A paper is not one page and it is not only the draft. Around the draft sit the things you need while writing it: the marked-up manuscript your editor sent back, the journal's guidelines, a sample paper you are matching, the interview transcripts, and the thoughts that outgrew this paper and belong to the next one. An academic who tested MojoPad put a number on it without meaning to — in his own project the draft was eight items out of twenty-eight. The paper is the smallest thing in the workspace.
So MojoPad does not ask you to make a "project". A paper is a folder. Put the sections in it in the order they will be read; put everything else in folders beside it. Nothing is special, which is the point: the same tree, the same pages, the same search.
View ▸ Writing View arranges the app for exactly this job and nothing else. Your folder tree on the left, the section you are writing in the middle, and down the right, about that section:
The strip follows the cursor: move to another section and it is about that section instead.
Everything above this point a dedicated writing app has had for twenty years. This is the part none of them can do — because none of them is also where your reading lives.
As you write, MojoPad reads what is in front of you and offers what else in your wiki belongs beside it — notes, quotations, pages, and the PDFs and books in your library, because their text is read and indexed too. Not by folder, not by tag, not by anything you remembered to file: by meaning. The list changes as you move between sections.
The academic above described the problem before he had seen the answer: "later, when working on another paper, I remember writing notes that could be used, but I cannot remember which app I had written them in." That is the tax every researcher pays for notes spread across apps and folders and years. Here there is nothing to remember, because there was nothing to file.
Click one and it opens beside your draft, not instead of it — the split pane, with your sentence still where you left it. This needs a local embedding model; MojoPad tells you in the strip if one is missing, and nothing leaves your Mac either way.
Two panes are not always enough — translating a quotation wants the source, the draft and the translation. Right-click any page ▸ Open in Small Window, then click the ✎ pen in that window's header: the card becomes a full editor, with the same formatting controls and the same Find and Replace as the main window. The eyeglasses turn it back into a clean reading card.
Each card is its own window, so open as many as you like and put them where you want — across two displays if you have them. Drag one wider and the formatting bar uses the room.
Windows are one answer; the other is more columns inside the one window, which is what a very wide or very tall display is for. Right-click any page and choose ＋ Add a Pane: it appears beside everything already on screen, with its own ✕ to close it and its own ⤢ to send it out to a window. Add as many as the screen fits.
How a page opens, everywhere. A wiki link, a row in a database view, the header of a pane, and a page picked out of a list all follow one choice in Settings ▶ Links & Pages: open in the same view, or open in a new tab. Hold ⌘ while clicking to do the opposite once. Two things sit outside that choice on purpose — a stored paper always gets its own tab, because a document replaces the whole reading surface and following a citation is looking something up rather than leaving; and anything arriving (a clip from your browser, a file imported, a page fetched from a URL) always opens beside what you were reading rather than over it.
Sizing them. Every pane has a divider on its left edge, the same one the reading pane and the sidebar have: drag it to make the pane wider or narrower, and double-click it to put it back. The width is remembered, so the next pane you open is the width you like a pane to be.
This is a different button from ⇄ Open Beside This Page on purpose. That one is the reading pane — there is one of it, it swaps with the editor, and it is the pane that holds a PDF, a book, a video or a canvas with the controls each of those needs. An added pane holds a page: a title, the words, and the two buttons. Ask one to hold a document and it will say so and offer you a window instead, rather than showing you something that cannot work.
How many, and what happens when the screen runs out. There is no limit. Panes narrow as you add them, down to a width prose is still readable at; past that the row scrolls sideways, and the page you are writing on always keeps its own room. Ask for a pane on a PDF, a book or a video and it opens in the reading pane instead — that is the one with the controls a document needs.
Asking for a page that already has a pane gives you the pane you already have rather than a second copy of the same words. Panes are live — edit the page in the editor and its pane follows — and they last for the session rather than being remembered between launches.
Everything a paper needs is here, though it is taught in the Properties chapter because a reference is a page with properties on it. In short: File ▸ Import ▸ Import References (BibTeX)… brings a library over from your reference manager; Format ▸ Cite a Source… drops a citation where the caret is; Insert Reference List puts a bibliography at the end that keeps itself in step with what you have cited; and Caption This Picture or Table plus Refer to a Figure or Table… number your figures and keep the cross-references right when you move them. See Citing as you write and A reference list that keeps itself under Properties.
Compile walks the folder in order, so the binder order is the paper's order: front matter first, sections in the middle, references and any closing abstract last. Depth decides heading level. See Compiling a Long Document.
Things that must not appear in the output have three ways out, and it is worth knowing which is which:
Compile writes a new page in the wiki, a PDF, a Word file, Markdown, or plain text. For a paper the PDF is the one that matters, and Settings ▸ Files & Backups ▸ Paper sets page size, margins, line spacing, typeface and page numbers so it comes out looking like a paper rather than a printed web page.
Which file you send depends on where you are in the process, and the order surprises people who have only ever submitted once. A paper goes out for peer review as a Word file — usually through the journal's own submission system, sometimes still as an email attachment — and it must meet the submission and formatting guidelines or it is rejected before anyone reads it. Review is normally blind, so that copy has to be stripped of your name and your affiliation. Only at the very end, once the reviewers' changes are agreed, does the editor want a ready-for-print PDF — a page they can send straight to the printers, which is increasingly the author's job rather than the publisher's.
So both matter, at different moments: Word is what gets you read, and the PDF is what gets you printed. Set the paper up once and Compile produces either.