Some thoughts on transliterature

by Jack Seay - Oct. 3, 2005 - freely distributable

Display and edit the following lists:

Content list

{url: start, length} A document content address and a span of text in it.



Format list

{document address}

{start, end: bold} External markup applied to a content list.

{start, end: italic}


Dimension list - Lists connecting the other lists

{content} Text of a document, can include transcluded portions of other documents.

{version} What edition of a document

{format} External markup

{view} 2D, 3D, mindmap, outline, etc.

Version list

{base document}

{version 2} - pointing to a content list

{version 3}


To display and edit documents:

Point to a content list

Display it's dimensions, which point to versions, formats, views, etc.

Choose a format, version, and view type

The display program keeps everything linked to it's origin, so any copying, moving, pasting, outlining, formatting - creates new lists that point to the original documents. You don't point to D, which points to C, which points to B, which points to A. You just point to A. Every revision points directly to the first source of everything contributing spans of text to the current document.

Version lists would not contain content, just pointers to content lists: doc 1 is the base document, doc 2 is the second version of it, etc.

Special web servers could be set up that guarantee stable addresses for all documents placed on that server.

Normal files and http protocol would be used. All HTML, CSS, Javascript etc. would be stripped from the files before content is used. Only the text is wanted. All else is embedded markup, a useless and evil pollution to be purged; replaced by separate format lists applied to the content. Any editing programs must keep all markup separated from the content (invisibly). The user should see and edit the finished product, but the programmers need to be clever enough to always keep them stored separately. HTML web pages could be deconstructed into separate content and format lists. No need for the current bizarre combination of embedded HTML, internal and external CSS, server side includes, javascript, blah, blah, blah. All markup would be external. Why do I say embedded markup is an evil pollution? Because I have wasted large parts of my life trying to keep track of changes, trying to reformat documents for different end uses, then losing track of which formats have which version of the text. I have wasted thousands of hours attempting to learn today's incredibly complex methods of creating web pages (a grossly inferior and severely broken form of hypertext). Millions of other web developers endure the same creativity killing, unnecessary complications, wasting hundreds of millions of hours of valuable time.