The On-Line Document Conversion Newsletter

December 1995: Volume 1, Issue 12

Internet Assistant Version 2.0: First Look

by David B. Demyan

Version 1.0—Worth What You Paid

IA Version 2.0—Some Improvements

Well, here I am grudgingly admitting that Internet Assistant Version 2.0 is an improvement over Version 1.0. It is more reliable than the previous version and performs a few new functions. It still doesn't do tables or graphics, but at least it allows you to use them (Version 1.0 choked on any HTML codes it did not recognize). IA Version 2.0 at least passes unrecognized codes on to be properly interpreted by your browser-of-choice (not likely to be the one included with Internet Assistant in either version).

Where Can I Get It?

Specific Improvements

• Probably the most important feature is that IA Version 2.0 is less likely to crash and corrupt your files. I used Internet Assistant Version 2.0 for a few days without any crashes. Furthermore, I was able to open previously-saved HTML files, even ones with unsupported coding like Netscape Table Extensions, without significant file corruption. After a few futile attempts to re-open existing files with the previous version, I finally realized that Version 1.0 was only a converter: it could not be relied on as an editor.

• IA Version 2.0 takes advantage of Word 7's ability to handle bigger files. Because of the 32-bit architecture of the newly-designed Windows 95 programs, IA can probably handle bigger conversion tasks. Furthermore, any WordBasic macros you use in your conversion activities will run faster and you probably won't get that annoying "Out of Memory" message. More on the use of customized Word macros in a future issue of Conversioneering.

• Due in part to some of the factors above, it is now possible to open several Word/HTML documents at the same time and create or troubleshoot cross-file hypertext links. The previous version thwarted this primarily because you could not re-open a document in HTML view once it was saved due to the possibility of file corruption. With that threat gone, you can safely open several at once.

• Style tag conversion is a bit more predictable. While this area is not yet perfected, it is much better. If your headings are all consistently tagged, for example, you can expect they will be accurately mapped to the proper HTML heading codes. A notable exception is the quirky numbered lists, described below under "Shortcomings."

• While IA Version 2.0 does not convert Word's own tables to anything resembling a table in HTML, at least it allows you to pass the coding through. More on the table issue in "Shortcomings."

Notable Shortcomings

• Tables in Word documents are not allowed. Imagine that: You can create all the fancy tables in Word you want but you cannot put them on-line. They can neither be created in IA nor converted. This problem requires you to convert the tables elsewhere and cut-and-paste them into your final HTML files. There's a decent Word table to HTML filter at www.cadd.nps.usace.army.mil/tbl2html.htm). You'll probably want to do your final cut-and-paste outside of Internet Assistant, since it does nothing with the codes. I found it tolerable to edit the tables in the new MS WordPad, then incorporate them into the final converted files by link or by pasting in-line. A reasonably automatic table filter that works during conversion is desperately needed.

• Word-embedded graphics are not handled. They will show up on screen in Word/HTML view, but they will not be converted to anything in the results. A good workaround: before conversion, manually convert graphics to GIF format, then import them into the Word document. The conversion then creates an in-line link to the graphic and it appears where placed in the result.

• Word linked tables of contents are not converted. My trials still required manual linking to bookmarks that I had to manually create for each heading I wanted to include in the TOC. This should be more automatic, as in Quadralay WebWorks (for FrameMaker document conversions).

• There is a problem with lists: you know, those strings of numbered, lettered, bulleted, and dashed items that most documents have? This problem is shared with other conversion programs: you cannot reliably convert "nested" lists. For example, if a numbered list is interrupted by a plain indented paragraph, you should be able to continue the numbering scheme after the interruption. No commercial converter allows this today. Typically, the numbered list ends and a new one is started after the interruption. See my example. This is an area that needs attention and perhaps IA will feature a better way of handling nested lists in a future release. Meanwhile, does anyone know of a good macro that supports this type of conversion and gives the user some control over numbering?

• Some other miscellaneous document features that IA throws away:
  • annotations
  • borders and shading
  • captions
  • character formatting (e.g. superscript)
  • drawing layer elements
  • embedded objects, or "cut and pasted" objects, such as equations, clip art, Word Art, and MS Draw objects
  • fields--only the field result is converted
  • footnotes and endnotes
  • frames
  • graphics embedded via the Clipboard
  • headers and footers
  • indented paragraphs in any paragraph style other than OL or UL
  • index entries
  • page breaks and section breaks
  • revision marks
  • tabs in any paragraph style other than PRE and DL
  • TOC entries

Conclusion

Conversioneering will devote more review space to Microsoft Word Internet Assistant in a future issue. In addition, we'll cover any new conversion programs that come along and discuss general conversion topics. Some of our future articles:

• On-line Document Conversion Based on Standards and Templates
• Detailed Review of Microsoft Word Internet Assistant 2.0
• Quadralay WebWorks Converter for FrameMaker Documents: First Look
• Quadralay WebWorks: Detailed Review

Call for Ideas