Zettel ID

Threepwood

I have read several comments* in the forum pushing back against the idea of the Zettel number in note titles. However, I was ready to accept some minor inconvenience for the advantage it brings.

Accordingly, I renamed all my existing notes to add a Zettel ID based on file creation date, using the app "Better Rename 9". After doing so, I checked & found about a dozen duplicate IDs. Not sure how they came about, but I've noticed before that file meta data can be unreliable, so duplicates were not surprising. However, it was easy enough to locate & correct these errors.

A little later, I happened to check my notes (stored on Dropbox) in 1Writer on my iPad. Because the Zettel ID is long, and 1Writer allows a fixed space for it's title list, it was difficult to find the note I wanted.

So, I have hit two problems related to Zettel ID. This got me thinking - are there any other issues which might cause trouble ? It didn't take me long to find another scenario in which Zettel ID could be duplicated:

Let's say I'm a playwright, and I have just typed a list of the characters in my play. I decide I want a separate note for each one. Luckily, creating notes is really quick, so I can create several in one minute - all with the same ID.

So, does all this add up to a real problem with Zettel numbers ?

I think that Zettel numbers are a problem for two important constituencies:

Some of the most likely customers for 'The Archive' will already have a collection of notes. If they add a number themselves, some will introduce duplicates (even if they simply use a shorter, sequential number instead, it could easily match with a section of an existing real Zettel number)

It would make sense to encourage these 'customers' to join the ecosystem by providing a method to batch add Zettel numbers to pre-existing notes within 'The Archive'. But, under the current system, this too is likely to introduce some duplicates.

The second problem is more troubling. I suspect a sizable number of serious note-takers store their data in a cloud service, and sometimes want to address them via a mobile device. 1Writer is one of the best options to do this - but it doesn't play nice with Zettel numbers. I don't know of a mobile app which would.

The fact is, Zettel numbers are much longer than needed to guarantee uniqueness - and, even so, it is not a reliable guarantee.

I therefore tend to feel that the 12 character Zettel number is an unnecessary handicap - it needs to be shorter. Selling the idea of a Zettel number to a wider audience will require the least possible friction in it's implementation.

Here are a couple of alternative suggestions:

a simple 6-7 digit sequential number

An 8 digit hybrid number: YYMM followed by a 4 digit sequential number
(This combines the advantages of some date information, plus the guarantee of uniqueness which sequential numbers can provide)

I apologise for raising this matter again when clearly a lot of thought & discussion has already gone into it - but I am otherwise so delighted with 'The Archive', that I really want it to succeed.

Geoff

• https://forum.zettelkasten.de/discussion/61/writing-index-notes-for-your-saved-searches

  https://forum.zettelkasten.de/discussion/comment/921/#Comment_921

ctietze

You summarize good points here, and I took note of your ideas to help users "migrate" their regular notes to unique Zettel ID-based ones!

For practical purposes, you could of course experiment with:

• putting the ID at the end of the file name (increasing visual noise, though, because the numbers don't line up anymore)
• leaving the ID out of the file name, putting them into the note body only as a header (both The Archive and 1Writer will find them thanks to the full text search; of course you'd render filename base shell scripts useless, if that's a concern)

Rehashing the argument against sequential numbers: they are not "state-less", i.e. they cannot be created without knowledge about already used-up numbers. (You can argue that the date is just as much a variable as the numbers that are already in use; but in practice, using the date to produce an ID is trivial on all operating systems, while the sequential number isn't.)

If our numbers increase, maybe we can pitch the idea to hide ID prefixes to the very helpful and responsive 1Writer developer

Threepwood

Thanks for your comments - and your patience. I notice that the 1Writer developer has already provided several different formats for his list view. How hard can it be to provide one more ?

Geoff

rrieke

One of the things I like about Markdown is that it's human-readable. Someone with no knowledge of HTML or any other documentation markup could see a MD document and pretty well understand what the extra characters mean.

For the same reason, I initially started with my Zettel IDs at the end of the filename, like /FileName-20180403155000. In my mind, it lets me wander through my note list in a friendlier, more "recreational" way. It encourages aimless perusal (fun and sometimes idea-stimulating) as well as intuitive visual navigation.

But this morning I finally understood a little more about how our minds are so date-friendly (see Christian’s post,) and so I'm trying it with the ID first. I figure it won't be a bear to rename either way, if it comes down to it.

tonari

ok, so the best way to handle this is to have seconds included, since it is almost impossible to create multiple notes in the same second. Even if you could, the difference in time is negligible.
it creates a perfect "unique" name, and is significant (i.e. meaningful).
I usually scan documents with this file name format, and just rename by adding a meaningful prefix...
also, it is a nice feature that if you want to just type a search "name" and then hit enter to name the new note... skipping the "zettel id"...

this is best.