RC2 on paper, part 4. Pages 14 to 19, graphs and Where do you land?, where the pen asks the book to draw: sixty notes, two new themes, Wardley maps before graphs, and one page that holds every note so far

In short

• Pages 14 to 19: chapter 3, graphs, and the start of chapter 4. Sixty notes, 84 to 143.
• Draw it. "make the (visual) case of why graphs"; nine notes ask for a picture where the page has a paragraph.
• Wardley maps first, then graphs. The answer to the For Dinis box on page 15; and the presentation is graphs.sgit.ai.
• The voice, again. "DEMEANING", "harsh", "too harsh", "weird", "meaningless is too strong".
• Too early. Wardley maps, refactors, ILC and the graphs book are used before the reader has met them; code as a graph goes later.
• What works. Connections as the way in; "It does not matter where you start; it matters that you start." as a theme for the whole book; the three personalities.
• Every note, one page. RC2 on paper: every page, every note: 19 pages, 143 notes, by theme and by chapter.

Every page, every note

With this batch Dinis asked: "are you tracking all pages so far in a ux that allows the easy viewing of all comments and mappings?" The honest answer was no. Each batch had its record and its article, and the only way to see note 12 beside note 120 was to open two articles. So before this batch I built the page that was missing.

RC2 on paper: every page, every note has three parts. The pages: a card for each photograph, with the chapter it is in and its notes, opening the callouts. The map: a table of chapters by theme, so you can see, for example, that chapter 3 is mostly about the voice and about meeting the reader where they are. And every note, in one table, with its own address (/reviews/paper/index.html#n120), the words as written, what they mean, and a link to the article: filter by theme or by chapter, or search. The same notes are data in notes.json, for the editor and for any agent working on the next candidate. The build refuses a batch without its article, and notes that are not numbered one to N across the batches, so the page cannot quietly lose one.

Part of a series

The fourth article of RC2 on paper: part 1 (pages 1 to 3), part 2 (4 to 9), part 3 (10 to 13). The record of this batch is admin/content/reviews/paper/rc2-04.json.

Page 14

Above the chapter's title, the job: "Basically here we really need to make the (visual) case of why graphs & why they are so powerful & why they are everywhere". Then the voice. "I have met people for whom a graph is a picture with boxes": "it's not their fault, the industry terminology is all over the place", and on the left, "Reword, too harsh". "A node connected to nothing is literally meaningless": "meaningless is too strong", "each edge is a signal". "Easier than my generation did": "DEMEANING". Down the left margin, a long note: "never a chart, never a figure" is wrong, because "a graph node can be anything, since a chart is a graph too"; see thinking in graphs, or fractal semantic graphs. The graphs book is "not relevant at this stage of the book"; the three meanings of graph "needs rewrite"; the 2018 outline is struck. The For Dinis box asked which presentation on graphs the outline meant: "see graphs.sgit.ai". And of "In Follow the rabbit holes I said": "these kind of references are weird since the reader has not read them".

Page 15

"See article and entire section in graphs.sgit.ai about thinking in graphs"; and the graphs book, again: "reader doesn't know about this book". "That line is for you too": "harsh". The graph of your field in your head: "we need to show lots of examples here", and "also show here why some of the biggest companies in the world are based on graphs". Then A map is a graph with positions: "we need to qualify this better", and "reader has no idea what a Wardley map is (by page 15)". The talk's shifts are "not relevant to the reader"; the definition of a Wardley map, "show this visually"; shadows of reality, "yes, and show example", "visualise this". The For Dinis box asked which comes first; the answer runs down the right margin: "First we need to show the Wardley maps & how to read them, then go into graphs". "That I now cannot unsee": "weird". And for graphs that nest: "First show example: reality, systems, nature".

Page 16

The page the pen likes most in the chapter. Fractal semantic graphs "is a good example that should make the reader think & ideally make a paradigm shift": "Focus here and explain it", and "visualise this". The world made of connections, groups of groups as the way we think: "this is a good way to start explaining graphs, connections". Smaller things: "finding" needs a better word; "A refactor is" needs defining. The honest answer on AI-generated code: "rewrite", with the newsroom's sgit example struck. And the whole of Code is already a graph, bracketed: "good but for later in the book since most ZX founders have no idea what we are talking about here; on this topic see what we did at secrets.sgit.ai".

Page 17

Reading code at the altitude you need: "show example of this working in practice". Then the line the pen wants everywhere, "It does not matter where you start; it matters that you start.": "this needs to be a common theme across the book". Pick a node and follow an edge: "yes but they need a place to put those nodes & edges". The From the register box: "what is the context (& audience) of these sections". The hands-on chapter Your first graph, which the page says you do not need yet: "this should happen now & we should be using prompts & QR code to examples & the text of the prompt". And a question, under it: "what about if every page had a QR code to that page (& version) online? or section", whose visits "will drive revenue".

Page 18

Chapter 4 opens by asking the reader to draw a map without naming it: "visualise this". On the right, how to build it: "first we got the reader to draw a graph, top down; then we add the evolution, Wardley map". The models moving right: "this needs to be visual, maybe even with a storyboard, asking the questions the reader will have". "Gave you the graph": "??? what does this mean". The four stages: "show other types of evolution". Three sentences are struck. Then two arrows. Down the left, from the paragraph on why Dinis cares and his 2022 talk: "Don't start with the where". Down the right, to "Wardley Maps gives me inner peace": "start here". And before the four stages: "explain ILC, movements & attention".

Page 19

The rows of evolution: "visualise this". "Once you see it, you see" becomes "experiment". "Where do you fit?" is struck, and the left margin offers a new heading: "Let's meet the cast", or "players", or "personalities". "The question I most want this book to ask you": "harsh". Wardley's names in 2015 and since 2023: "this evolution is not relevant". "What I can tell you is what I have seen" is struck, with "visualise comparison". Down the right margin, the picture: "visualise: the real world, explorers of a new land, followed by the villagers, followed by the town planners; is a good one, as population grows". And at the foot, of the three descriptions: "these are great :)".

The themes, after four batches

Two themes are new. Draw it: where an idea is a shape, the book draws it (84, 103, 105, 110, 123, 125, 134, 141, 142). Answers to For Dinis: the pen answering the candidate's questions (92, 106). The table of every theme by chapter is on the hub; in this batch the voice leads (14 notes), then drawing it and meeting the reader where they are (9 each).

The pattern under them is one note, said many ways: the reader is on page 14, and the page uses words, books and talks the reader meets later or never. Wardley maps before graphs is the structural answer; a line of explanation, a picture and a QR code are the local one.

What Dinis answered

• The presentation on graphs the 2018 outline names: graphs.sgit.ai (note 92).
• Graphs or Wardley maps first: "First we need to show the Wardley maps & how to read them, then go into graphs" (note 106). Chapters 3 and 4 swap in the next candidate, and the map is built in steps, as note 124 describes.

Dinis then read the four uncertain notes back to me, and said what he meant:

• A chart is a graph too (notes 85, 86, 90). "My use of graph" is right, and points at the circled nodes and edges, the way Dinis uses the word. In ordinary use a graph means any diagram, so it is not the reader's fault their definition is different, and the book should not make them feel guilty for it; it should give more analogies of what a node and an edge are. "Never a chart, never a figure" is too brutal, and wrong: in fractal semantic graphs anything can connect to anything, so a node can be a chart or a figure, and a chart is itself a graph, two-dimensional, but still nodes and edges. That is why the note points to thinking in graphs and fractal semantic graphs: meaning comes from connectivity, and the value is in what is on the other side of an edge.
• The map in two steps (notes 124, 125). First the reader draws a graph, top down: this leads to this leads to this, a kind of tree. There the positions mean nothing; you can move the dots around. Then the evolution is added, and the position of each dot gains significance. That is the moment the reader has a Wardley map. The line I had joined to note 124 belongs to note 125: draw it, perhaps as a storyboard that asks the questions the reader will have.
• Don't start with the where (notes 129, 130). "Where" is right. The book keeps saying where something was said (I did this, I talked about this, here is the article) before the idea, and the reader has not read any of it; it reads as Dinis validating himself. Start with the concept, "Wardley Maps gives me inner peace", which is where the "start here" arrow goes, and put where it was said behind a QR code, if it is relevant at all. The editor should ask that question each time: does the reader need to know where?
• Meet the cast (notes 139, 142). "Let's meet the cast", or the players, or the personas: introduce the explorers, the villagers and the town planners first, then show them, with examples and how they connect. The picture down the right margin is the settling of a new land: the explorers arrive; as more people come, the villagers build the village; as the village grows, it needs town planners, with rules, laws and roads. The population growing is what makes the analogy work.

The strip

A new land, six panels, is note 142 drawn: Zara arrives first, explorers of a new land; followed by the villagers, Priya; followed by the town planners, Tomás; as population grows. The pacing experiment this time is accumulation: every panel keeps the people of the panel before, so the reader feels the land fill up instead of being told.

What I need from Dinis

The four readings of this batch are answered above. Two remain from part 3:

• Note 61. Page 11, right margin, about a third of the way down, beside the list of transformations: "good", then, as I read it, "format better".
• Note 78. Page 13, left margin, a little above the middle, beside the paragraph on how to use the rest of the book: "Reword from an '...' approach"; I read "adult first".

What this does not claim

• The reading is mine, word for word; the four uncertain readings were confirmed or corrected by Dinis, and the record says which.
• The regions on the photographs are placed by eye. A box shows which note is which, not an exact boundary.
• The hub counts the notes, not their weight. "visualise this" counts as one note, the same as the paragraph that set the chapter's job.

Where this comes from

Dinis's six photographs of 19 October 2026, pages 14 to 19 of the printed release candidate 2, and his question about tracking every page. The candidate's text is v2/versions/v2.0.0-rc2.md. Written by the journalist, in its own voice, as decision 007 asks.