I’ve been going on here about using styles, but some concrete examples are probably worthwhile as Bryan D. Wilhite has asked on his blog about what I mean by this:
why not use the built-in word processing features and layer your semantics on top of them? Use the methods already provided for expressing your own semantics. Method number one is styles. Method number two is tables. This approach will work in versions of Word that don’t have XML export, and with forthcoming Word formats that have been announced but not released and with other word processors that save in XML such as OpenOffice.org Writer.
Mr Wilhite shows how he uses his clean XTHML tool to write his blog in Word 2003, and wonders whether I use tables where he uses XML elements mixed in with Word’s markup. The answer is no. I don’t use tables. I use styles.
First of all, there’s one area we have in common. A paragraph is a paragraph.
In his screenshot it appears that Bryan has used Word’s built in structure to represent paragraphs. It looks like links are handled in a similar way – using Word’s built-in hyperlinking function. Again, when I am producing a blog entry that’s what I do too (only I am using OpenOffice.org Writer, not Word).
Where we differ is in the use of inline elements like
(In the spirit of Bryan Whilite’s self-referential post I have added the style names I used to write this post to the rest of the paragraphs).
So, to answer Bryan Wilhite’s questions of me:
For the example you have given I would not resort to tables, but would use the in-build word processor structures for paragraphs, links and images. In your example I’d use a style like ‘p’ for every paragraph and then have the back-end code turn that into an XTHML ``
For the inline elements, such as acronym I’d use a character style. In the ICE system, we use character styles based on XHTML element names so you would mark XML with the character-style
i-acronym. (For a list of the styles we use see my word processing site).
Where I would use tables would be in a case like a sidebar or in the educational context, a reading or an activity. Say I want to put in an activity, I would use a table with two cells in one column. The top cell would contain a paragraph in style
h-activity, and the bottom cell would contain the activity itself, a mixture of text and graphics. The back-end processing system would recognize that it’s dealing with an activity and could treat it specially, probably by throwing away the table code and putting the activity into a iv
I don’t think much about the Word XML task pane, I have never used Word’s XML features in the way that you’re using them, because as noted before here I think they are a bad idea for writing documents.
It’s a bit tricky to answer Bryan’s final question about what’s wrong with the system he’s built. I think it is solving a problem that I don’t have. I wouldn’t like to try to roll it out to authors here are the university because:
It uses Word 2003, which means it locks out Mac and Linux users and probably working from home for many people.
It uses WordProcessingML which is about to become obsolete and code that runs in Word that I would not like to try to run in a server environment.
It requires users to deal with new interface elements like the XML task pane.
We have our own way of making XHTML which uses styles to do most of the heavy lifting, and can create course-packages for upload to a Learning Management System.
So Bryan Wilhite, I have a question for you. Would you be interested in exploring with me whether we can get your clean XHTML system to use styles to produce its output?