One scientific (occasional) blogger has a few amazing posts and videos about using tinderbox in a scientific thought process.
I’ve been borrowing some ideas from her blog, especially about ‘conversations’ with notes which improve them by updates and links over time. It made me wonder if there might be an at least superficial way to measure one note’s ‘value’ over time, compared to other notes.
Let’s say we simply use the number of times the note has been touched. Even if we were to simply increment some attribute counter every time a note is ‘hit’, then over time we can see which notes are ‘improving’ while others are becoming less ‘interesting’. Maybe that would even help us better ‘weed the garden’ in occasional reviews. I think the hyperbolic view is especially helpful here in seeing how to improve a note…
but maybe we want something at a simpler level that we can sort… then choose out of 1000s of notes we wish to improve the ‘value’ of, and weed the ones which are simply old, way out of date, or have become useless.
Has anyone thought more about this? Any ideas? Thanks for reading and any consideration.
John
and just realzing that the blogger mentioned is also a consistent contributor here… Probably for the 2 people in the world who already haven’t watched her videos now you know where they are
I believe there was an example of this in one of the Tinderbox Weekend videos / examples back when creating dashboards with Tinderbox was a thing. I’d have to look back at my archive.
Of course $ReadCount is purly mechanical. For quality measurement it might be better to create a numeric attribute that you could update manually as a scalar (say, 1 to 100) that reflects the pleasure or interest or confidence (etc.) you have with that note when you visit it. That, combined with the delta between $Modified and $Created for a note could give you an indication of how long you have been cultivating that note.
thanks… I just found $ReadCount and $SelectionCount… I agree that these are very close to what I was describing.
And yes, I think a ‘value’ that becomes some kind of weighted ratio between attributes (# of links, # of reads, # of children, time from last modified etc) could make a great ‘value’. I think I have enough to start this.
Now to try to work on it without excluding doing ACTUAL work. Now that would be a great metric. Ratio of ACTUAL work / tinderbox tweaking
ISTM, the hard point for the code is to figure if you intentionally looked at a note or just selected it for any one of plenty of incidental/accidental reasons. I’d treat the count with a degree of caution - not because the count is wrong but because the count accrued may reflect other factors than the user assumes.
