unbound_book - generative typesetting and visually semantic publishing notes

Aug 26, 2026

A set of personal notes outlining a thesis trajectory from programming as an open-source and auto-didactic practice toward generative typesetting and visually semantic document workflows. The material connects LaTeX/ConTeXt and WYSIWYM with WYSIWYG, emphasizing implicit semantics, external styling constraints, and the conditional production of both HTML and PDF. It also traces how early peer-to-peer systems like email and their emphasis limitations shaped later markup conventions (Markdown, MediaWiki) and publishing formats for essay-like writing.

LanguageEnglish

CategoryTechnology

FormatPDF

Pages6

Reading Time15 min

AccessFree to download

Introduction

Auto-didactic programming and tool ecosystems

Typographic considerations and LaTeX/ConTeXt

Thesis focus: generative typesetting

Typesetting as a self-referential case study

HTML and PDF as conditional outputs

Visual semantics and WYSIWYM

Email as an origin of textual subtext

Markup, readability, and semantic ontology

Publishing formats and practical limits

MediaWiki and essay fragility

No universal pre-format for typographic workflows

notes for unbound book john c. haltiwanger my.intro (in reference to Rakim, who said “it’s not where you’re from, its where you’re at” i’d like to propose a new variation: “it’s not who you are, it’s where you’re at”) auto-didactic programming skillset free software new media theory (easily one of the most important “open source”tools i’ve ever come across; i’ll cometothe topic of my thesis ina moment) r11y mindset (has anyone here heard of i18n? it originates from trying to solve a language barrier; thus my contribution ‘r11y’ -a shorthand for ‘revolutionary’; this is not meant that these or any of my thoughts are revolutionary, only that i like to concentrate on problematics with potentially macro-level consequences; any approach that is concerned with its own approach) + typography had known LaTeX as a common libre tool for some time without comprehending exactly what it was. my friend’s college institutionalized LaTeX for typesetting theses and he mentioned the typographical and, further, the referencing management advantages. later my sister showed me her PhD workflow in LyX. what you see is. . . WYSIWYM is an adjacent paradigm to the venerable and popular What You See Is What You Get. it is different in the formulation of its focus: implicit semantics over explicit choice. it abstracts from the underlying layer in an act of [re-presentation. in](re-presentation. in) the end it is feeding a typesetting engine that speaks a dialect of a programming language that is over 30 years old. this hybridization relies on external styles. this is not in itself a problem, but it turns out that extensive stylistic customization can be a tricky deal in LaTeX. i began exploring other options along the same lines and came across ConTeXt. i’ll be returning to the importance of this in a second. grammatic processes my thesis was a shift in focus from large-scale, peer-to-peer platform design based on social solutions to “piracy”. there were moments in the class that led to revelationsonthe integrability of programming tools with humanities essay writing. especially in a collaborative sense. geert lovink proposed a concept for a new kind of journal based on moving beyond old academic publishing paradigms such as traditional peer review and paper-focused issue formats. i emailed him some thoughts about how web interfaces tied to translation layers could potentially enable for experimental peer review systems and cohesive visual design. the next thing i knew my thesis was about generative typesetting. the ‘what’ is generative so i wrote a thesis whose case study was the typesetting of itself. the typesetting was conditional: as a primary order, in order to address its own conditions of relevance, the thesis was determined to be set in both HTML and PDF. such a methodology implied automation. the easiest option is a direct PDF rendering of HTML. which has no hyphenation. EEErrr! but even in the first place, who would write their thesis in HTML? (HTML is a machine-oriented semantic markup, making it non-trivial to attain either literacy or instinctual readability; due to time constraints i will not going togo into detail with the history of HTML and machine-oriented semantic markup;) email as an origin of visually semantic textuality email was the first time we truly went peer-to-peer. any node that had access to the internet could send and receive emails because the protocol was designed that way. (this, of course, led to heavy spamming.) email was also a system where one could not rely on physically derived cues for textual emphasis. for instance the repeated etching of ink or double stamping of typewriters could not be represented by the plain-text (read: ‘semantic free’) environment. [DO YOU REMEMBER ALL THE SHOUTING?!?!] this led to systems of expression which relied on textual clues to provide subtextual meaning. the most obvious are all-caps shout-outs, bane of the unintentionally

FAQs

How do these notes define WYSIWYM compared with WYSIWYG?

WYSIWYM is described as emphasizing implicit semantics over explicit visual choice, abstracting the presentation layer while feeding a typesetting engine that behaves like a programming language dialect.

Why is conditional generation of both HTML and PDF important to the thesis?

The thesis uses self-addressing typesetting conditions of relevance, choosing HTML and PDF as the dual representation targets, which implies automation as part of the workflow.

What role do email and early peer-to-peer protocols play in the argument?

Email is presented as the first true peer-to-peer system and as a text environment lacking physical cues for emphasis, leading to alternative textual signals like all-caps, emoticons, and underscore syntax.

Aug 26, 2026

Upload by Callum