What new workflow ideas do I try, discard, and keep?

I’m reorganizing my research workflows. It’s long overdue, and I know from experience that going slowly and carefully through seemingly dull process-oriented work will settle all the reading I’ve been doing into a better mental order. One of the disincentives is that I know I’ll try some things that fail and have to backtrack. I’m curious about what the churn will be, and also hoping to use this curiosity to take some of the sting out of having to backtrack. This project begins with a simple spreadsheet in which I’m recording the new processes as I try them. A sample row is here:

Idea Source Date Found Date Tried Date Ended Notes
iAnnotate for PDFs Research Methods for Historians by Robert Karl July 14, 2020 July 14, 2020 Want to read while walking, but hard to read PDF’s on my iphone…

I predict my try:keep ratio will be 2:1.

1 Like

Curious to know, do you zettlekasten?

:grinning: that is the second item on my “workflow ideas” list: :

Zettlekasten Method Academic Workflow by Beck Tench

It’s an interesting one because it suits some aspects of my current chaotic practice with books: I make marginal notes in pencil in books and try to efficiently record the relevance of the underlined section to what I’m currently working on. But they are stuck in the books, mostly, until I go back through the book piles and review. I can’t believe I made it this far with this method. Notes from Internet reading are mostly copied into a Google doc with the link as a reference, without comment, and then just read through, with the key bits then copied into a very rough draft. I normally make dozens of drafts, and of course the source material is digested in this process and only the key references remain. In my current workflow there are two key Zotero folders: One contains references to all the sources used in the drafts; the other is a final reference list. (Only the final reference list is carefully reviewed and corrected; the other is messy.)

As you can imagine, the process is glitchy, full of double work and with frequent loss of items. (“I know [some author] talks about this somewhere…”) I sort of like this mental searching process and find it creatively useful, but I’m also kind of sick of it; so I’m not trying to reach ultimate efficiency but just recreate my workflow using up-to-date tools and reduce some of the annoyance.

Today I’m trying to figure out where to collect the notes from books and other reading. Trying SimpleNote… which fails after only 15 minutes of testing. I can’t see any advantages over my random collections of Apple notes.

But I do come out of my overnight thinking about what notes app to use with a new idea. One of my favorite things to do with book and web passages is to take a photo or screen grab. Sometimes I post the picture to my @agaricus Twitter account (“random and untidy”) so I know where to find it. (I know, absurd.) Therefore what I want from my notes solution is: easy sharing of images from screen grabs and iphone photos, with Zettlekasten-type summary/reflection text.

I found this video by Shu Omi very useful for a brief Zettlekasten description; I have not yet investigated the Roam software he mentions.

Had a chance to spend some time today looking at other workflows and help documents. My new draft workflow has the following elements and processes.


  1. Reading Notes

—Image+Note (immediate comment)

  1. Reference Notes

—Reference Note Title (Article or Book Title)

—Minimal bibliographical reference (Title/Author/Date/URL)

—Biographical/bibliographical detail if wanted.

  1. Zotero entry.

  2. Permanent Notes

—Image (optional)

—Written Note



Tools: Apple Notes, Zotero, and Roam

Collecting Reading Notes

  1. Reading Notes are sent to Apple Notes from whatever reading environment I’m in.
  2. Reading notes are copied into a section on the daily notes page in Roam.
  3. Reading notes are grouped by source using “block reference” in Roam.

—Create or Open [[Source: Title]]

—Add Reference Link [[Reference: Title]] to Source if needed

—Add reading notes using “block reference” (ALT+drag)

Creating Permanent Notes

  1. Highlight the key text on the reading notes, and add double square brackets around the key text to start a permanent note.

  2. Add Source links to the top of the permanent note. Source: [[author]] [[title]]

  3. Add Keywords: #keyword #keyword phrase . (All permanent notes should have #permanent notes as a keyword)

  4. Add Relevant Notes by opening [[Permenant Notes]], reviewing notes either by skimming or using keywords, and adding links to the relevant notes.

  5. Add reading note using “block reference” (ALT+drag)

  6. Edit/rewrite reading notes into full sentence permanent notes that can be incorporated into a draft.

1 Like

Lmao it sounds like you’re a lot further along than I am I’ve been using Tiago Forte’s PARA method since the beginning of this year and really enjoying that. Though upon learning about zettlekasten the other day I’m slowly gathering info and impressions about Roam and Obsidian to see if it’s worth diving in to. Check out Obsidian if you haven’t already, I’m a really visual person and their node / network visualisation looks incredible to me. Just not sure how many different softwares I want to introduce into my workflow, so far Notion has been a godsend for organisation but the level of connectedness which roam and Obsidian bring to the note network looks soooo appealing, especially for that creative part of the process where you can let your thoughts… roam through the archives and just find connections.

Hacker News often posts and comments on various methods of more efficient note-taking and research, including Zettlekasten, Roam, Evernote, etc. (A recent one is here). I’m a sucker for these various discussions, though over time I’ve concluded that the effectiveness of a given system depends on the dedication you apply to learning it. It’s easy to waste valuable time learning such-and-such method when the time would have been better spent just working on your project.

1 Like

Very much agree with this, which is how I got 3 decades into a research-y career with such a messy process. I prefer diving in and keeping going. But I’ve reached a point where the amount of time I spend searching for what I need is annoying enough to give at least some limited attention to trying to find a better way. One of my mental tricks for overcoming resistance is to turn the search itself into a research project – even if I go back to my previous way, I’ll have some semi-well formulated lessons.

1 Like

OK, so far so good. Although I learned that Roam won’t take my photos of book pages (too large), I don’t really need them in Roam. I open them in Apple Notes, then use them to make my permanent card in roam, with tags to the source:

I’m finding Roam intriguing but I’m not really sold yet. A few things are bothering me:

*The learning curve is rather steep, and I’m not sure I want to immerse myself in conforming to a complex system.
*It is a proprietary system and I have some concern about lock-in. So many things can go wrong with these systems: the most risky of all is that they aren’t sustainable and disappear, along with the entire structure of your data.

Going to keep using it for a few days but starting to have my doubts.

Honestly this is something I’m terrified of and mentally preparing for. If Notion blew up overnight I’d be heart broken.

I know that Obsidian is totally local (no cloud storage), I’d have already gotten into it if it weren’t for the zettlekasten learning curve.

OK, I’ve made quite a bit of progress through using and learning Roam and I’m now more optimistic it can work, though far from fully convinced. One breakthrough was key for me: Many of the tutorials and guides focus on implementing specific workflows. This is useful for learning but not easy, because you have to mentally adjust for the differences between the teacher’s use case and your own. I haven’t yet found a tutorial that explains how to approach the tool from a more conceptual level, giving you a clear mental model to carry around and refer to as you customize.

However, I did have an understanding breakthrough that helped me a lot. In Roam, the fundamental unit is called a block. Behind the scenes, Roam creates a unique identifier for each block. (You can see these, but they are normally hidden.) The unique identifier is a string of alphanumeric characters. It’s good to remember than every block has its own unique identifier, because it helps you get in the habit of not caring too much about the apparent hierarchy of the organization as it is displayed to you in the interface. Once I began thinking of every block as a free floating element (my mental picture was a loose index card), it became easier for me to experiment and learn how to best create connections to support my own workflow.

For instance, Roam tutorials advise you to start working in the “Daily Notes” section. You just start entering your info on today’s page, and then add tags to each block so you can find it again. This is a bit unnatural if you are used to hierarchical file systems, and so you start asking yourself: OK, now that it’s in my Daily Notes, do I tag it with its source, idea, project and other tags, so I can find it again? But what if I’m not sure what project I’ll use it in? Or what if it’s not a fully formed idea? What if I have more than one tag for the same idea, won’t this get confusing. And Roam has a bunch of different ways to build links, all slightly different, so it’s easy to get lost watching tutorials and then getting confused by others’ workflows.

After I started to think of my cards (in Roam, “blocks”) as truly independent elements, all this became much easier to handle. I now give a little more time to thinking about what goes on the card, and, especially, what the first words of the card are. These first words are my “card title.” I always mark them using square brackets, like this [[card title]]. Once this is done, I have a good unit of thinking I can use in the future. After this, I don’t worry too much if I have all the right tags or links or embeds, or whatever else Roam offers. I’m sure these are useful, and I’m slowly integrating them into my approach, but the main purpose is being served well so I’m motivated to continue.

1 Like