Main Area and Open Discussion > General Software Discussion
I'm thinking of going primitive, with discursion into zettelkasten
Dormouse:
Vulnerable how? Like you link to the directory with the images in it, and that link breaks? Or they break when trying to sync between different systems (something I have had problems with)?
It just seems like that is something that should have been solved.
-sphere (February 21, 2024, 05:40 PM)
--- End quote ---
Solved by textbundle, except no-one supports it.
My experience is that directories move around and that when you move files around they may not have access to the original directory. Or they get renamed. Irritating rather than insurmountable. If you have a highly structured approach and are disciplined enough to stick within it, there won't be a problem.
Except in the end I think they all break. Like images on web pages. Or web pages themselves. I worked out that if I have all the assets I refer to in the file, then nothing ever gets lost.
sphere:
I worked out that if I have all the assets I refer to in the file, then nothing ever gets lost.
-Dormouse (February 21, 2024, 06:03 PM)
--- End quote ---
I have had issues using Window's documents folder or download folder. I try to go up to something like c:workdir because if i change computers then I can likely create that same folder.
So file, like the same folder/directory? Yeah that is hard for me, because I like to keep things in folder based on the contact/project/interest/category.
Solved by textbundle, except no-one supports it.
--- End quote ---
Was not familiar with textbundle. When checking out textbundle I looked at the supporting applications http://textbundle.org/ and I saw keepmark among many many Mac apps.
Keepmark is not open source. Single developer but it has some merits. Not as nice as tangent but there are a bunch of ways to dispay information and also it looks like it can pull in and process data from bookmarks, web pages etc.
https://keepmark.io/
https://www.youtube.com/channel/UCHBJWlv35p4jEqx4jrpIq6Q
looks like this is the developer https://github.com/tchudyk/
Dormouse:
many many Mac apps.
-sphere (February 21, 2024, 11:00 PM)
--- End quote ---
Yeah, quite a few Mac apps, but very few on Windows. Only one I've used is Zettlr, but I've not checked how that goes since I haven't used it in ages. Seems like there was a fair wind of Mac enthusiasm a decade ago, but then it died and has never picked up again.
I've read a Zettlr response to a request for textbundle to be settable as a default format:
thanks for the issue! Unfortunately, this is not possible. There are several reasons to that:
Both TextBundle and TextPack have been designed specifically for inter-application sharing of notes and cross-device sharing. They were never intended to be used as a default storage container.
TextBundles are basically just folders within folders, so using these by default would inflate the amount of folders within your application unnecessarily; additionally, information will be stored redundantly, thereby consuming more space than with the current setup. Additionally, it would take a longer startup time, as more information needs to be processed, and the current File System Abstraction Layer already has a lot to do parsing 10,000+ files due to the lack of caching methods.
The "good" way to work with Markdown files is to store notes in one directory and use a single img, assets, or images-directory to collect all images being used across the notes in that directory. This means: If you need an image a second time, you can simply link it to both files. Additionally, you'll have more control over what is stored where. Zettlr already gives you the option to automatically store pasted images in a predefined "collection" folder.
Markdown is intended to be used as a format that is both machine-readable and human-readable; to add additional layers of information makes it harder to write programs to address these issues.
TL;DR: The default way is and will be to have single Markdown files with the images in a different folder. TextBundle and TextPack, however, can be used to share specific information with other users, devices, or applications.
--- End quote ---
So that wouldn't work for me. I don't mind duplication if it guarantees security and I dislike the approved markdown default setup.
So file, like the same folder/directory?
-sphere (February 21, 2024, 11:00 PM)
--- End quote ---
No, as in a text file with images.
I'm okay managing folders of digital assets, but if I write something with images I want the image locked in with the text and not relying on a link.
docx okay, databases with good export okay, but .md isn't.
Luckily few of my notes or writing needs images, and my writing is in docx now anyway which is fine.
Dormouse:
The thread has moved on, with no further comment on Upnote?
-rjbull (February 09, 2024, 04:55 PM)
--- End quote ---
I decided to move it into test mode.
-Dormouse (February 17, 2024, 05:43 PM)
--- End quote ---
I also read a post on the Workflowy Slack from user Frank G - “I have been using Msft Word for Mac (Windows before) for as long as I can remember. I have also played with Craft, Ulysses , Scrivener, Speare, Author , Typora , all excellent alternatives. Today WF finally ate them all up too for my long form docs . The blank page with a few basic formatting choices plus WF’s speed (once opened) , simplicity, and flexibility, has made WF my go to choice for long form doc drafts . I will still need to export to Word (or Craft) for final formatting and PDF conversion, but 95% of my time I will be with WF .”
Essentially this refers to the Roamlike feature where bullets can transform into text blocks. Logsec has had this for some time too.
It struck me that it might be worth comparing the editors in (some of) the apps that I use, what makes them good and how Upnote compares.
One group is writing apps: ie those where there is a published output, especially those with longform output is in print formats. Important features include:
* a look and feel that sustains focussed concentration for long periods;
* a presentation that shows the reader experience, especially the white matter balance;
* the ability to manage long documents, whether by putting together many small documents (eg Scrivener, yWriter) or working in a single long document with multiple sections (eg Word);
* plus word count and other feedback about the output. Many editors, especially those markdown derived or inspired, are weak in 2) because of the lack of first line indent and blank lines between paragraphs. Ones that tick all these boxes include Word, Inspire Writer & Ulysses. iAWriter misses different features depending on the OS and has first line indent only in Review mode. Tangent is pretty good but lacks the ability to manage longform (and I haven’t personally tested the first line indent CSS); Upnote isn’t quite as nice, can’t show white matter balance, but has quite nice document feedback and can manage longform; Typora is nice, can (to an extent) manage longform, has first line indent but also blank line between paragraphs (although this ought to be modifiable with CSS), and word count is okay.
otoh, if we look at note and PKM features Word, Inspire, Ulysses, Scrivener, Typora all fail (Word sort of passes if you include OneNote to make a dyad). Tangent and Upnote pass. For my limited usage, Upnote works well on Android (and it includes a Quick Action widget for new notes). Images are an issue for both (Tangent link only, Upnote limits quality). Both are comfortable for writing notes. Tangent better for following a thread, Upnote better for web clips. Workflowy can do notes and PKM but remains primarily an outliner (which automatically gives longform); its big advantage/disadvantage for writing is that it is block based (which can help for working out structures, but hinders seeing the document as a whole) and that it has an alternate kanban view. All work with tasks: probably WF>Upnote>Tangent. Writers for print/publish using a notes app such as these for their writing will typically export into Word for the review and editing passes and sending to editors or publishers.
There are many notes/PKM apps. The editors are generally adequate (occasionally some delay in keyboard response) but tend to the generic. They compete on their other features, not the editors. Upnote’s editor is better than most of them. I find Tangent better for reflective writing in depth, but Upnote feels better for notey notes - even quite long ones: there are many little features (eg open a note in new window + keep on top button, for taking notes when browsing or using another app).
I'd just add that this
the Roamlike feature where bullets can transform into text blocks
-Dormouse (February 29, 2024, 06:37 PM)
--- End quote ---
is pretty well the perfect design for a zettelkasten as used by Luhmann. Every note has a parent. There is no distraction from bullets. Links are possible.
Purely relying on wikilinks is a non-reflective semi-automation that dissipates the pondering and sorting that made his method productive.
A mindmap view of the outline/cards might enhance it but his use of the system was essentially conceptual and verbal, not visual.
Dormouse:
I'm afraid this is a very long post. It seemed worth giving enough details for anyone else to be able to follow what I'm currently doing; it does assume some knowledge of outliners, Workflowy, zettelkasten and Luhmann.
I haven't done much with zettelkasten in the past - my usual need is for longer notes - but I have started one with Workflowy, and thought I'd explain it here.
* The first point I need to make is that, despite "Smart Notes", Luhmann's zettelkasten is not a system designed for students - at least not until they are following their own streams of thought; Luhmann was always a researcher who published articles and books. But, if the phrase honing stage is ignored, it is a perfectly applicable system for anyone who is following their own streams of thought whether that be RPG, making bookcases or tracking health. It ought to help develop expertise in anything.
* Second point is that it is a high effort system. Part of the productivity gain is from the repeated pondering when trying to locate the best place for a new zettel. The rest of the gain comes from the structure enabling brick to be placed on brick until a building emerges.
* The third is that Luhmann's practice (aka "the rules") does matter.
* Zettels have to be placed with time-consuming care in a sequence. Some may be new starters (for different subjects), but they need at least the same amount of thought before that decision.
* Zettels have to be brief and to the point: they're not a string of sequential thoughts - if they're worth it, sequential thoughts need to be sequential zettels. This is hard and requires discipline. If a thought isn't worth that effort, it should not be put into the zettel. I don't put my frequent long notes into the zettelkasten, but I do write a zettel fo each one, with the the reference. It's worth emphasising that a brief atomic note implies two conditions that have to be met, not just one.
* Workflowy supports this to a degree because it works best when each zettel is a bullet/paragraph.
* I find that, for the purpose of future linking, it is useful to append a time/date stamp to each zettel. My sequence is title (in bold) + time/date, then the note/thought. All in a single mass of text. I have a time/date stamp as my (most used) shortcut in a text expander so that it can be used in any program. I put references and #tags in the bullet note (#tags are used as a form of index entry)
* (Technically, all the content could be put into the note rather than the bullet - and I might do that if Workflowy offered the same viewing options for notes (show, hide, 1st line) as Dynalist - but it has no options, and always shows 1st line only.
* I also find it useful to have an external library/hoard (which is part of the usual zettelkasten system anyway), Items can be highlighted & referenced. Zettels need the essence and not the clutter.
* I'm aware that it's easy to add a zettel to the head of the list in an outliner. It should also be possible in a card index: idk if Luhmann ever did it; I doubt it since he was probably restricted by his numbering scheme, but I do sometimes.
* If I want to see, or play, with the zettelkasten in a mindmap, I do OPML export to Mindomo. And export back again if necessary. Most mindmap programs should be able to do this. (Many might be able to work as the outliner too. I simply prefer to use Workflowy).
* Workflowy doesn't have a wordcount. Which is fine for zettels, but is a pain for anything longer.
* There's also the possibility of using notes or comments to add detail or successive reflections. At some point the comments could be made into zettels of their own.
Luhmann's phases are essentially: read/think - write zettel - refine note language - place in zettel sequence. Every note placement is an opportunity for further thinking, refining language and adding links/references.
The system I am describing has neural pathways through folgezettel outlines/mindmaps, wikilinks and backlinks, tags as well as optional structure notes. Plus search and filters.
There are a number of reasons why systems like the daily notes and wikilinks of Roam and Obsidian don't create a functional zettelkasten.
* They encourage verbosity and lack of focus. (A frequent reflection from Roamans after leaving the cult.)
* (Everything should NOT go into the zettelkasten; the thinking about exclusion is important.)
* There's no targeted review.
* The long folgezettel debate on forum.zettelkasten.de largely compares the use of alphanumeric folgezettel with numeric date/time stamps as links. The latter's review system is based on Structure Notes (aka index notes, or Maps of Content in ObsidianSpeak); but folgezettel placing can take place multiple times a day and cover all areas potentially receiving new notes, whereas Structure Note review will be infrequent and partial (I've seen weekly recommended apparently with an acceptance that most users won't manage it that often). The debate made me think of angels pogoing on an needle: everyone is righteous and no-one changes position.
* No mental effort is used to find the precise place for a zettel/note. Everything relies on links, the automated backlinks and a graph.
* Structure is expected to be self-emergent (ie automated) rather than the consequence of reflection during placement. There's an assumption that note value is demonstrated by its link density.
In use, I find that this system highlights trains of thought rather than individual notes, and that going through it later does sometimes stimulate further thoughts. Not only are thoughts/notes not islands, but they have active (or inactive) trading networks in a way that graphs of wikilinks don't. I therefore use it for all topics where I am interested in the chain developing (and, maybe especially, branching).
If I write a long note, I decide whether I'm interested in a chain, and, if I am, I add a zettel to reference it. I write as many long notes as I did; the type of mental focus used when writing a long note is quite different to working on the zettelkasten. Sometimes I might write the zettel and then the referenced longer note later.
I happily mine old notes, highlights, webclips, articles, books and add a zettel to reference them when it seems right. I never directly add an old note to the zettelkasten.
Everything is written, described or summarised in my own words. When I don't like my words or phrasing, I have never been able to stop myself working on the language whenever I read something I've written.
Navigation
[0] Message Index
[#] Next page
[*] Previous page
Go to full version