Content Development tools
Common Gateway Interface (CGI) is a programming specification for web servers. It makes it possible to have HTML pages tell the web server to run programs that make your web pages interactive (Wagner & Englemann, 1997). HTTP was designed to move information around, not process it. In order to interact with databases, perform searches, or otherwise manipulate information, HTTP requires a well defined means of handing off this work to other programs. CGI does just that.
- Gateway can be written in any interpreted or complied language supported by the web server platform. UNIX--PERL, C, C++. Windows NT--Visual Basic and Visual Basic C++. CGI tend to be more predominate because of their platform independence and easier use versus other programming interfaces.
- CGI is a standard because all HTTP servers support it and because it is a platform independent means of programming.
- CGI Routine
- Client -- Browser calls the CGI script on the web server.
- Web server (HTTP server) -- The HTTP gets the script.
- CGI script executes on the server and returns the information to the browser as HTML.
- Preferred technique because;
- Generally easier than programming server specific Application Programming Interfaces (APIs).
- Costs tend to less.
- Use true programming languages but remain independent of the web server (powerful and flexible), platform independent.
- Weak performance -- The gateway program must be started by the server each time for each CGI request.
- Security is an issue -- Problematic.
Some examples of vendor specific APIs include familiar names such as Netscape's Server API (NSAPI) and Microsoft's Internet Server API (ISAPI). "The general principle for either API, as for any proprietary extension to a shared technology, should be to use it when nothing else will do" (Bennett, 1996, 121).
HTML Editors and Tools
Intranets would not exist without the presence of HTML editors and tools. Just as these applications have fueled the Internet publication boom, so too have they influenced the Intranet world. Libraries can certainly make use of the multiple products available to facilitate web page publication and expansion and most already have the means to do so because of the Internet library web site publication.
Web tools for receiving and publishing information are deceptively free. All you need to start an Intranet are a free server and free web clients. A skilled user may be able to set up a Website from scratch in an afternoon. And that is where the problems can begin. (Flohr, 1997, 69)
Flohr's concern stems from the fear and overwhelming presence of the Internet and Intranets on the issue of document management. The ease with which Intranets can be set up means that it is possible for a library or any other organization to have an Intranet without the organization's awareness or so he claims. Whatever one may think HTML editing has become easier because of the demand for products which provide greater flexibility and performance. Many powerful editors for Intranet use have full editing features such as web templates, design wizards, drag and drop insertion of interactive web elements, built in conversion from RTF and ASCII formats, and translation of a wide variety of image formats to GIF and JPEG.
Web Browsers
Intranet web browsers or search engines are prolific. The marketplace has an enormous amount of competing products and as stated by DeJesus "Like a book without an index or a library without a catalog, an intranet without search engines is just a pile of information, no good for the people its intended to serve" (DeJesus, 1997, 92NA1). It is important to match your searching needs with the right search engine for your Intranet. Many libraries have large mainframes with databases that produce data streams in native 3270 format. Some Intranet browsers are able to convert that data into web pages and make use of the transactional systems that have been created for the mainframe. Two such products are Simware's Salvo Server edition, which converted the Smithsonian Library Catalog form 3270 screens into HTML (Salamone, 1996) and Intelligent Environments Inc.'s Amazon tool (Zelnick, 1998). Keeping Intranets organized is a difficult task but as Gustavson states,
whether high powered intranet search engines can be sensibly implemented, or corporations recognize the need for human filters and librarian, one thing is beyond argument: without the network, no enterprise publishing is possible. (Gustavson, 1997, 48)Callaway provided an interesting breakdown of the top five web based document management vendors
| Vendors | Market Share |
|---|---|
| Documentum Inc. / Pleasanton, CA. | 14.7% |
| PC DOCS Group Int'l/ Toronto, ON. | 13.8% |
| Altris Software Inc./ San Diego, CA. | 7.9% |
| Information Dimension Inc./ Dublin, Ohio | 7.7% |
| Interleaf Inc./ Waltham, MA. | 6.7% |
(Callaway, 1998, 80)
Callaway also noted that Open Text Corp of Waterloo, ON. was currently numbered 7 in terms of market share, but gained more new customers then any other document management vendor between the end of 1996 and the end of 1997. [Another noted Canadian product is Fulcrum Knowledge Network]. Most of these Intranet search engines come with a substantial price tag to accompany their corporate target market (Guennette, 1997; Zorn, 1997) but ther are free search engines available for Intranet searching. Some have already been mentioned such as WAIS and SWISH (C Program), but Zorn also evaluated four others in comparison to some corporate vendor based products.
- Excite for web servers (EWS)
- ht://Dig (UNIX based)
- Htgrep (UNIX based , CGI Script written with PERL)
- Webglimpse (UNIX based)
If the intranet site is small and does not contain documents in formats other than HTML and ASCII text, the freeware search engines may be enough to do the job. The frequent downside of these free tools, however, is that advanced technical knowledge is needed to configure and customize the software for site specific use, and that advanced searching functionality found in commercial engines is not available. In addition, little formal technical support is offered by any free intranet search engines, except for Excite, which charges for its support and maintenance contract. (Zorn, 1997, 47)
Another interesting application of Intranets is the ability to access databases with search engines that convert 3270 data streams into HTML. The option to access legacy information is certainly valuable for many organizations including libraries. Another product that has won over people is the Resource Organization and Discovery in Subject - based Services (ROADS) suite of software that attempts to overcome many of the limitations of OPAC database software. ROADS is a UNIX only solution, is free, well documented, works well, maintenance is web based and configuration is done through the web interface with ASCII files sent via ftp to their proper locations. Adding and deleting records to this database is a matter of using HTML forms (Morgan, 1998). I would think that such an effort would certainly be compatible with any intranet implementation that was on a UNIX platform. Certainly worth investigating.
A side note to the web search engine foray is a recent development using "a JAVA based three dimensional data animation tool that displays multiple relationships between objects in a sort of perpetual information dance" (Marlatt, 1998, 25). The product is called Thinkmap and designed by a company called Plumbdesign.
Thinkmap's first commercial application, launched in early March, is the Smithsonian Web site exhibit "Revealing Things" which explores the cultural importance of common objects, such as a pair of jeans, a chemistry set, ... (Marlatt, 1998, 27).
The 3-D navigable thesaurus may be the future for library OPACS let alone web search engines.
Part 1: Intranets
Philosophy