[ptsefton.com] | [CV & Bio]

Another defensive (yet not huffy) post about ICE


View this page as PDF

Dorothea Salo returns to the subject of Word processors, this time with a quote from somebody who as at the Extreme Markup conference, talking about issues with word processors and XML. (And she linked to my recent explanation of why ICE works.)

It seems that one thrust of Ms Salo's piece, and the commentary she's quoting is that word processors are not a good way for people to author for complex XML systems.

I agree, more below.

She also seems to be of the opinion that authors will always not use styles, and with that I can't agree. Sure, we'll never get all the authors using styles all of the time, but I think we can establish communities of practice where people do use styles.

Why do I think that?

Been there (Kannegieter and Sefton 2000).

Done that (Sefton 2004).

One scenario mentioned in the stuff quoted by Dorothea Salo is that authors don't do what you tell them:

The core of this talk was an extended lament on how authors insist on using Word; even if you provide specialized authoring tools, they compose in Word and then cut and paste, more or less incorrectly, into the specialized tool.


So what does this tell us? It tells me that if you can't get your staff or stakeholders to use your specialized tools then the tools are probably inappropriate for the authors and/or the task at hand. We had a system like this at USQ - it was an XML courseware system, but virtually none of the authors ever used it. Guess what? They used Word. This was noted and we responded with a different system that is more appropriate and which no longer requires back-room staff to remove all formatting and start again.

We now have the ICE system, which has 100-plus courseware authors at USQ with numbers increasing about as fast as our support staff can cope. (Yes there are editorial staff involved, it's a real publishing enterprise and yes they do clean some stuff up.)

As reported in the material quoted above there are real problems trying to match a word processor editing system to an XML system. That's why we don't try to do that with ICE. We take a pragmatic approach that recognizes that the most useful outputs for many purposes are HTML (with microformats for domain specific content like activities in learning material) and PDF. With that in mind we work with rather than against the word processor.

The ICE system was designed to target a generic document format, HTML, using generic word processors, not to target some specialised schema or use any specialised tools. We don't lament our authors using Word. We help them write their documents and look forward to seeing them embrace e-research and produce not just dead PDF but live web documents, as seen in a previous post of mine about chemistry.

I appreciate that word processors have caused problems for serious publishing people in the past, but look, on one level all I'm trying to do here is something that Microsoft and Sun should have done a decade ago:

  1. Design a complete default stylesheet for generic documents with sensibly named styles.

  2. Set up smart HTML export using said styles.

That's the core aim. We wrap it up in an application that does some other stuff, but we're tying also to break ICE into services so you can use the bits you like.

Dorothea Salo says:

I will add that testing such tools on a small, highly-selected author population (as Mr. Seftons blog post indicates that he has done) leads to tools that work very well for a small, highly-selected population of authorsand fail utterly once they move beyond that population.


I wouldn't put it that we have a highly-select author population they're just the lecturers and support staff at the University of Southern Queensland. To a large degree they have self-selected by asking to use the system that they've seen peers using, which I take as an encouraging sign. The population is small, but I don't take it as given that it can't get any bigger.

What I can say is that if you are a teacher who wants to write course notes in a word processor and you have materials like the ones we produce at USQ (see some of the courses in this list) then we've tested the system on dozens of users so it might suit you.

We are not confining our testing to courseware at USQ, using our Australian government funding we are reaching out to other communities of users around Australia, I've just been to Perth and Adelaide talking about ICE to receptive audiences. If we fail utterly we'll report that in January 2008 but we're not giving up yet.

I don't think we're failing. Yesterday I sat with a PhD candidate in Adelaide who's been wrestling with Old Icelandic quotes, and she was delighted when we worked out together how do get rid of rows of tab characters and replace them with styles. She seemed happy to see her work in HTML, with the promise that her supervisors will soon be able to look at it via the web and leave their comments via a web browser. Welcome to ICE-land Lisa.

We have had a couple of less successful trials with more distributed teams with less clear goals who had trouble with the system, and based on feedback from those groups we're now concentrating on building an easier to use, centralized web version of ICE with a much lower barrier to adoption.

Ms Salo has expressed some interest in trying ICE, and we've set an account for her and encouraged her (and the rest of you) to sign up to the ICE user list and post what you all think. Does it work for you? Problems installing? Need more styles? If we don't get specific feedback we can't make the system more relevant.


Kannegieter, T., and P. Sefton. 2000. Content Mangement for all of us: The Standards Australia Experience. AUSWEB. http://ausweb.scu.edu.au/aw2k/papers/sefton/paper.html .

Sefton, P. 2004. Publishing versus Handouts, contrasting approaches to courseware online. AUSWEB. http://ausweb.scu.edu.au/aw04/papers/edited/sefton2/paper.html .