2.2 - Areas of responsibility during the internship

2.2.1 - Hip Webzine production
2.2.2 - Analysis of Web server logfile data

During the eight-month internship period I had two primary areas of responsibility with Hip, both of them involved with the Webzine itself. The first area of responsibility was in the production of new issues of the Hip Webzine. The second area was to investigate and analyze of some of the information recorded in the Web server's logfile, in particular with regards to the readership of the Webzine.

2.2.1 - Hip Webzine production

Shortly before the start of my internship with Hip, the three people who had been responsible for the production of previous issues of the Webzine left Hip. As a result I was, in effect, the de facto editor and publisher of the Webzine throughout the period of the internship. Since Dick had had minimal involvement in the day-to-day production side, it was my responsibility to pick up the loose threads of the issue then in production. While doing this task, I realized that there were few guidelines on the production process. Since my involvement with Hip was short term and since our work arrangements meant that I would be working at Hip only one day per week, Dick and I agreed that it would be useful to try and rationalize and document the Webzine production process.

2.2.1.1 - Rationalizing the Webzine production process

The goal of this rationalization was to set things up in such a way that issues of the Webzine could be put together by other Hip staff members as time permitted or by contract (or volunteer) labour on a part-time basis. There were a number of factors that influenced the directions I took in setting up new production procedures:

Hip's hardware and software environment was in a state of constant flux. Some of the workstations were leased and were replaced mid-way through my internship. During this shuffle, machines were renamed, access to files and servers was altered, and the Web server software was replaced by a competitor's product. (8)
the operating environment at Hip consisted of Macintosh, Windows 3.1, and Windows/NT operating systems. One of my goals was to locate the Webzine material on a file server that could be accessed over the network from any of the workstations.
like many small startup companies in a volatile field, Hip's workforce has had a high turnover rate. During my time with them, approximately eight to ten people came and went. This lack of continuity, combined with the shifting work environment, meant that some kind of written documentation was crucial.

The results are shown in my report, which is included as Appendix A to this paper.

2.2.1.2 - HTML markup of Webzine material

Material for the Webzine was submitted in several forms. Ideally articles came to Hip in electronic form, either on floppy disk or via EMail. Occasionally however, they were only available on paper and had to be re-keyed.

Publishing on the World Wide Web requires that all material be "marked up" according to a standard known as HTML, or HyperText Markup Language. HTML uses a system of formatting tags, enclosed in < > characters, which are imbedded in the document's text and sent to the Web browser over the network along with the text. The tags are then interpreted by the Web browser, which formats the page for display to the reader (Graham, 1995:2-3).

As an example, consider the following selection of text and HTML tags:

<TITLE>The simplest HTML example</TITLE>
<H1>This is a level-one heading</H1>
Welcome to the world of HTML.
This is one paragraph.<P>
And this is a second.<P>

which a Web browser will display as shown in Figure 2.

Figure 2: A simple HTML example

Markup of the Webzine material was done using the BBedit software for Macintosh. Other alternatives were explored, but they were found to be either unreliable (a beta version of Internet Assistant for Word for Windows crashed repeatedly during testing) or not suitable for regular production work. Once the material was marked up, I proofed it using the Netscape browser. After final proofing, each new issue was published on the Internet according to the procedures described in Appendix A.

In addition to copy markup, I did a minimal number of editorial revisions on the copy, and communicated with authors to keep them informed of upcoming publication deadlines. I also proposed some changes in the page layout and design within the Webzine in an attempt to improve the readability. In early issues the text pages were all manually "double-spaced," with fixed length lines separated by a blank line. My redesign eliminated this double-spacing, and added a cleaner "look" to page headings through the use of centered titles and an enlarged cap at the beginning of each page (Figure 3).

Figure 3: A Webzine page after layout redesign
(click to see full size version)

2.2.1.3 - Restructuring the Focus catalog

As an outgrowth of the Webzine redesign, Dick had me participate in redesigning and restructuring of the Focus catalog. Focus International was among the earliest of Hip's paying clients. Based in New York state, Focus describes itself as "Your Sexual Health Education Store," and it markets a range of sex-education material (primarily videos) thorough its online catalog (housed on Hip's Web server).

The catalog redesign centered around some new Perl scripts written by Greg Lipowy, a Hip employee during the spring and early summer of 1995. These scripts provided a simplified ordering interface that allowed browsers to order products from the Focus catalog directly over the Internet. Items could be paid for with a credit card, with Hip receiving a percentage of all sales.

2.2.2 - Analysis of Web server logfile data

My second major area of responsibility during the internship was to investigate the logfile that was kept by Hip's Web server. The logfile recorded each page reference, or "hit," providing a chronological record of all activity on the server. Since Hip's server holds pages from a number of different clients, all page requests are intermingled, creating a disorganized, but potentially very rich, collection of data.

Dick was interested in coming up with analytical software tools that could sort through these Web server logfiles and extract useful information on the readership of the Webzine. These tools might later serve as prototypes for the analysis of activity on other Hip clients' sites.

A number of goals were established, and existing logfile analysis tools were examined to see how well they could meet our goals. While the information that we hoped to obtain through logfile analysis was specific to the Hip Webzine, I also hoped to be able to obtain some insights into reader behaviour within hypertext publications in general.

With these goals in mind, we decided that I should work towards extracting the following information from the logfiles:

the "hit" rates for each page of the Webzine, to gain an overview of the relative popularity of the various Webzine pages
the completion ratios of the two-page Webzine columns, to obtain a sense of which Webzine columns were being read to completion and which were only being sampled
the sequence of pages read within the Webzine by any one reader, to determine how long was being spent on each page and which inter-page links were being used most frequently

The process of extracting this information, and the development of suitable methods for analyzing the data, are described in greater detail below.

Footnote 8:
The malleable environment which made well-documented procedures necessary was one of the things that made it difficult to keep the document up to date. (back)