Granularity of IDs
19/31 daily posts as part of WeblogPoMo2024. Expect (and forgive) more words and less editing.
While I was away last week, Lucy recorded herself creating her index. (We're still recording that section of the workshop.)
One of the items she added was a doctor's referral letter. My natural instinct would have been to create a very granular ID, something like:
12.20 Referral for scan for achy knee problem1
...or maybe broaden it to the subject rather than the specific action:
12.20 Achy knee problem
This is just how I've always done it. But Lucy did this instead:
12.20 Doctors and tests
That's much less granular than I would have done, and I really like it.
Because I don't need a super-specific ID for that scan. In the grand scheme of my life, does every little thing that I do need its own ID? No. I'd be drowning in them.
Lucy's answer fits better with an idea that we came up with while recording the workshop: IDs are like the manila folders that we show on the home page. When you sit down to do 'some work', imagine yourself opening that folder. Ideally, everything you need is in it, and not much more.
And imagine that they cost about a dollar. So you wouldn't create them willy-nilly, but nor should creating one be something you consider too deeply.
Put the date in the scan's filename
The crucial extra step here, which Lucy did, is to name the scan so that it includes the date at the front:
2024-05-19 Referral for achy knee scan.pdf2
This way, folder 12.20 can contain an almost limitless amount of this sort of information without ever feeling cluttered.
If I'm lucky, this one folder will serve me for the rest of my life.
Officially endorsed
I really like this approach. I'll be doing it myself when I rebuild my own systems.
There is a follow-up to this post at
22.00.0052.