Finding things
EtherPK gives you three ways to get to something, and they are deliberately different sizes. Quick find is the small one in the sidebar for when you know roughly what the page is called. Search is the big one for everything else - including finding words inside your notes. All documents is for browsing when you do not have a name in mind at all.
Note: Search and drafts are new. If you have used EtherPK before, two things have changed: clicking a link to a page that does not exist no longer asks whether to create it - it just opens, and creates nothing until you type - and quick find now offers ideas you have linked to but never written.
Quick find - when you know the name
The box at the top of the sidebar. Start typing and it offers matching pages and journal entries as you go. It matches names only, and it is forgiving about how you type them:
- Case does not matter.
physicsfinds Physics. - Any part of the name works, not just the start.
mechfinds Quantum Mechanics. - Aliases count. If a page is also known as QM, typing
qmfinds it and opens the page it belongs to. - Nested names read as plain words. A page called
[[Physics]] Quantum Mechanicsis found by typingphysics quantum- you never have to type the brackets.
Use the up and down arrows to move through the list and Enter to open. Escape clears the box.
If nothing matches what you typed, the last row offers to create a page with that name.
Search - when you know a word, not a name
Press Alt+F (or Cmd/Ctrl+K), or click Search everything under the quick find box. If you had already typed something into quick find, it comes across with you, ready to edit.
Search looks in two places at once and shows them stacked:
Names - the same name matching quick find does, over pages, journal entries, aliases, and concepts you have linked to but not written yet.
Text - every mention of your words inside your notes, grouped by the document they are in, so you can see at a glance which pages talk about something most.
The two run separately, so names appear more or less instantly and text results fill in a moment later. Each has its own paging controls, so a long list of one never buries the other.
Use the up and down arrows to move through both groups as one list - the highlight scrolls to follow you, so the last result is always reachable - and Enter to open. Opening a text result takes you to that exact line and puts your cursor at the start of it, scrolling there if it is further down the page.
Documents that mention your words most often come first, and each shows how many mentions it has. Every mention shows the line it is on, with your words picked out, and - when it sits inside an outliner list - the bullets above it, so a line like "yes, but only above 40°C" still makes sense out of context. Click one to open that document at that exact line.
A few things worth knowing about how text matching works:
- It matches whole words, and the last word you type is treated as unfinished - so
orphfinds orphaned while you are still typing, butgraphwill not find subgraph. - Put a phrase in quotes to match it exactly:
"quantum mechanics"finds the two words together rather than scattered. - Links count as words. Searching
physicsfinds it whether you wrote it as plain text or as[[Physics]]- you never need to think about which. - Everything else you type is just words. There is no special search syntax to learn and
nothing to get wrong: punctuation is ignored, so
C++,don'tandorphan - assetall simply search for the words in them. (For proper structured querying - "every unfinished task due this week" - that is what query blocks are for, not this box.) - Text results need at least two characters before they start - two characters as you typed
them, so
C++searches happily even though only theCis a word. Name results begin from the first character, so the modal fills in from the top as you type.
Search covers the graph you have open, and only that one. Opening a result closes Search; reopen it and your query is still there.
Encrypted notes are never searched
Anything you have encrypted stays out of Search completely - neither its words nor its scrambled form are ever put in the search index, so there is nothing to match even when the content is unlocked in front of you. Encrypted page names are still found by name, as they always are; only the content is out of reach. When a graph holds encrypted content, Search says so beneath the text results, so a search that finds nothing never quietly misleads you.
All documents - when you just want to look
The All documents button at the foot of the sidebar opens a full list of everything in the graph, with a filter box. It is the one to reach for when you do not have a name or a word in mind and simply want to see what is there. It is the same name matching Search does with an empty box - the difference is that All documents is a panel you can dock and leave open beside what you are working on, rather than something that closes when you pick a result.
Finding things you have linked to but not written
When you write [[Kanban]] in a note, that idea becomes part of your graph straight away, even
if you never make a page for it. Search and quick find both offer these, marked as not yet
written and showing how many notes mention them.
Opening one takes you to a draft: an empty page you can read and type into, showing every note that links to it underneath. Nothing is written to your graph by looking - the page is only created the moment you type something into it. So you can follow a link, see who mentions it, decide it was not what you wanted and leave, and your graph is exactly as it was.
This is also what happens when you type a name into quick find that matches nothing and press Enter: you get a draft, not a page, so a typo costs you nothing.
A draft takes its title from however you have most often written the idea. If eleven notes say
[[Kanban]] and one says [[kanban]], the page is called Kanban whichever one you
followed - so a new page never ends up disagreeing with the notes pointing at it.
Renaming an idea that has no page
A draft's tab offers Rename… like any other. Because there is no page to keep an old name
on, there is no alias to offer: renaming rewrites every link to the idea, at any depth, so
[[Physcis]] in forty notes becomes [[Physics]] in forty notes. The dialog says how many
notes that is before you confirm. If the new name already has a page, those links simply point
at it from then on - nothing is merged, because there was only ever one page.