Rich Text and Argument Maps

[gusl]

message I left for Danenberg at Wikisophia:

Rich Text and Argument Maps

Hello,

I think WikiTeX is a brilliant idea... and I would be happy to make contributions in my areas of interest.

I would like to define a language for semantically-rich text, including |Argument Maps (http://www.austhink.org/monk/Fenner/Fenner.htm) and finer formalizations. This formal structure would be reminescent of XML and could have many ways of being displayed, all of which I think could be encoded in graphviz.

AFAIK, LaTeX is meant for static documents only (DVI, PS, PDF), so I imagine that it doesn't support such dynamic structures. Or does it?

Is the language used to render a WikiTeX object always the same as the one used in the corresponding LaTeX package? If so, does that mean that the first step is for me to create a LaTeX package?

Gustavo 10:36, 16 Nov 2005 (CST);

Comments

Dynamic Structures

[danenberg.livejournal.com]

Did I miss your message, Gustavo? In any case, WikiTeX can handle any backend with a CLI, not just LaTeX; would you mind elaborating on what you mean by "dynamic structures?"

Best, Peter

Re: Dynamic Structures

[(Anonymous)]

Hello Peter,

Glad to hear from you.

By "dynamic structures" I mean what Ted Nelson means by "not paper-like". PDF documents are paper-like, except (I think) for the fact that you can follow links from them (although this may depend on the viewer).

So by "dynamic structures", I mean something that you can interact with, e.g. things that have multiple possible visualizations (like a file tree whose branches you can expand by clicking on the "+" next to each directory), or a 3D image that the user can rotate.

My idea is that we could encode such structures in a wikitext-like language. Their rendering would require some client-side action, through something like JavaScript.

Gustavo

P.S. I left the message at your user page. How did you end up finding this post?

Re: Dynamic Structures

[danenberg.livejournal.com]

> Their rendering would require some client-side action, through something like JavaScript.

Interesting; we're already exploring the use of alternate engines with IPA and Greek, where paper-like output shows its limitations.

Have you tried your hand at some code, or is it still in the idea stage?

> How did you end up finding this post?

Google, actually; hence the delay.

Re: Dynamic Structures

[gustavolacerda.livejournal.com]

Things are very much in the idea stage. I need to work these ideas out, before I can start doing something.

Suppose I want to support file interactively-visualizable trees (see comment above). Would it be a good first step to write a TeX package?

If I do write such an TeX extension, it won't be supported by PDF, since PDF is too paper-like for the purposes of the package, AFAIK.

TeX code using this extension would be better "compiled" by something like LaTeX2Html.


P.S. Don't you use LJ notification? Maybe I should be emailing you instead?

Re: Dynamic Structures

[danenberg.livejournal.com]

Would it be a good first step to write a TeX package?

Actually, I'd start with the wiki-list-syntax:

* Root
** Branch 1
*** Branch 1.1

since it's intuitive to wiki-users and sufficient for encoding a tree.

Now, we'd write a parser that takes in list-syntax and spits out an interactive tree using, say, JavaScript; then we'd use that JavaScript as an alternative to the paper-like output.

The parser could be written in any language you're comfortable with, so long as we can invoke it from within PHP or on the command-line.

In fact, some people already came up with an interactive tree (http://dhtmlkitchen.com/scripts/animtree/index.jsp) in JavaScript that we could use as a starting point.

Do you have the time or inclination to get started?

Re: Dynamic Structures

[gustavolacerda.livejournal.com]

Unfortunately, I'm struggling with getting my DSL to work on my Linux machine, so I probably can't get started yet.

I would like to avoid AnimTree, since it's not free software. This seems like a better idea.

If we can't find free software that will do Tree2HTML conversion for multiple kinds of tree visualizations, then it may be best to just start with a simple thing like below:

Here's the bulk of the JavaScript that need to be generated. You also need to include some .js file: