I already mentioned TreeSheets. The more time I spend with it, the more fascinating I find its idea. Here is an example what I like about it.  It was created with, you guess, yes, TreeSheets.

What I like about TreeSheets The idea Hierarchical Structured Compact presentation Intiutive GUI Easy to insert new cells Tricky ways to manipulate hierarchy My favorite: transposition left right <=> left right Formula handling More intuitive TreeSheets price qty total 2.5 3 1.5 * 4 2 6 10 6 9 sum 25 vs Other spreadsheets price qty total 2.5 4 10 3 2 6 1.5 6 9 Grandtotal: 25 Not documented yet

In this economic situation, everyone is interested in savings.  I am exploring here how you can save a few letters when typing.  A simple method exploited by shorthand and speedwriting systems is to find the most frequent blocks of letters (n-grams) and use an abbreviation instead.  The natural consumer’s question is, how can you save most?

Bigrams or trigrams? Suppose, the most frequent bigram (block of two consecutive letters) is th in English.  We could type 1 for it, so “the royal path” would be spelled “1e royal pa1″.  Suppose furthermore, the most frequent trigram (block of three) is ath in English, so using 2 as an abbreviation would give us the spelling “m2 is the p2″ for  “math is the path”.  As you can see, the longer block is abbreviated, the more letters are saved.  But longer blocks occur less frequently.  How could you save more, with bigrams or trigrams?

(See my previous post about n-back words)  I found only two in English: beriberi, hotshots. There are some more in Hungarian: elõkelõk, tisztiszolga, történetének.  It’s addictive, don’t read this post before going to bed.

I really can’t stop, I just checked if there are words with an amazingly repeating block of 5 letters.  None in English, only one in Hungarian: tisztességesség.

n-back is a method to test your short term memory.  According to some research, it may increase your fluid intelligence.  I’ve been playing with BrainWorkshop, a neat little open source program  for n-back.  It works, I’ve already become smarter.  My proof is that I invented a word puzzle.

The n-back value of a word is the number of letters repeated at the nth position, for example the 2-back value of dad is 1, because the last d is two positions from the first d.  A letter in bold shows a repetition.  The 2-back value of my name, Adam, is also 1.  Higher numbers can also play, the 3-back value of member is 1, while its 2-back value is also 1 (member).  Is 1 the highest n-back value for words?  No.  Some of my favorite ones, nonsense, contented, stomachache all have 3 as their 3-back value.  Or the simplest one, tartar, which is nothing else but repetition.  They look easy, because its a whole block of letters that are repeated.  But there are trickier ones, like fulfilling where the repeating letters are interspersed.  Which English word do you think has the highest n-back value?  I don’t know.

A key element of Autofocus is to pick a task from a list subconsiously, intuitively.  Or how Mark puts it, choose the task that stands out.  There are pages in my system that keep remaining active,  I have difficulty to choose what stands out, but I’m also reluctant to dismiss the whole page.  Here is how I managed to (subconsciously) work around this.  I scan the list, read each task, maybe pick one that doesn’t feel right.  Then I go out for something else, make a tea, make a pee, whatever.  And I notice my mind is busy with an item on the list, not the one I chose, but another one.  That’s my intuitive choice.