) and Photoshopped them to the specs I needed. After that, I fired up Notepad to hand-code the paper’s front page.

Finally, I would use two FTP programs to manually transfer each JPG image and story and index file to the Rocky’s webservers, then at Scripps headquarters in Cincinnati (Fetch for the images from the Mac, and WS_FTP for the HTML files from the Wintel box).

When I moved to the Los Angeles Times website in January 2000, I was delighted to find that the Times staff (which numbered in the dozens) had written a series of scripts to move every article and some images from the paper’s print publishing systems onto the Web. But human staff needed to check the feed every morning, to see that it had come through uncorrupted. Several mornings, it hadn’t, and tech staff needed to debug and restart the feed.

Still, Times online editors hard-coded most top stories in HTML, manually editing images in Photoshop and building index pages by hand.

By today’s standards, the work was ugly and mundane. But it had to be done. Online readers wanted to see the newspaper online. They wanted the freedom to read the paper on their computers at work, so that they could hit the road from home a few minutes earlier each morning.

On the advertising side, automated scripts at both papers helped bring classified ads online. And at the Rocky, inside sales reps “upsold” classified advertisers to put their ads online. At the Rocky and at many papers, that incremental revenue from upsold classified ads paid for the online production staff, in both editorial and advertising.

Why does this matter now? Shouldn’t those of us who remain at newspaper websites just be thankful that we don’t have to go through that hassle to get our sites online everyday?

Let’s remember that arc, though, and how what happened then has shaped what is happening in the industry now.

I believe that the hoops we had to jump through to get newspaper stories online influenced newspaper managers’ perceptions about the difficulty of online publishing. Sure, many of us had personal websites and knew how it easy it was to slap together a page in HTML (or by using an early page editor). But senior newspaper managers, the people plotting the business future of the industry, saw online publishing only through the prism of getting their content from their proprietary print systems through programs like Pantheon and onto the Web. That led many of them to see online publishing as something difficult, creating a high barrier of entry for potential competitors.

If newspapers were worried, it was about big-money rivals such as Microsoft’s Sidewalk. Individually published websites and blogs, when they started to appear, weren’t n anyone’s radar as competition. Managers saw online publishing as demanding complicated, expensive, technical solutions.

So online staff were put to that task, not to noodling around with online-only content, independently conceived and produced. Today, we can look back and see the opportunity missed. What if the Bay Area newspapers had developed a free online classified service, and attempted to upsell some of those free advertisers into a paid print ad – the opposite model of what so many newspapers pursued? If they had, perhaps there wouldn’t have been a Craigslist, and the future of the news industry could have developed along a radically different arc.

What if more managers had paid attention to the ease with which so many of us were cranking out our personal websites and charged us, on company time, to develop tools to allow all newspaper readers to do the same? Can you image what could have happened had newspapers developed and controlled the first blogging tools?

What if newspaper ad sales teams sold ads into those bloggers’ webpages, before Google got into that game with AdSense? What would the industry’s market share look like today?

But, of course, none of those things happened. Because, I suggest, 1990s newspaper managers looked at us, toiling with the likes of Pantheon and hacked-together Perl scripts, and concluded that online publishing was complex, frustrating and difficult. So by the time that online jockeys who didn’t have to struggle getting newspaper copy online had developed tools like Craigslist, Blogger and AdWords, the competition they unleashed overwhelmed the industry before newspaper managers could change their thinking.

About Robert Niles

Robert Niles is the former editor of OJR, and no longer associated with the site. You may find him now at


  1. says:

    In other words, you faced having some poor tools and never considered attempting anything else. Lost opportunity, no doubt, but the tools to blame?
    There were so many simple ideas like Craigslist that worked, but only for those who had the idea and did the work at the right time.
    I don’t see your point.

  2. says:

    I don’t think the only mindset problem is that the technical barriers were high. The two elephant in the room is that they are not really thinking about search (some examples here).

  3. I want to be clear that I am not blaming the tech here. I’m blaming the management mindset that the tech reinforced. What newspaper managers failed to recognize is that online production and porting content from legacy systems were, in fact, two issues. They were conflated in many managers’ minds, and that created the problem.

  4. says:

    From my experience, most publishers didn’t have a clue then about what was required, so I can’t agree with your assumption that one thing led to another.
    Unfortunately, they don’t have a clue now either.

  5. says:

    Atex was a big problem for you, where if you might have had Xyvision, you life may have been simpler…


    I think that this has nothing to do with why newspapers are tanking. It is the indeed the distribution model change. Like record labels failed to adjust to how music is now distributed, so goes it for newspapers.

    Book Publisher are next.

  6. Perry Gaskill says:

    Interesting take on the technical background to some of the current issues, Robert. I could be wrong, but I also have a sense that at least part of the problem, to this day, is due to an effort by newspaper software vendors to somehow augment legacy print systems to be web capable. There’s also a sense that instead of making the augment split off at the point of text entry, that it tended to occur further along in the workflow which made the web capability less nimble in terms of change.

    It’s also interesting to note your followup clarification about the split between what publishers perceived as recurrent operating costs and one-time capital outlay. A lack of understanding, one would assume, about the differences between the initial heavy lifting to lay a foundation, those things which are going to be anticipated costs to make a news site evolve as technology changes, and those things which are day-to-day expenses. There also needs to be factored into this the lack of a tangible business model to make the news site be self-supporting at a quality level equivalent to the print product. Something which is still unresolved.

    That said, it should be also noted that we are now working with the benefit of hindsight, and that some of the decisions made by news publishers more than a decade ago probably made perfect sense at the time. That they were proved wrong is because technology went in a direction no one could have anticipated. It also seems to me a more important question is what to do about it now. How do you jettison existing dysfunctional baggage and replace it with something better for the longer term? My own ideas for fundamental change tend to go something like this:

    The news industry needs to move beyond the idea of using proprietary vendors or custom solutions to build and maintain websites. For the sake of its own survival there needs to be a cooperative effort among publishers to develop a standard reference, an online news publishing platform using an open source model. If Wikipedia and WordPress can do it, why can’t news publishers?

    I could have missed it, but have yet to see any common online forum for discussion on those topics related to the broader view of the coding process involved in putting news online. Late last month, for example, Every Block released its Python source code and the event breezed by with very little mention.

    One of the major reasons the online news business model is broken is because legacy advertising models don’t work anymore. Instead of having print “up-sell” to the web, online news sites should be moving into marketing services to help local merchants sell stuff. There is a huge untapped realm out there of possibilities but so far what we’re seeing is news organizations waiting around until somebody else invents something.

  7. The word news and newspaper are always associated with the sense of speed. I believe I could quite catch the feeling of frustration that you tried to describe about publishing things onto web format in the last decade era…having to remain superfast with the news at hand while having to face with new learning curve of the new technology.
    Somehow, even though with current technology evolves the tools to be much better and easier to use, like the all available CMS like joomla / wordpress / VIVVO, the frustration then moved to some other new requirement to compete : the struggle to stand out from the gigantic amount of ‘Pages’ on the web.

    These days, any one can publish. High school kids have their own blogs, publishing their own creativity. Especially when people are starting to understand google’s algorithm. Professional journalist with ‘REAL’ content are struggling to compete against ‘Adsense earners’ who don’t have real content but know how to get their page ranked no.1 in google.

    I think this change should be much more frustrated. And i sense that another change will be brought into the flow of information highway again very soon.
    Big Change. With BING enters the market. I’m sure google will change their algorithm again, and let’s see….whose content…despite the technology used to create it, will get to be ‘saw’ by millions of people using search engines.

  8. says:

    What many industry participants fails to see is the role of bargain (the plain rules of economics).
    Tear away all software, all workflows and all efforts to make it into the web bizz; the bottomline was and still is: newspapers purely giving away every asset for free.
    If only this industry would have had a wide range of assets, but it did not. The industry had one: News. And noone dared to even try setting a price for it.
    It had to be free – said who?

  9. Having kicked around online publishing in the mid-90s, I think there are a couple of other big factors at play. First was the lack of real audience. As many papers started going online there just weren’t that many people with access to the Internet. Putting up barriers like a pay-model would just have insured that the few thousand readers with modems wouldn’t bother connecting to our site.
    But the real gating technology issue was the lack of viable e-commerce. Even if publishers had wanted to charge for their content, or tie it to print subscriptions, there just wasn’t an easy way to accomplish that at first. Even was launched with no way to buy tickets.

  10. says:

    when we look back to the time frame you mentioned, it’s important to remember that newspapers were fat and happy and enjoying margins over 25% year over year. In other words, they had little incentive to evolve beyond their print monopolies. They simply believed the status quo would last forever.