How to Use

How to use Textmind

The chronological ramblog

A ramble is a level-one heading in ' It is a strict chronological log of all the day's thoughts and occurrences, divided by timestamps. Each timestamp has its own line, like so: [2020-01-22 Wed 00:20]

The ramblog is lightly outlined, but the outline hierarchy is not strict. Chronology takes precedence, and pivots or digressions are encouraged, regardless of damage to the outline structure. A strict outline is maintained in ' to support work in ', so it isn't necessary to maintain a strict outline in '

Processing round

When to

While the ramblog is short, it's easy to remember and use its contents. But then it gets too long, or too much time passes, or there's a big pivot. As a result, using the ramblog causes mental friction. You waste time scrolling, looking for info. So it's time to do a processing round.

When is the best time for a processing round? Well, memory decays exponentially, and memory consolidation (including forgetting) occurs during nightly sleep. Spaced repetition is based on the principle that reviewing at spaced intervals stores info in long-term memory. Therefore the natural Textmind rhythm is to do one ramblog per day, and process yesterday's ramblog in the morning.

Ramblog processing decomposes yesterday's ramblog into topical headings suitable for filing in Textmind. This is a good way to review material and make decisions for today.

How to

Follow the ramblog processing steps on the ' checklist. To summarize:

Duplicate yesterday's ramblog. File one copy in the journal directory. Destructively analyze the other into tasks and background information. Then file the atomized headings appropriately, to depth=1. Prioritize any new tasks.

Video demo below:

Checklist steps

Here's each step of the ramblog processing checklist, in detail:

(find-file "~/1-mansort/1-Textmind/2-Linked/Hud/")

Run C-x C-e eval-last-sexp on the above line to open Dired to the directory that will contain the ramblog copy to decompose.

- [ ] proc rambled & reorient
  - [ ] (org-reset-checkbox-state-subtree)

This clears all boxes in the checklist.

  - [ ] (leo-create-new-rambling)

This creates a new rambling heading in

  - [ ] git commit

Fumble-fingers might delete the ramblog somehow in transit. Safer to commit before proceeding.

  - [ ] double file prior rambled

This step is too important to automate. I'm afraid some unexpected edge case could cause undetected data loss, creating a mess. Better to do it manually so I can spot any mistakes.

  1. Demote the old rambling heading from level 1 to level 4.

  2. Kill the heading and open Hud/'

  3. Insert the heading under two level-3 headings: "Log copies: send to /4-Time" and "Work copies: send up with H-g"

  4. The contents of the first level-3 heading are batch exported to /4-Time on demand. No further action now.

  5. The contents of the second level-3 heading are immediately sent up to an with leo-up.

  - [ ] deletion

Delete everything obviously extraneous from the ramblog. Go from bottom to top, because it's easier to tell what's obsolete if you already know how things ended.

Don't think hard at this step. Any difficult questions will be resolved in later steps.

Deleting too little doesn't matter. Just don't delete too much.

  - [ ] headingfy

Insert headings wherever appropriate. Divide unlike things.

  - [ ] lazy title and quick-publish

Name the headings. Quick-publish anything relevant to social media or blogs.

  - [ ] promote

Promote any wrongly-subordinated headings.

  - [ ] file

Refile headings into Textmind's tree.

  - [ ] file to depth=1

File headings one level deeper.

 - [ ] sort tasks
  - [ ] file agenda files to taste
  - [ ] (treefactor-refresh-org-search-scope)
  - [ ] rank #5 tasks - search org agenda for #5's
  - [ ] consolidate tasks

File and prioritize Org Agenda tasks.


Pick a task from the Org agenda and execute it as usual, using ' and '

If necessary, use a checklist to help decide the next action. This can ensure one remembers to e.g. check the day's appointments before perusing the Org agenda task list.

JIT sorting

One may need to sort the Textmind tree to complete a task. Never sort just for the sake of sorting. Always have a task that narrows focus and resolves ambiguity. In other words, use "Just-in-time" sorting.

If you can't choose a next task, then create a task, "Figure out what to do next," and sort until an answer appears. Even an open-ended question is better than purposeless sorting, and provides a surprising amount of clarification for ambiguous sorting decisions.