It'll probably cost about $100 on a stand-alone basis, but as a client on a network it could get up to about $600, plus another $4000 to $5000 for the server.

All of this is allegedly unattended, though experience leads us to suspect that the main attendant will be a service technician. But we'll wait and see.

This level will set you back about $5000 for the software alone and a gazillion dollars for the hardware.

But hold on, the ever-crafty Israelis have come up with a solution they call Intersite: Send a Netscape browser search off to an Iota database and you'll get back not a bunch of pages, but a bunch of strips (Headline: Hillary Rodham Clinton Strips on Israeli Computer System!), each containing a hit found by the search engine, extracted from the page and encoded to HTML format.

That means you'll get citations, in a few lines of context, returned to your screen relatively quickly. When you find the one you want, double-click on it and the full page will start oozing across.

What you have when you get done is a full page, with jumps, pictures, sidebars, etc., that you can view, print or store on your own machine or server.

The historic Iota problem persists: Interesting, even potentially watershed technology, but no one is running it out in the field.

But this might be the year; stay tuned.

SRA International
"Everything we're doing these days is web-based," said Frank Roche, media systems marketing director, clearly reading the tea leaves strewn over the shoulder of the Infobahn.

SRA, which used to cozy up to intelligence agencies before government was reinvented and it ran out of money, has been shopping its data-manipulation services for the last few years to newspapers and other giant warehouses of nouns and verbs.

This year, Fairfax, Va.-based SRA is particularly proud of four products:

So you can put together a "who was in the news this week" list, dynamically, for your web page, for instance. It works with a search engine, but doesn't require one. Think of this weapon in the hands of the New York Times crossword puzzle editor.

In the outside Internet world, you could feed it a URL from, say, city hall and keep the search engine on your newspaper's web site. The possibilities are interesting.

One of the unique things SRA claims is its ability to mix apples and oranges -- to use the above tools to pull together, for example, a BRS-based text archive with a Sybase photo archive and make them searchable through the same browser.

SRA is installed at the Chicago Tribune and U.S. News & World Report, but a couple of big newspapers are in negotiations for the latest software and could be announcing contracts soon. That should provide the industry with a real-world test.

Gannett Media Technologies International
This offshoot of the media giant has embraced the Web with open arms.

"We don't ever expect to write another client again," said Bill Toner, vice president for operations. "This whole [Intranet] thing is fine with us."

"We're using Java, windowing, putting toolbars along the side of the window, even using a crawler to put messages up" on the Netscape browser, Toner said. "As soon as we can do drag-and-drop, we will."

The DiGiCol archiving system that Cincinnati-based Gmti sells is growing, too -- in addition to web browser support, DiGiCol (finally) supports Adobe Acrobat's PDF format for storing, indexing and searching full pages.

But Toner's been talking to the Iota folks, too.

"We're talking about integrating the Iota indexing utility into our system for the specific things that use bit-mapping, where OCR [optical character recognition] doesn't do a good job.

"We want to be able to handle hard copy a little more elegantly. With some large drawings, you can't index the text -- Iota will solve that," Toner added.

He also pointed to an interesting Web advertising function that DiGiCol will be able to support: Say you're running a web site. You sell some ads, store them in the DiGiCol database.

The archiving system can read the Internet addresses of people calling up your site, determine their locations and send a targeted ad out to the web site, just for that "customer"

It's kind of an Internet Caller ID program, which should make your newspaper's database marketing people salivate.

Lexis-Nexis
The giant database company based in Dayton, Ohio, is really a front for tiny Tribune Solutions, the R&D division of the Morning Tribune, a 27,000-circulation newspaper in Lewiston, Idaho.

Glenn Cruickshank, manager of Tribune Solutions, takes great delight in describing his firm as a Chevy among a fleet of Lexuses.

"We started off as a text archive company in the late '80s," Cruickshank said, after developing a system for in-house use. About three years ago, when images needed to be added to the NewsView system, "we already had the Folio engine, which had the ability to display text, pictures, sound and video," Cruickshank said, so it wasn't hard to add photos.

Now, like everyone else, he's doing web things, but not browsers -- Lexis-Nexis Connections software sets up text and photos for other on-line suppliers, for example.

This year it's evolved into a "complete drag-and-drop electronic newspaper system." Cruickshank said. "You just move stories around on the page, hit the Build button and it builds the whole paper for you."

As proof, he offers his own paper's web site: http://www.lmtribune.com.

Instead of dedicating two to 10 people to process stories for the Web as most papers do, Cruickshank claims one person at the Tribune can place 70 stories a day, and do it in a half-hour.

The Detroit News, Cruickshank said, is doing the same thing at http://www.detnews.com.

Another new wrinkle in Connections is byline rights management, which sounds like something anybody could add to a database but nobody did, pre-Cruickshank.

"When you're sending stories out to other on-line services, you can't just throw everything out there that they want because they may not own it -- if a stringer wrote it, for example." Cruickshank said. "Connections manages that on a byline-by-byline basis."

Also new this year: a completely re-engineered version of PhotoView that now stores EPS files in addition to TIF, Iptc and JPEG formats.

Tribune Solutions also has redesigned the interface, and added about 40 fields supported by the SLA (Special Libraries Association) guidelines. Before, there were 18.

All of this is in the same IBM database that can sit on a small, Chevy-type front-end system and purr along contentedly, Cruickshank said.

T/One Inc.
T/One's Merlin photo archiving system is Netscape-ready for anything the 'Net can throw at it, and is hanging onto its old browser at the same time.

"We see advantages to having both," declared Pete Leabo, director of marketing.

"The browser is great for the occasional user who just wants to see if the photo exists," said Leabo, "and if it's there retrieving it from the archive. From the user's point of view, it's wonderful."

But in the library, Leabo sees the dedicated client as the better choice: "When you're doing extensive work with the archive itself, for example, a librarian who is going to be appending keywords and adding published captions, correcting information and doing that quickly and effectively, that's where the client pays off.

"But we're doing both and will continue to do both," Leabo said.

New this year is Merlin 3, which is already in production at the Boston Globe and The Record of Kitchener-Waterloo, Ontario.

It is the first complete redesign of the Merlin application, which is about 2¢ years old, and it is, Leabo said, faster than a speeding bullet.

Merlin 3 has autopurge and autoprotect features which enable it to be used as a repository for incoming wire photos.

"Some papers like The Virginian-Pilot [of Norfolk, Va.] were already flowing wire photos into the older software," said Leabo, "but that puts an undue burden on the librarian to manually delete about 300 pictures a day. Now, with autopurge, any image not protected can be purged from the database."

Similarly, you could set the system to autoprotect any locally scanned images by looking at the credit field, so you'd know you'd never accidently let the system purge them, Leabo said.

Other new features are the ability to save the results of several searches -- similar to AP's lightbox feature, but to save the searches themselves.

"You could have a search you'd be repeating regularly -- say, all sports photos of the 49ers in the last six hours, using multiple search criteria," Leabo said.

"You'd create it once, execute the search, save it as a 'Query Shortcut' and give it a name. Then it becomes a menu item."

And all of this with faster throughput as well? How did they do it?

"We were not re-engineering for speed," insists Leabo, "because we have the fastest archiving system anywhere to start with."

But as they reprogrammed to give the system a nicer user interface and add features, they tightened the code, optimized some of the search routines, and presto!

"We got a speed increase of at least fourfold and in some cases up to tenfold," Leabo said.

By NEXPO, he adds, his team may have another goodie: The ability to issue a "Save As" command on a graphic and have the Merlin system save it in separate text, photo and graphics databases, while maintaining the same links to the original.

-- John Bryan

The Associated Press,
(212) 621-1732;
Gannett Media Technologies International,
(513) 665-3777,
e-mail: dzito@gmti.gannett.com,
http://www.gmti.com/;
Iota USA Inc.,
(203) 227-5602,
e-mail: glellevre@aol.com,
http://www.iota.col.il/;
Lexis-Nexis,
(800) 227-9597, Ext. 1819;
SRA International,
(703) 803-1500,
e-mail: frank_roche@sra.com,
http://www.sra.com/;
T/One Inc.,
(617) 328-6645, e-mail: pbl@crl.com, http://www.t-1.com/.

From THE COLE PAPERS, May 1996, Copyright © 1996, All Rights Reserved.