Repository navigation
GraphPaper 0.4: preserve authorial intent and add Polemic - #7
Conversation
Codex Review SummaryThis comment shows the latest Codex review activity on this pull request.
ℹ️ About Codex in GitHubYour team has set up Codex to review pull requests in this repo. Reviews are triggered when you
Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings. |
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 9c5230027c
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
| const polemicPreviousReview=reviewHTML; | ||
| reviewHTML=function(){let value=polemicPreviousReview();const a=state.p?.review?.authorial_assessment;if(a&&Object.keys(a).length)value+=`<div class="evidence-card mt20"><span class="caps">Authorial fidelity</span><p class="hint mt12">${esc(a.summary||'The editor assessed the intended argument separately from factual support.')}</p><div class="row wrap mt12">${[['intent_preserved','Purpose'],['stance_preserved','Position'],['voice_preserved','Voice']].filter(([k])=>typeof a[k]==='boolean').map(([k,label])=>`<span class="chip ${a[k]?'green':'gold'}">${label}: ${a[k]?'preserved':'needs attention'}</span>`).join('')}</div></div>`;return value;}; | ||
| const polemicPreviousAction=handleAction; | ||
| handleAction=async function(el){const a=el.dataset.action;if(a==='writing-mode'||a==='choose-polemic'){if(!editable())return;await flush();modeDialog();if(a==='choose-polemic')$('#writing-mode').value='polemic';return;}if(a==='apply-writing-mode'){await flush();const mode=$('#writing-mode').value;state.p=await api('/projects/'+state.p.id+'/writing-mode',{mode,version:state.p.version});syncSummary();editorialUI.statusKey='';closeModal();render();toast('Writing mode changed. Your existing work is preserved.');return;}if(a==='save-brief'&&$('#b-thesis')){const force=Number($('#b-rhetorical_force').value);if(!Number.isInteger(force)||force<0||force>100)throw new Error('Choose rhetorical force between 0 and 100.');state.p.brief={...state.p.brief,thesis:$('#b-thesis').value,stance_policy:state.p.mode==='polemic'?'preserve':$('#b-stance_policy').value,rhetorical_force:force};}return polemicPreviousAction(el);}; |
There was a problem hiding this comment.
When a Science project is on its default research tab and the user switches to Polemic, Nonfiction, or Fiction, this handler keeps state.tab === 'research' and immediately calls render(). The non-Science projectHTML() has no research entry, so it destructures undefined and throws, leaving the interface stuck in the old Science workspace even though the server has already persisted the new mode. Set the tab to a valid standard-workflow tab before rendering when leaving Science.
Useful? React with 👍 / 👎.
1 participant
Fix the editorial workflow, not only its label
Adds Polemic as a fourth mode and a reversible mode switch for existing projects. Removes forced counterbalancing from the general writing boundary; nonfiction can explicitly argue a case too. Author voice and a shared authorial contract now reach graph-candidate evaluation, angle generation/scoring, outlining, drafting, review, revision and prose polishing. Factual premises remain checked separately from moral judgments, satire and rhetorical force.
Core thesis, purpose, rhetorical force and evidence detail are independent. Stale generated angles are marked and can be refreshed without rebuilding the graph. Short approved argumentative outlines are drafted as one coherent whole instead of restarting the same caveat per section. Reviews report stance/voice fidelity and treat unwanted softening as meaning drift. Both opposing author positions use the same fidelity policy; no opinion is hard-coded.
Final validation
Verified Windows release artifact
GraphPaper-v0.4.0-Windows-x64.zip— 187,529,591 bytes.SHA-256:
c9fb03c6b293b8505b14ec427c7266aba1e3be8f0391be35ecead18a91399b53.Reports:
docs/validation/v0.4.0/. The 221-file publishing manifest was verified against staged Git objects. Runtime binaries are distributed as release assets, not committed alongside user data. These are actual local Windows/native results; hosted CI platform results remain visible separately.