Imaging & Document Solutions - First Looks: Server-Side File Conversion for the Web

featuring

CONTACTS

Server-Side File Conversion for the Web

By Lowell Rapaport

More and more companies are communicating with their branch offices, suppliers and customers using Web technology. But ordinary browsers need help handling the wide variety of file formats used in business. Last month's First Looks examined a pair of client-based file and image viewers that integrate with just about any browser. Once installed on the user's PC, they make viewing, downloading and/or printing almost any type of file a seamless, browser-based activity.

This month we look at three products that approach Web-based file viewing from the server side. Why would you want a server-based solution for file delivery? The main advantage is that you won't have to install viewing software at the client level to make information accessible. This is particularly important if you have large deployments, multiple locations or applications extending outside your enterprise - to suppliers, service providers or customers. Massive internal client installs are costly and time consuming, and it's much harder to guarantee success if the desktops are outside your enterprise.

Server-based file delivery gives you near universal compatibility with any Web client. Preprocessing files into a Web-compatible format at the server will ensure that all your users and customers will be able to get the information they need, without installing viewers or plug-ins.

The downside of server-side file processing is that your server needs to have much more horsepower than an ordinary file or Web server. In addition to serving the files, the server also has to convert them to a form that the browser will understand.

Another challenge with server-based viewing is the browser's limited features as a document viewer. There are no zoom, thumbnail or annotation functions like you get in a full-blown file viewer. If you want to provide these features, they have to be implemented on the server via scripts. Two scripting alternatives are to use either downloadable Java applications or Javascript. These tools let the server turn file processing and viewing functions over to the client machine. Since Java and Javascript are built into most browser platforms, users won't have to install helper applications or plug-ins.

In developing NetVue, Java was the technology of choice for Accusoft, Westborough, MA (www.accusoft.com). The NetVue client is a Java application with an assortment of common image-viewer tools, and it will automatically download and cache within the browser. On the server, NetVue maintains a document repository in a SQL or open database connectivity database. This lets it integrate with existing document management systems with a compatible database.

NetVue converts more than 100 file formats, including an add-on module for Microsoft Office files. It is popular with application service providers (ASPs) because it lets them support their customers' existing archives without converting the documents for Web compatibility. It also allows them to support any hardware on the client side. ASP Network Computing Devices of Mountain View, CA (www.ncd.com), uses NetVue to support an array of financial services and manufacturing clients. Datahub of Westlake Village, CA (www.datahubonline.com), uses NetVue as part of a service that allows municipalities to distribute government documents such as title deeds and birth certificates online.

NetVue uses a high-speed streaming protocol to download documents faster than conventional HTTP. The product also offers an annotation module that supports up to 256 layers of annotations. According to AccuSoft, annotations can be used to turn NetVue into an ad-hoc workflow solution.

The latest version of NetVue, version 3.1, adds a number of enhancements. Released in September, version 3.1 offers an optional Content Manager that can search NetVue's document repository for keywords and/or phrases. An optical character recognition (OCR) engine lets you search content in nontext files such as TIFF images. AccuSoft says these documents are automatically processed in batches during system idle time, and the search metadata is saved. This prevents users from having to wait for OCR processing when they initiate a search. The content manager flags keywords found within documents and allows easy navigation between occurrences of those words.

Scheduled for future release are Unix and Linux ports for the NetVue server. The company hopes to have the Linux version ready this winter and full Unix support available next spring. Current prices for the NT version of NetVue start at $5,000 per server plus $1,600 to $5,000 for the PDF, CAD, Microsoft Office and Content Manager add-on modules

The venerable Outside-In server is another alternative for delivering a range of file types through ordinary browsers. Developed by the Chicago-based Information Exchange Division of IntraNet Solutions (www.outsideinserver.com), Outside-In converts all documents to HTML text and all images to GIF, JPEG or PNG formats that can be directly rendered by a Web browser. There are no plug-ins, Java applications or Javascript applets, but there are also no annotation, image navigation or image processing features.

Outside-In has been around since the days of Windows 3.X, and for many years it was embedded within Lotus Notes. Today it is positioned as a Web publishing solution for more than 200 file formats, including, Microsoft Office, WordPerfect and other leading desktop application formats. The emphasis is on HTML formatting, and the server software includes templates for all formats and browser types supported. This makes Outside-In very flexible. The HTML code it generates can be custom designed for Internet appliances like Web TV, handheld computers using Wireless Markup Language (WML), and standard Netscape and Internet Explorer browsers. An upgrade specifically supporting handheld computers is expected to bow by the end of this year.

One caveat: Outside-In server has to co-exist with the Web server on the same machine. Because conversion and rendering are processor intensive, scaling up can be difficult, but it is compatible with load-balancing clustering software, a popular solution for scaling-up servers.

Outside-In is file-system-based; integrating it with a document management system's database is possible through the use of scripts, but it is awkward. The HTML rendering engine is available fully integrated with Filenet, Documentum and PC/Docs document management solutions, so a careful look at the Outside-In product will give you an idea how those document management systems have been Web enabled.

Outside-In is of particular interest to businesses that encounter a wide range of document types. Yahoo!, for example, uses Outside-In as the online document viewer for Yahoo Mail. It is also useful in environments such as schools and hospitals, where there is a wide variety of client computers. Outside-In is also useful for businesses that need to make unstructured documents, like human resources information, resumes and freeform documents made in a wide variety of application software available to online users. Outside-In is available at $10,000 per server license or through software leasing plans.

One of the simplest server-side solutions available is the Doc-to-Net image conversion tool from Skyline Tools, Woodland Hills, CA (www.skylinetools.com). Doc-to-Net is a file format filter that converts scanned document images such as Group 3 compressed TIFFs into JPEGs, GIFs or PNGs that can be displayed by a Web browser. Like Outside-In, it is a pure server-side product; no scripts or applet are downloaded to the client. Doc-to-Net does include image-processing tools, like deskew, despeckle and zoom, that are not offered through Outside-In.

Doc-to-Net comes in two versions. A CGI script application that can be added to a Web server and a dynamically linked library (DLL) for conventional application development. On the Web a client accesses Doc-to-Net's image-processing features by clicking on links to CGI or ASP scripts embedded in a Web page.

"Doc-to-Net is a tool for Web developers," says Skyline President Jan Dekkers. "Our customers include a number of Internet fax companies that use Doc-to-Net to make Group 3 compressed faxes available to Web users."

Doc-to-Net is also available as a traditional DLL object to be added to any document imaging application. The tool gives developers an easy way to Web-enable a document archive without converting an archive of scanned document images into browser-compatible formats.

Because Doc-to-Net relies entirely on server-side processing, it will even work with browsers older than version 4.X. It won't format Microsoft Office application files for the Web, but at $600 per server license, it is highly affordable. If your Web publishing needs revolve around scanned document images, it's the perfect solution.