Twenty Years in the Making
by David M. Cole
If you ever sat through an ANPA Wire Service Committee meeting in the
1980s, you know that they were designed to make your teeth hurt.
Well, maybe not designed. But I can remember a throbbing molar or four after sitting through a debate on where to place the symbol for a "club" so that syndicates offering card columns could transmit bridge hands to newspapers.
I am told I was lucky to miss the earliest meetings, where members were worrying about important issues—like whether to support the Bell 202 protocol, which went whizzing by at a then-blazing 1,200 bits per second. (The modem on your laptop or desktop now goes at 56,000 bits per second, and my DSL line goes at 384,000 bits per second.) The result of those meetings was a document called ANPA 1312; it was published Feb. 1, 1979.
The role for ANPA 1312 was as important as it was simple: It provided a way for wire services to consistently transmit stories to newspapers. It also ensured that the coding embedded by the wire service—bold face, spacing, typographic markup—would not only survive the trip, but really be bold face or a thin space when it got to the newspaper.
Essentially all North American wire services and all the suppliers of wire collection software adopted ANPA 1312 as the "standard," though I have this niggling memory that UPI did something a little different than everybody else with the header. I seem to remember that we had to run a patch on the UPI port to compensate.
Anyway, for many years we thought the Wire Service Committee had fallen into inactivity. But, lo and behold, the group—made up of newspaper, wire-service and supplier stalwarts who have more patience and resilience than a flock of cat-trainers—came out of hiding in the early part of the decade and began to talk about a Universal Text Format. The UTF would acknowledge the fact that wires were going a little faster than 1,200 baud, as well as support more sophisticated markup and coding.
The Wire Service Committee perked along through the middle part of the decade building a protocol that would expunge all the weaknesses of ANPA 1312. Until committee members heard about SGML, that is.
Standard Generalized Markup Language—and its cousin HyperText Markup Language, or HTML—had become the "in" thing among standards organizations, and after trying out a couple of other formats, including Microsoft's Rich Text Format, the committee picked SGML as the basis for UTF.
Then, last year, everybody got excited about eXtensible Markup Language, or XML (TechNews, March/April 1998, p. 21). Fortunately, XML is another cousin of SGML, and with a little work, the NITF (News Industry Text Format as it had become known) became XML compliant.
Shortly after the Wire Service Committee (which had taken to calling itself the News Information Task Force—NITF, get it?—and also included representatives from the International Press Telecommunications Council, or IPTC) got into the XML world, there was a bunch of discussion about using XML to "tag" stories, so they would work well in media other than print. XML tags could also include information reducing the need for archivists to add "keywords" when sending stories off to the digital-asset-management system.
But it was just discussion—the NITF people said they'd get the News Division of the Special Libraries Association involved (and they were, for a while), but nothing much came of that work.
Until about nine months ago, that is, when an offshoot of an American Press Institute meeting decided that there needed to be a "grammar" to new media, and part of that grammar was a standardized set of tags. Though it's unclear whether the API Grammarians understood what was happening with the NITF, it became abundantly clear a few months into the project that the NITF group believed that the API Grammarians were stepping all over their territory.
At a big lunch in suburban Washington, the bruised feelings were soothed and the differences—which were minimal—between the NITF and the Grammarians' work were ironed out.
Then, out of nowhere in March, a company that doesn't even directly supply the newspaper industry introduced its own version of the NITF with Grammarian-like XML tags. It was shot down within a week.
In late April, the NITF and the Grammarians announced that agreement had been reached and that everybody endorsed everything (though there's still a vote that has to happen at June's NEXPO; see related story, p. 22). At least one supplier put out a press release in late April announcing that its product "will be the first digital-asset management system for the news industry to completely incorporate NITF as its core data structure."
Twenty years to get a new standard? Yes, but not as bad as a 20-year trip to the dentist. n