I'm also having another go at doing a serious journal/diary.
I've been inspired by Virginia Woolf's diaries which I hadn't read until I was recently gifted a copy. (Ebook versions are available for free on the internet, like all of her work.) I've always been impressed by the famous diaries (Pepys etc) , but they'd never seemed relevant to my own potential use. These diaries fit much better. They seem to cover everything. She took them seriously, trying to set aside a small amount of time each day, but there are still many days with no entries. There's description, observation, activities, introspection, planning and recording her reading and her writing programme. Recording and trying to analyse her feelings and emotions and health (frequently poor). Practicing phrasing and style (and reporting her feeling, after some years of use, that it had helped her writing to flow more easily).

In the past, I've often been caught out by the physical methodology getting in my way. And limited time. That led me to focus on what seemed important (with the habit stopping when it no longer was), and books (printed diaries/plain notebooks) which then became inconvenient. And I've tried so many digital methods without ever getting going seriously with any of them. At least with the pen and papers ones, I still find the ones made years ago interesting today. But in recent years I've never been able to do more than a few days at a time in a paper diary, so digital it has to be.

I tried a plain document approach using Obsidian, but that became messy and not very useful. So as my system is opening again, I've returned to Diarium, which I first looked at in March last year before encountering Obsidian. AND I'm having another go with dictation on the phone. The availability suits me well and I'm accustomed to dictating, even if it is never as fast as typing, when it feels as if it ought to be faster. Worked okay so far, but very early days, so mostly likely another false start; maybe not because the Woolf model could easily work for me. And I'm not limited to dictation or the phone. I expect to export regularly into .txt files.

But the dictation works  remarkably well. 'Marques de Riscal, Rioja Reserva 2016' without a hitch. I was well impressed.

Wikilinks in Obsidian are a wonderful tool for writing and research. I'm still investigating the best way to use Obsidian (& et al) for my writing, in particular looking for the best method for organising and manipulating longer pieces of writing.

One option is very long documents and using headings. Headings produce outlines and can be linked directly.
Another option is multiple documents with a MOC/Index. This makes manipulating and re-sequencing very easy.
Another is simply using files and nested folders - the way most writer's programs do it.
However, all  are very linear. And I inevitably find linear constricting.

I have always used spreadsheets as part of my planning process, and am aware that many writers have constructed complex systems using multiple spreadsheets.
Tables can help as a cutdown way of doing the same, but their limited functions restrict what can be done. And markdown tables are a PITA (though Typora's are more usable than most).
Ideally I wanted to continue to use spreadsheets. Obsidian allows links to the files which can then be opened in the default spreadsheet program, which is manageable.
But I have worked out a much, much better solution.

1. I use a spreadsheet. Scenes, people, places, concepts (anything I choose) get put in wikilinks if I think that I might want to link to them. I can switch and move things around however I want.
2. I copy the spreadsheet and paste it in Typora (doesn't work if I try pasting in Obsidian or WriteMonkey, but haven't tried anything else).
3. I copy the Typora table and paste it into Obsidian.
4. Obsidian will automatically link to any notes that already exist and will offer to create notes for the wikilinked titles that don't already exist.

This gives me a supercharged index that isn't simply sequential. If I hover over the link, I can see the content. I can use the spreadsheet itself (I'll stick a link in as the title of the table) for analysis and development.

The one fly in the ointment is that this seems to be unidirectional - you can't take the markdown table and paste it into a spreadsheet. I assume It might be possible with a few conversions, but that doesn't feel like a productive workflow in normal circumstances. It's not a major issue - it means always making changes in the spreadsheet and copying back into Obsidian, but that's all.

Probably of no interest to most people, but I find the possibilities exciting.

that's a pretty manual process.  It would drive me nuts lolll...

Indexing, I basically use the classic zk of idea of creating a note that links to all the sub-topics.  So i create an index note.  Multiple index notes can link to the same sub-note, that's fine.

I think the  difference here is that this part of my process has nothing to do with note-taking or zettelkasten. It's about planning, writing and organising an MSS. If there will be enough for more than one book, moving parts between them so that each one is well structured. Making it easy to see gaps that need filling. Ditto for generating multiple articles from one research programme. And that's the same for a series of articles on aspects of the same issue. Irrelevant if you're hand-to-mouth but essential when you're in a position to plan the series.

If you're a pantser writing fiction it has no value at all. Though, when I think on it, if you write lots of bits it might help you stitch them into coherence.

Has to be manual because all the decisions require thought.

And at the end, you do have an index. If there’s a book series, and each column is a book, with scenes on the rows, all that's needed is to copy a column, put a ! In front of the wikilinks to insert the transclusions, and then export/print the whole MSS.
You'd have other columns in the spreadsheet of course. Word counts, targets, appearance of characters and locations, whatever it is that's helpful for planning or editing/reviewing. But you probably only want that in the spreadsheet.
Would work for a PhD too. Be overkill for an UG essay, but could help with a thesis.
Or multi-stage business plans. Or years of committee meetings. Any whole that splits into multiple sections, with highly detailed components, where each section has a similar structure.

I wouldn't use it at all for zettelkasten type notes (or the value of using it there hasn't struck me yet).

I just found something new- but I'm not sure about its practical use for me - https://orgmode.org/

It's primarily for organization to create a living document. It's more of a note-taking language for technical purposes, with formatting and such a second-class citizen. It reminds me of Jupyter notebooks. As I use my Zettel to format and track my blog posts, I think it wouldn't work for me. It has features that make me tempted to try it out, though.

There's an extension for Visual Studio: https://marketplace.visualstudio.com/items?itemName=tootone.org-mode- source at https://github.com/vscode-org-mode/vscode-org-mode, but it doesn't seem to have the primary power of Org-mode (blocks, code execution, macros, etc).

I'm really not up on going back to EMACS, so perhaps this is a non-starter.

Org-mode

I looked at this a while ago. Concluded that it was better than markdown but lack of ubiquity made it too much of a lock-in, as with AsciiDoc. As time has gone on, I'm glad I learned markdown and have an awareness of these alternative options, but I've moved away from using them all. Now it's mostly text with occasional Obsidian commands.

I have developed a very barebones approach to Obsidian. No community themes or plugins (every now and then users report data loss from one of them, and I just don't need the aggravation). And nothing confidential (it's moved from json to a database so no longer immediately readable, but still saves in User folders); I could manage it securely but don't want to have to design a system around a still moving target.

As it is I have a highly functional system requiring little effort.

My journal has continued successfully, still informed by Virginia Woolf's practice. Once she'd got into the swing of it, she wrote her journal in the afternoon/evening with her serious writing in the morning. Journal entries varied: sometimes she used them to help develop new ideas and issues; sometimes she practiced styles; sometimes she recorded events, sometimes, observations, sometimes feelings; sometimes she analysed her beliefs and aims about writing. Gaps, of course, because of her illnesses and engagements that took all her time.

I'm not partitioning by time of day, but I am partitioning by entry and storage system. Continuing to use Diarium, even though Obsidian will become a practical option with the mobile versions. I can export the relevant entries from Diarium (as text) into Obsidian (rename as md). I| doubt the Obsidian option will ever be as slick, and I can already recognise the value of the different mindset when I'm writing the diary. If I didn't need mobile, I would probably just use The Journal but I do find Diarium very smooth. Effectively, it means that the journal is one path into my notes.

FWIW I'm also trialling Instapaper (potentially also comapring it with Pocket) and Readwise, again as part of the entry system. Also Readly. I'd never seen the value of Readwise which struck me as expensive, given that it is easy to find alternative ways of doing the same thing. But convenience has a value and one of the advantages it might bring is a greater use of highlighting and notes on everything rather than just where it will obviously be useful.

I'm aware that there have been one or two Obsidian updates that have resulted in (a tiny amount of) data loss (quickly addressed). And somewhat greater issues arising from plugins. Some reports of data loss seem most likely to be due to conflict with a user's sync system. For the moment, I don't use community plugins at all and I am exercising caution over what I use Obsidian for and making sure I have sufficient backups. (Ditto for the Diarium database.) My plan is to do a major review once the pace of development has slowed; I expect then to move back to heavier use, but with a clear idea of how to manage security and what limitations I should impose on use. So far it has moved from all data being in the vault folders, to adding jsons in User profiles, and then adding a database in the User profile; the databases aren't essential - they will be recreated from scratch if deleted, but there's no way I've seen to turn off the persistent storage automatically. It's moving too fast to be worth attempting such a review yet.

I've started to combine a new editor with my VS Code workflow - Deepdwn (https://billiam.itch.io/deepdwn).

Looks quite nice, though I'm not sure I have a real use for another editor.
Not quite sure where I stand on YAML. I originally stuck a YAML section into Obsidian notes, for potential future use, but then deleted them all when they said that the YAML was being reserved for plugin use - would be beyond irritating to develop a use and then have it disrupted by plugins. That seems to have reversed and it's now used by some people, though I'm still not sure why I would need it. Easy enough to add should I want to.

I use Typora regularly because pasting excel cells produces a markdown table that I can copy and paste elsewhere. I don't know which other editors will do that, but don't have an incentive to look.
FWIW, if I paste the cells into Diarium,  I get a formatted table in an image file and the unformatted text in the body of the entry. I always ue Excel because I don't get the same response from Sheets - even when I paste the cells into Excel and then abstract them again. It all seems very odd to me, but I don't really need to understand why they work in this fashion.

Also briefly tested the txt-as-md plugin on a test vault.
My interest was Obsidian commands rather than markdown,  but successfully inserted a transclusion into the txt file.
Will save a smidgen of time when exporting from Diarium and a heapfull if want simultaneous access from a program that won't read md.

Also briefly tested the txt-as-md plugin on a test vault.

That's interesting. I see in https://forum.obsidian.md/t/handle-edit-other-plain-text-file-formats-rmarkdown-tex-txt-code-etc/1656/50 that there is discussion about two possible features.
1. Have Obsidian read .txt (and other plaintext) files as if markdown (apply markdown styling, and so on)
2. Have Obsidian read .txt as plain plaintext (no markdown styling)

The plugin does 1 AFAICT. But if this gets wings perhaps both 1 and 2 will get implemented, and maybe also a further feature
3. Have Obsidian read files with extension .NNN as some other, perhaps user customized plaintext format that differs from markdown. For example, AsciiDoc.
That is, users would in a config file tell Obsidian to treat filetypes {.md, .txt} as MarkDown, {.asciidoc, .adoc, .asc} as AsciiDoc, and so on.

Obsidian would expand to be a more general system-of-plaintext-notes editor/viewer. Compare to how VS Code is an editor/viewer for code in different programming languages. That ties in with our previous discussions here for/against the markdown format(s) and alternatives to it.

1. Have Obsidian read .txt (and other plaintext) files as if markdown (apply markdown styling, and so on)
2. Have Obsidian read .txt as plain plaintext (no markdown styling)

The plugin does 1 AFAICT. But if this gets wings perhaps both 1 and 2 will get implemented, and maybe also a further feature
3. Have Obsidian read files with extension .NNN as some other, perhaps user customized plaintext format that differs from markdown. For example, AsciiDoc.

The plugin does 1, but only partially. It does 2 in the absence of any markdown in the file; up to a point. I'm not convinced that the holes in these will ever fully filled, but I see no chance of 3, although there must be a possibility that a plugin could be written to do this to some extent, maybe.

The plugin allows .txt to follow markdown styling. It allows links and transclusions, in .txt files. It doesn't allow .txt to be read natively or to be transcluded itself, and links to .txt files have to be typed out in full. Neither are tags in .txt recognised and nor do .txt files appear in graphs. Having hit these limits, I am now wondering how much use I will be able to make of it. For some things it will work fine, but not being able to be transcluded means that the technique of embedding chapters into a file to produce a complete MSS won't work. Of course, there's no difficulty in renaming them all at that point but working out which format to work in at each stage doesn't seem straightforward now; quite a collection of swings and roundabouts.

Quote from: wraith808 on March 23, 2021, 01:08 PM

I think a lot of this relates to the direction the developers want to take the application. I haven't seen any indication that they want to make it a general purpose application.

I'm not sure. I agree it's not their own intent, but they do follow users and are very open to almost anything being done with plugins and frequently consider extending the API to make some plugins possible. But how much can reasonably be done via plugins I have no idea.

Oh, I'd agree if it can be done via plugin they'd make it possible to be done that way. I was just talking about core intent. The way I'm using VS Code right now isn't MS's intent. However, it's made possible by extensions. And they've shown that they are open to including particularly useful functionality from extensions in the core app.

Actually, pretty much sold on signing up. Unfortunately desktop + mobile requires a subscription, but it always takes me at least a year to try things out properly (Obsidian was an aberration), and the price seems pretty standard for the paid mindmap apps. Ticks all my required boxes. Some clashes in markdown syntax, but that's a straightforward conversion. Pessimism has turned to optimism.

I have asked myself what it doesn't do better than all the programs on the list for the short review I'm doing. The answers are:

• cost
• direct link to the editor used for writing
• enforced simplicity (ie risk of being pulled into unproductive ornamentation)

This feels like a remarkably short, though not insignificant, list.

In truth, Mindomo will be the best of these options for brainstorming, organising and reorganising lengthy writing if most development work is done in it from the beginning.

One issue for me that I've noticed is that the text editor for attached notes doesn't seem to have a dark mode option, which risks a glaring white panel. But that should be avoidable. Actually on the web version, it comes up in dark mode on Vivaldi anyway, so no problem at all.