Posts

An analogy for app switching

Last updated Dec 19, 2024 | Originally published Aug 16, 2024

I buy a house.

The new house is amazing. It has so many rooms, and so much room. It’s got a nice fridge with a cool little ice maker. The island countertop is huge. The garage has one of those peg board organizers and all of my tools are so easy to find. It even has a little sewing room with a automatic sewing machine so that I don’t need anyone else to hem my pants.

I move in. I live there. I live there some more.

I see some other houses in the city for sale. Those houses look nicer than mine.

Our house has the same number of rooms, but there’s no room for anything. I can never find what I want in the fridge and the ice maker broke and I haven’t had time to order the part that I need to fix it. The island countertop is always full of toys and birthday invitations. I can barely open the garage door because of all the half-finished projects in the way, and besides, nothing is where it’s supposed to be on the peg board. I can’t hem my own pants, either: we put all of our saved search queries and tags in the sewing room. Besides, I never learned how to actually use the sewing machine.

I buy one of those other houses in the city.

The new house is amazing. It has so many rooms, and so much room. It’s got a nice fridge with a cool little ice maker. The island countertop is huge. The garage has one of those peg board organizers and all of my tools are so easy to find. It even has a little sewing room with an automatic sewing machine so that I don’t need anyone else to hem my pants.

I move in.

Tools like note-taking apps and task management tools feature a special kind of cruft: the detritus of old parts and ideas that builds up as someone works on projects. Cruft leaves you with the feeling that it needs to be intentionally and deliberately sifted through and recycled. What can you remove? You can’t just throw it all out, as it might contain important parts that are hard to replace! But you never have the time or energy to deal with it, so it accumulates, and sort-of gets in the way although you can kind of step around it if you know where to put your feet.

In software development, cruft can lead to technical debt and dependency hell. In knowledge/task management, cruft leads to knowledge management debt (“where should I put this? In the projects folder in my role folder? Did I tag it with the new-project and topic tags? Maybe I’ll just put it in one of my inboxes for now.”) and to-do dependency hell (“before I start on writing about this new idea I really should finish this reading to-do I had in this related project because it might be relevant and I need to know that before I start”).

The hidden problem here is that these tools typically feature a special, easy solution to cruft: just start using a new tool.

When you start using any kind of new knowledge or task management tool, there’s a sense of shine and wonder to everything. Yet the most important reason for this is not because the new tool is better than the old one: it’s simply because you haven’t moved in yet.

Crucially, this effect is more than “tabula rasa,” the blank slate. Yes, as you start using the shiny new tool, you’ll end up adding the projects you’ve been anxiously avoiding, the repeating reminder you keep pushing off, and the notes on that blog post you’re never actually going to write. Yet an even more sinister problem than crufty stuff is crufty softerware. Your knowledge and task management system contains to-dos and notes, yes, but it is also made up of the habits and behaviours and idiosyncratic use cases that you develop as you write those to-dos and notes. In those first moments of adopting a new tool, you forget about all of that cruft. However, even though it may take months or even years of use to build up new piles of crufty stuff and softerware, it will eventually happen — and this cycle is the engine of the industry.

Is it always a bad thing to switch tools? Of course not. Sometimes the new software is absolutely better than the old, and sometimes the new softerware you develop empowers you to do things more effectively or efficiently than before. But it is important to be aware of both the cost of switching and the reasons why the new thing feels better than the old. So, when you are tempted to switch, try to look for your own system design smells: are you switching because there’s too much cruft built up in the system around the new tool? If so, take a moment to reflect: setting up a new tool will probably cost you as much time and energy as cleaning up the cruft of the old tool. Which is the better expenditure of time? Only you can decide.

Originally published in the MacPowerUsers forum

Apps

Set up a hyperkey and globe key on iPadOS with a QMK keyboard

Last updated Jul 30, 2024 | Originally published Jul 30, 2024

I really wanted the experience this person is having, but with an iPad:

So I recently picked up a Nuphy Air75 v2 keyboard (Wisteria switches) with the Nufolio case/stand.

For a quick review of the experience: it’s pretty great! The keyboard itself sounds and feels excellent — much better than the iPad Pro’s Magic Keyboard, which I used begrudgingly with my previous iPad Pro. It has all kinds of great features that I’d consider basic at this point:

multi-device pairing (so I switch between using it with my phone and my iPad, a feature that has already been helpful when I needed to type a message on an app I don’t have installed on the iPad)
built-in kickstand feet at adjustable levels
F-row keys
customizable backlighting

But most importantly, the iPad doesn’t need to be attached to it for it to work. Thus I can sketch and layout diagrams with typing, touch, and stylus, all at once, while the iPad is in my hands or lying flat. It’s quite nice.

However, the keyboard’s layout is not what I’m used to. On the Mac, I’ve conventionally customized my keyboard with BetterTouchTool and other third-party services. Unfortunately, the iPad is a “console computer” and iPadOS isn’t a real operating system. So, iPadOS doesn’t allow developers to develop cross-system customization tools, and iPadOS only offers a very limited set of keyboard customization options. You can use Settings → General → Keyboard to change a few modifier keys around, but that’s about it unless you want to turn on full “accessibility mode”-customization. (Accessibility → Keyboard and Typing → Hardware Keyboard → Full Keyboard Access, IIRC.) This gives the keyboard all kinds of power over what you can select and act on in the OS, but it’s a bit intrusive and not necessary for my use-cases. Regardless, even the “Full Keyboard Access” options fail to let you do the really fancy keyboard customizations I’ve come to rely upon, such as setting up capslock as a hyperkey: changing the capslock key to function as esc when tapped and shift+alt+ctrl+cmd when held.

Fortunately, the Air75, like most modern mechanical keyboards, is a QMK keyboard. That means that the keyboard itself is customizable on a firmware level: you can use tools like the VIA keyboard customization console to fundamentally change what the keys on the keyboard do. So, that’s what I did! Now, even on iPadOS, tapping capslock enters esc, and holding capslock gives me the shift+ctrl+alt+cmd modifiers all at once. This functionality is now just how my particular keyboard works, so it will work the same way on any device I connect the keyboard to, no operating system-level customizations or third-party services (such as BetterTouchTool or Hyperkey) required.

Here’s a quick guide for you to do it, too (provided without warranties, guarantees, or support):

Purchase a QMK VIA-compatible keyboard
Follow the manufacturer’s instructions to set up the keyboard to work with VIA, if necessary (Nuphy’s instructions are here)
Open VIA and connect your keyboard (again following the manufacturer’s/VIA’s/QMK’s instructions to make sure it’s connected and authorized properly)
In VIA’s configuration tab, select the capslock key
In the customization options available at the bottom of the screen, choose the “Special” tab, then select ANY
Enter MT(MOD_HYPR, KC_ESC).¹ This translates to:
- MT: Modifier-tap, as in “act as the modifiers I specify when held, enter the keycode I specify when tapped”
- MOD_HYPR: shift+control+alt+GUI, where “GUI” means cmd on macOS and win on Windows and… something on Linux, presumably (I avoid the warren of rabbitholes that is Linux for reasons that should be obvious if you’re reading this post)
- KC_ESC: the esc key.

Now, what about the “Globe” key? It has become increasingly useful on iPadOS in recent years, but for complicated reasons, it is currently not possible to add a globe key to a keyboard via VIA/QMK. Fortunately, iPadOS allows you to modify your modifier keys (as mentioned before: General → Keyboard → Hardware Keyboard → Modifier Keys). We can remap capslock to the globe key! Never mind that you just got rid of the capslock key: just use VIA to place capslock elsewhere, e.g., on your now-redundant escape key. Then, change your capslock to globe on your iPad, and you now have a globe key on your keyboard, too.

Here’s my final layout:

I think you could actually use HYPER_T(KC_ESC) here, which is basically a shortcut to what I’ve done, but I got MT(MOD_HYPR, KC_ESC) working so I decided not to try to experiment further. ↩︎

Note-taking apps (and practices) do make us smarter

Last updated Aug 26, 2023 | Originally published Aug 26, 2023

Platformer’s Casey Newton published an interesting piece on note-taking apps yesterday: Why note-taking apps won’t make us smarter.

It’s a fairly rich piece in that it clearly lays out a number of challenges experienced by people participating in knowledge management and innovation. These challenges — such as the fundamental idea that these tools should be designed to help us think, not just to collect things — are important.

However, I have to disagree with the headline.

Does a hammer¹ make someone stronger?

No.

But you can do a lot more with a hammer than you can do with your fists.

Same goes for note-taking apps.²

These are simply tools³ that offer us different ways to work with the material of our thoughts — our notes! — to shape them into whatever we want to.

But, as many have described already, it’s how we use the tool that matters.

What won’t make us smarter is to do what Casey describes in the article:

I waited for the insights to come.

And waited. And waited.

Marginalia

This was originally published in a Mac Power Users forum discussion.

…There will come a day when my hammer metaphor is bent so far that it breaks, but today is not that day. ↩︎
And same goes for our note-taking practices, which is often overlooked in articles like the OP but, as many have already discussed here, is the thing that actually matters. Latour and co. had it right. It’s not the person, and it’s not the tool, it’s person + tool. Or: we shape our tools, and they shape us. ↩︎
Or thinking environments, even. ↩︎

AI is the new plastic

Last updated Jul 18, 2023 | Originally published Jul 18, 2023

Data was the new oil, and now AI is the new plastic.

From user yabones on HN, discussing media companies’ blatant strategies for using “AI”-based text generators to spam content for Google Search Engine Optimization.

Permissionless integration

Last updated Jul 3, 2023 | Originally published Jul 3, 2023

In a recent blog post, kepano writes about the power of files for enabling users to have access to and use of their data in the long-term:

File over app is a philosophy: if you want to create digital artifacts that last, they must be files you can control, in formats that are easy to retrieve and read. Use tools that give you this freedom.

File over app is an appeal to tool makers: accept that all software is ephemeral, and give people ownership over their data.

This reminded me of a related insight I had about files over apps a few years ago: the friction-free power of permissionless integration.

User data should be like a piece of wood on a workbench: you can pick up hammers, drills, screwdrivers, nails, paint, saws, and all kinds of other tools and materials and make that wood into what you want. No special access or permission is required to cut or sand or shape that block of wood. You just pick the right tool for the job and do what you want. It’s your wood, your workbench, and your tools.

Digital tools should be built so that users can work with their data in the same fashion. Apps should be able to interact with one another to help users shape and learn from their data — their notes, models, drawings, spreadsheets, or whatever — without needing special interfaces to do so.

This is possible with files. Using files (especially files with open, standard file formats) removes the need to develop special ways of working with user data.

On the other hand, app-specific data structures create friction and lock-in. To read and change your data in one of these apps, you need to deal with exporting and importing, or only use tools that have been custom-designed to work nicely together (i.e., via an API). Use one of these tools to create and save your data and suddenly that data can only be shaped by a limited selection of other tools.

I’m sorry, your subscription for this pen has expired. Please use another Ink Pro-compatible pen or resubscribe for just $3/month per pen (billed annually).

This creates some ferocious friction. Imagine picking up a piece of paper with your latest grocery list on it. You go to add “Bananas” to that list … only you don’t have the pen you first wrote the list with, and none of your other pens will work with that sheet of paper. Then, when you find the original pen, your monthly subscription to it is expired, and so anything you wrote with that pen is now read-only.

That’s a scary thing. Remember, we shape our tools, and our tools shape us.

Being shaped by tools you haven’t shaped is not something anyone should want.

Towards a theory of leverage for strategic systemic change

Last updated Feb 24, 2023 | Originally published Feb 24, 2023

My article “Leverage for Systemic Change” was recently published in the inaugural edition of Contexts, from the Systemic Design Association.

The article ultimately proposes a few key directions for a research agenda on leverage in systemic design (see the table below).

Table 1. A research agenda for leverage theory in systemic design

Research area Research questions Existing research Possible studies Possible contributions

Dimensions of leverage - Is Meadows’s (1997) typology complete?
- What other features of the “physics” of systemic change might matter?

- System characteristics (Abson et al., 2017)
- Conditions for systemic change Kania, Kramer, & Senge, 2018)
- Other types of phenomena (e.g., bottlenecks, signals; Murphy & Jones, 2020)
- Relative leverage: chaining leverage points (Fischer & Riechers, 2019)
- Relative leverage: the context of the changemaker (Klein & Wolf, 1998)
- Recursive leverage

- A systematic literature review (Okoli & Schabram, 2010) of leverage points, especially using forward citations (Haddaway et al., 2022) from (Meadows, 1997)

- Understanding the nature of leverage and other mechanisms of change potential in systemic change

Methods for leverage - What methodologies are best to identify and select leverage points?
- What kinds of evidence will help validate leverage?
- How might systemic designers design theories of change (Gregor & Jones, 2007) for leverage theories?
- How might systemic designers limit indeterminism (Lukyanenko & Parsons, 2020) in leverage theories?
- Meadows’s (1997) typology’s order of effectiveness
- Leverage analysis [Murphy & Jones, 2020]
- Assessing potential for change (Birney, 2021)
- Surveying practitioners in systemic design on how they identify, assess, and address leverage points to identify common habits and best practices - How to identify phenomena useful for leverage
- How to evaluate and compare possible leverage points in the analysis phase
- How to evaluate the effectiveness of chosen leverage points with evidence gathered from implementations

Strategy with leverage - How is leverage best used in developing strategic plans for systemic change?
- How are leverage-based strategies best presented and communicated?
- How are leverage-based strategies best evaluated and measured?

- Systemic strategy (Murphy & Jones, 2021)
- The epistemic benefits of a leverage points perspective (Fischer & Riechers, 2019)
- Identifying conditions for systemic change (Kania et al., 2018)
- Relative leverage: chaining leverage points (Fischer & Riechers, 2019)
- Relative leverage: the context of the changemaker (Klein & Wolf, 1998)

- “Systemic change labs” tracing and comparing the impact of interventions using different kinds of leverage
- How to use leverage to develop better strategies for systemic change
- How to account for relative context in the design of high-leverage strategies

Execution on leverage - What are the best ways to target different kinds of leverage for systemic change? (E.g., how might we help actors in a system track all of the relevant paradigms?) - Fruitful friction as a tactic for transcending paradigms (Buckenmayer et al., 2021)
- Systemic change happens via multiple dimensions of change (Mulder et al., 2022)
- Design Journeys offers several chapters on taking action after identifying leverage points (Jones & Ael, 2022)
- “Systemic change labs” tracing and comparing the impact of interventions using different kinds of leverage - How to design innovations for each type of leverage

Research area	Research questions	Existing research	Possible studies	Possible contributions
Dimensions of leverage	- Is Meadows’s (1997) typology complete? - What other features of the “physics” of systemic change might matter?	- System characteristics (Abson et al., 2017) - Conditions for systemic change Kania, Kramer, & Senge, 2018) - Other types of phenomena (e.g., bottlenecks, signals; Murphy & Jones, 2020) - Relative leverage: chaining leverage points (Fischer & Riechers, 2019) - Relative leverage: the context of the changemaker (Klein & Wolf, 1998) - Recursive leverage	- A systematic literature review (Okoli & Schabram, 2010) of leverage points, especially using forward citations (Haddaway et al., 2022) from (Meadows, 1997)	- Understanding the nature of leverage and other mechanisms of change potential in systemic change
Methods for leverage	- What methodologies are best to identify and select leverage points? - What kinds of evidence will help validate leverage? - How might systemic designers design theories of change (Gregor & Jones, 2007) for leverage theories? - How might systemic designers limit indeterminism (Lukyanenko & Parsons, 2020) in leverage theories?	- Meadows’s (1997) typology’s order of effectiveness - Leverage analysis [Murphy & Jones, 2020] - Assessing potential for change (Birney, 2021)	- Surveying practitioners in systemic design on how they identify, assess, and address leverage points to identify common habits and best practices	- How to identify phenomena useful for leverage - How to evaluate and compare possible leverage points in the analysis phase - How to evaluate the effectiveness of chosen leverage points with evidence gathered from implementations
Strategy with leverage	- How is leverage best used in developing strategic plans for systemic change? - How are leverage-based strategies best presented and communicated? - How are leverage-based strategies best evaluated and measured?	- Systemic strategy (Murphy & Jones, 2021) - The epistemic benefits of a leverage points perspective (Fischer & Riechers, 2019) - Identifying conditions for systemic change (Kania et al., 2018) - Relative leverage: chaining leverage points (Fischer & Riechers, 2019) - Relative leverage: the context of the changemaker (Klein & Wolf, 1998)	- “Systemic change labs” tracing and comparing the impact of interventions using different kinds of leverage	- How to use leverage to develop better strategies for systemic change - How to account for relative context in the design of high-leverage strategies
Execution on leverage	- What are the best ways to target different kinds of leverage for systemic change? (E.g., how might we help actors in a system track all of the relevant paradigms?)	- Fruitful friction as a tactic for transcending paradigms (Buckenmayer et al., 2021) - Systemic change happens via multiple dimensions of change (Mulder et al., 2022) - Design Journeys offers several chapters on taking action after identifying leverage points (Jones & Ael, 2022)	- “Systemic change labs” tracing and comparing the impact of interventions using different kinds of leverage	- How to design innovations for each type of leverage

Some other key takeaways:

The concept of “leverage points” dominates modern discussions of leverage, but as Meadows (1997) herself proposed, that is just one paradigm we can use to view the best ways to produce systemic change.
There are good and bad kinds of leverage points! See figure 1.
A few promising insights about leverage have been proposed recently, such as the notion of “chains” of leverage points (Fischer & Riechers, 2019) and the idea of assessing potential for change (Birney, 2021).

Leverage points can be futile, catalyzing, or transformative, and they progressively reform or regressively deform our systems.

Keeping a Working Copy git repo synchronized with Shortcuts

Published Feb 23, 2023

Use these Shortcuts to automate syncing a git repo on iOS via Working Copy.

How to learn the most from the Obsidian community

Last updated Feb 14, 2023 | Originally published Feb 14, 2023

Think of a curiousity or a question
Search!
a. … the web (e.g., via the unofficial Obsidian community search engine) b. … the Forum for information ( use the forum’s Advanced Search capabilities)
c. … the Discord server ( learn more about power search tools in Discord here)
Ask on the Discord or the forum! (Make sure you review the list of channels in Discord to find the best place to post.)
Get more ideas and start again at (1) 😉

If you learn to make the most of the different pools of knowledge, you can always find rich answers — and you’ll ask better questions, too.

A Templater script for updating file titles and dates in YAML

Last updated Dec 15, 2022 | Originally published Dec 15, 2022

This Templater script gives the current Obsidian note a proper YAML title, date, and lastmod field. This facilitates integration with Quartz/Hugo.

It requires MetaEdit (as well as Templater, obviously).

Update file titles and dates with this Templater script

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60


<%*
let file = tp.file.find_tfile(tp.file.title)
const {update} = app.plugins.plugins["metaedit"].api
const {getPropertyValue} = app.plugins.plugins["metaedit"].api
const {position, ...rest} = tp.frontmatter;
let content = tp.file.content;
let newFileContent = content.split("\n");
let isYamlEmpty = Object.keys(tp.frontmatter).length === 0 && !content.match(/^-{3}\s*\n*\r*-{3}/);

/* reset selection to the top of the document to make sure the action doesn't clear any text */
let editor = this.app.workspace.activeLeaf.view.editor;
if (editor.getSelection === "") {  // no text is selected
} else {
	editor.setCursor(editor.getCursor());
}


async function updateCurrentFile(someContent, someFile) {
	someContent = someContent.join("\n");
	await app.vault.modify(someFile, someContent);
}

let propNameForLastModified = "lastmod";
let fileLastModifiedDate = "\"" + tp.file.last_modified_date("YYYY-MM-DD\THH:mm:ss") +"\"";

if (isYamlEmpty) { // No YAML yet
	newFileContent.unshift("---");
	newFileContent.unshift(`${propNameForLastModified}: ${fileLastModifiedDate}`);
	newFileContent.unshift("---");
	await updateCurrentFile(newFileContent, file);
} else if (rest.lastmod === undefined) { // YAML exists but no date field
	newFileContent.splice(1,0, `${propNameForLastModified}: ${fileLastModifiedDate}`);
	await updateCurrentFile(newFileContent, file);
} else { // YAML exists and a date property exists
	await update(propNameForLastModified, fileLastModifiedDate, file);
}

/* Now we can assume YAML exists, so let's add the rest of the metadata */
let propNameForDate = "date";
let fileCreatedDate = "\"" + tp.file.creation_date("YYYY-MM-DD\THH:mm:ss") + "\"";

if (rest.date === undefined) { // YAML exists but no date field
	newFileContent.splice(1,0, `${propNameForDate}: ${fileCreatedDate}`);
	await updateCurrentFile(newFileContent, file);
} else { // YAML exists and a date property exists
	await update(propNameForDate, fileCreatedDate, file);
}

let propNameForTitle = "title";
let fileTitle = "\"" + tp.file.title + "\"";

if (rest.title === undefined) { // YAML exists but no title property exists
	newFileContent.splice(1,0, `${propNameForTitle}: ${fileTitle}`);
	await updateCurrentFile(newFileContent, file);
} else { // YAML exists and a title property exists
	await update(propNameForTitle, fileTitle, file);
}

new Notice (tp.file.title + "'s title is now " + fileTitle + ". Last modified date metadata updated to " + fileLastModifiedDate + ".", 2000);
_%>

Why the grass is greener: Making sure that shiny new alternative tool is actually going to help you

Last updated Nov 3, 2022 | Originally published Nov 3, 2022

“The grass is always greener on the other side.” The popular idiom discourages whoever’s listening from seeking out alternatives, suggesting that other options always look better from wherever we’re currently standing. But it has a funny problem: nobody ever explains why the grass is greener on the other side.

That’s because it isn’t. The truth is that your side is just yellower/trampled on/eaten… and that’s because you’re on it.

Moving to a different place will be fine at first. Then you’ll use it, too, and eventually it’ll look the same as where you started.

(In this metaphor, you’re a goat. 🐐)

In workflow design, in addition to the novelty of “shiny new object,” new and alternative tools are great simply because they don’t have the cruft you’ve built up in the old tool. That cruft might be noisy notes, a lifetime of guilt-inducing task management, or even just bad habits and behaviours. The problem isn’t the tool. It isn’t you, either. It’s you and the tool.

So, after switching, the problems seem to go away… only to re-emerge (possibly in a new form) later because the issues are generated by your usage, not by the tool.

The solution is to fall in love with the problem, not the (shiny, potential) solutions.

Determine what your issues actually are, and try to figure out why they’re happening.
Then, abstractly identify how you might be able to mitigate the problems.
- Don’t say “I’ll use Bunch,” say “If I standardize certain work spaces on my computer, I can develop muscle memory for using those workspaces, reducing distraction and allowing me to spend less cognitive energy on finding everything I need to get engaged.”
Last, identify some tests or success conditions that will tell you whether the solution is actually working. This’ll help minimize irrational perspectives on how well the honeymoon stage is going.

Only after taking those three steps should you choose a tool. Find something that can implement the abstract principles you’ve articulated, and be sure to follow-through on the tests.

In doing this, you’re actually creating and implementing a rough design theory. You’re using design science to make your work as easy and engaging as it can be! High-five for that.

𖧹/△ Fulcra

All Posts