Software

DEVONthink or EagleFiler (or Finder?)

Posted on December 23, 2022 by Jack Software

One of the many things I waffle about is the choice of software for managing my many PDFs and other files. I often wish I could just keep everything in organized folders and use my Mac’s Finder to manage it all. That sounds great on paper, but never sticks.

I’ve used DEVONthink for years because it is so mature and powerful. I continue to find new features or techniques even after all this time. It does everything!

EagleFiler comes into play when I’m feeling overwhelmed by DEVONthink. EagleFiler is much closer to the metal, so to speak. It’s basically a thin, smart layer over a set of folders. Everything is accessible “natively” and its capture and organizational features are simple and useful.

But after using EagleFiler for a while, I start to wonder why I’m not just doing everything in Finder. I mean, if we’re going to mess about with files in the filesystem, why not just do that then? What was EagleFiler offering me, again?

After a short time, I end up back with nicely organized files and folders and it’s such a relief not needing to deal with databases or any of that nonsense!

Except it’s harder to get things into the right place using just Finder. It requires me to, for example, save a PDF to my Desktop then find and open the destination folder, then drag it on in. This makes quickly capturing stuff cumbersome. Search is fine, but harder to use than in the other apps. And I miss the way DEVONthink would intelligently sort and rename files for me. So always return to using DEVONthink, until I end up back at the beginning.

Where are we today, then? Today, it’s DEVONthink. My feeling right now is that if I’m going to abstract myself away from the actual files, I may as well use the tool with the coolest, fanciest ways to do that.

However, I can feel the pull of a simple set of folders and files. I am thinking about ways of making 2023 the Year Of Less Software, so stay tuned. 🙂

Neofetch

Posted on December 6, 2022 by Jack Tech

I just learned about Neofetch:

Neofetch displays information about your operating system, software and hardware in an aesthetic and visually pleasing way.

Here’s my MacBook Pro…

A tweak to the photo workflow

Posted on June 25, 2021 by Jack Misc

I’m trying to stick with the Adobe suite for processing, editing, and managing photos.

I prefer Capture One’s editing process, but Lightroom Classic has everything else going for it, (ecosystem, tooling, ubiquity, etc.) so that’s where I’ve settled for now.

But I’d love to take advantage of Lightroom CC on mobile and my laptop. CC and Classic will sync, but if not handled properly the whole enterprise can quickly turn into a mess. What I was doing is to import into Classic, edit, export, then add the “keepers” to a synced catalog (or “all synched photographs”) so that those photos would be available everywhere. The problem is that this takes diligence and consistency. It takes work. I’m not good at consistency, and I end up frustrated and bailing on the whole thing.

So here’s what I’m trying. I’m reversing the process and importing directly into Lightroom CC instead. I cull and rate the photos there. I do simple edits and enter captions. For any images I’m more “serious” about, I launch Lightroom Classic which automatically syncs all the images from CC. While I’m there I copy the files to my usual places on the filesystem and rename if desired. All this can be done in Classic and the photos still remain synced and available in CC.

One downside is that when syncing from Classic to CC the photos don’t count toward my subscription’s storage, which is nice, but going the other way takes up space. I think this will be OK. If I do come home with cards chock-full of images I’ll just start in Classic instead.

This also means I can enable auto-import from my phone’s library and have everything show up automatically. I have to be careful here, because if I want to keep Apple Photos app as my final library (for sharing, showing people, and ease of OS integration) I can end up with duplicates this way.

Lightroom CC is a more pleasant place to live than Classic, so for 80% of the time it’s good enough. For the other 20% I head over to Classic.

Update July 11, 2021: I’m mostly back to only using Lightroom Classic. Too many moving parts trying to wrangle both.

A visual thinker using text-based tools

Posted on June 18, 2021 by Jack Misc

Yesterday I was asked something about a project I’d worked on two years ago. At that time I’d used Curio to help manage the project. I opened the Curio project and within thirty seconds of just looking at the workspace I had a handle on the project and easily found an answer to the questions I’d been asked.

Whenever I revisit something that I’d created in TheBrain or a mind map or Curio or Tinderbox , I find the spatial layout of the information to be instantly useful.

And yet I use Org mode in Emacs for nearly everything. You can probably tell that I’m having another one of my moments.

I love plain text. Or maybe it’s more accurate to say that I love the idea of plain text. Nearly all the arguments for using plain text are good arguments, but that doesn’t make plain text any more useful for me.

Plain text’s usefulness depends on what it’s for. As an example, a simple log of things that happen throughout the day makes sense as plain text. It’s almost always going to be accessed via search, and text is made for searching. Journaling can be done in plain text, although it’s made better by including images.

The way text is presented can make all the difference. An example is the display of backlinks in Roam and Logseq. Those tools use a nicely-formatted display, including context. Compare it to something like org-roam, which, as powerful as it is, can’t compete visually. It’s hard to parse backlinks in org-roam just by looking at them. And that’s a problem system wide. A wall of text is less useful than a purposefully-arranged and formatted visual display of that same information.

Anyway, I launched Curio and Tinderbox and TheBrain and now I’m in big trouble.

Using Zotero as a bookmarking and read-later service

Posted on June 16, 2021 by Jack Misc

I’m almost certainly using Zotero wrong.

Instead of for citations and research, I’m using Zotero as a bookmarking tool and read-later service, and it’s working really well. Is no one else doing this?

I’ve used many tools meant for saving links for later, from del.icio.us to Pinboard to Instapaper to Pocket to Raindrop. All of them are fine. Some focus on social bookmarking, some on archiving, some are meant as “read later” services. And all of them are prettier than Zotero. And yet…

I installed Zotero while tinkering with an Org mode note-taking workflow. Soon after, I installed the “Save to Zotero” Safari extension and started using that instead of my usual “Save to Pinboard” bookmarklet, just for something different. I was surprised to find that this has become my default.

For free, I get smart metadata parsing and tagging along with old-school hierarchical organization. I get full-page offline snapshots and sync. I get PDF annotation and storage. Oh, and I get citation management I can use if I ever want to sound smarter than I am.

It’s only been a month or two, but it feels like I have a good start on building a nicely-organized reference library as a byproduct of bookmarking things to read later.

A reluctant Lightroom user

Posted on May 18, 2021 by Jack Misc

I’ve never loved editing photos in Adobe’s Lightroom (Classic) . It does the job fine, and it has all the tools one might need, but it’s no fun. I prefer editing with Capture One Pro .

As much as I enjoy the editing process in Capture One, it otherwise feels like working on an island. C1 has no way to sync photos, the plugin/extension options are very limited, and while it works with other editors, it doesn’t do it as seamlessly as Lightroom. And so on.

Lightroom’s ecosystem is hard to beat. It works with nearly everything. I can sync with the mobile Lightroom CC library. It works with every plugin one could possibly want (most notably, Jeffrey Friedl’s and Negative Lab Pro .) There’s no end to the presets and styles available. If I want to do something with a photo, Lightroom is more likely to be able to handle it.

Lightroom’s cataloging is more capable than Capture One’s. Or at least it feels easier to use. I’d prefer not having to rely on a specific vendor’s tool for managing my lifetime of photos, but it’s better than only having them scattered in folders. Believe me, I’ve tried it that way numerous times. It’s liberating, but only for a moment, then it quickly becomes frustrating.

And of course I don’t like having to pay a monthly subscription to Adobe. That said, for $20 a month I get Lightroom Classic, Photoshop, Lightroom CC on all my devices, and 1TB of cloud storage. It’s a pretty good deal.

While I’m still looking for excuses to go back to Capture One, I am, reluctantly, back in Lightroom Classic for the majority of my photo management. For now.

Writing everything in TiddlyWiki and publishing just the public parts

Posted on May 8, 2021 by Jack Misc

I take all my notes in TiddlyWiki now, and publish most of them to wiki.baty.net .

For the past few years, I’ve published my wiki using TiddlyWiki. I write daily, publicly sharable notes there. Private stuff goes elsewhere…or did, until yesterday.

It’s the “elsewhere” part that drove me nuts. I have a private Roam database in which I would track things I don’t want to share. Or maybe I should write it in Org mode . Or Obsidian , or Craft , or or or. The difficult part for me has been that I want to take a note about, say, a new camera purchase. There are two components to it, the information about the camera itself, and information about the purchase. The former is public, the latter is private. This means I create one note in TiddlyWiki and one in, let’s say, Roam. There are dozens of examples like this, and it’s crazy-making. I thought I could manage this using links or copy/paste but it sucks trying to do that. I could also make everything public or private. Neither of these are feasible.

If only I could keep everything in one place, but only publish things I wanted public. Then, a few days ago, Soren Bjornstad came to the rescue with his video, A Tour Through My Zettelkasten .

Wow, other than building an amazing Zettelkasten, Soren has implemented nearly everything I needed in order to go all-in with TiddlyWiki for my own wiki.

A few highlights:

Public and Private tiddlers
Sensible tagging and organization
Override the “copy permalink” feature to substitute public URL when on localhost
Scripted rendering and publishing of public wiki
Specific behavior when viewing public vs private editions
A number of other nice touches

I borrowed some of these and integrated them into Rudimentary Lathe . Now, I’m taking all my notes in TiddyWiki. I’ll describe the process a little.

Editing the wiki locally.

I use TiddlyWiki as a local Node.js app. While one of TiddlyWiki’s great features is that can be just a single HTML file, running it locally as a single-page web app via node.js makes things a bit more flexible. Also, it’s the easiest way to allow for saving changes in Safari. The file structure looks like this:

├── files/
├── plugins/
├── tiddlers/
└── tiddlywiki.info

All tiddlers are kept as separate “.tld” files in the tiddlers folder. Here’s an example:

created: 20201220181044760
creator: jack
modified: 20210505182021507
modifier: jack
revision: 0
tags: Public
title: Leica APO-Summicron-SL 35mm ASPH
type: text/vnd.tiddlywiki

[img[files/2020/leica-apo-summicron-sl-35mm.jpg]]

I prefer primes, so this is the one I've chosen for the [[Leica SL2-S]]. Watching Peter Karbe admit it's is desert-island lens and suggesting it's the best lens Leica has ever produced made the decision a little easier.

I have over 2300 of them currently. Another nice side effect is that git diffs are much more usable on individual text files than on a giant HTML file.

Public vs Private content.

Any tiddler I want to be public gets a “Public” tag. That’s it. The export script is smart enough to automatically include all system tiddlers, etc so that everything works.

As a useful helper, each tiddler displays a “Publish this tiddler” checkbox to make adding the tag easier, as well as serving as a handy indicator of private vs public status. The export script updates one of the configuration tiddlers so that the published version doesn’t show this checkbox.

I can’t tell you how huge this is. Not having to choose the tool or app for new notes is so liberating. I can now write and link freely with everything and can still share most of it publicly.

Hosting

I’ve never used Github Pages for hosting any content, so thought this would be a good opportunity to try it. Basically, I keep a separate repo of the public version and pushing to that repo automatically publishes it. Super easy to set up.

Publishing workflow.

Soren was kind enough to share a version of the script for publishing his wiki (publish.sh), which I’ve modified slightly. Here are the highlights.

PRIV_FOLDER="rl-wiki"
PUB_FOLDER="public-wiki"

FILT='[is[system]] [tag[Public]] -[[$:/plugins/tiddlywiki/tiddlyweb]] -[[$:/plugins/tiddlywiki/filesystem]] -[prefix[$:/temp]] -[prefix[$:/state]] -[prefix[$:/sib/StorySaver/saved]] +[!field:title[$:/sib/WriteSideBar]]'

WIKI_NAME="index.html"ext_image_folder="extimage"

“FILT” is the tiddlywiki filter for determining which tiddlers to include (and exclude). The [tag[Public]] bit is the key to the public/private thing.

Then we export tiddlers based on the filter and settings above.

"$(npm bin)/tiddlywiki" "$PRIV_FOLDER" --savewikifolder "$pub_wiki" "$FILT"

Next, generate a single HTML version of the wiki and copy over the separate image files..

"$(npm bin)/tiddlywiki" "$pub_wiki"
    --render "$:/core/save/all" "$WIKI_NAME" text/plaincp -r "$pub_wiki/output"/* "$pub_ghpages"cp -R "$PRIV_FOLDER/files" "$pub_ghpages"

Isn’t TiddlyWiki amazing!?

Finally, we commit and push the public wiki to Github…

if [ "$1" = "--push" ];
then
echo "Pushing compiled wiki to GitHub..."
cd "$pub_ghpages" || exit 1
git add .
git commit -m "publish checkpoint"
git push
else
echo "Not pushing the wiki to GitHub because the --push switch was not provided."
fi

And voilà!

A few nice odds and ends.

Soren’s “Reference Explorer”, seen at the bottom of individual tiddlers, replaces my handmade backlinks display. His is much fancier. I removed a few tabs I don’t use, and may exclude the tags at some point. I conditionally exclude the explorer from my Daily Notes pages. (anything tagged “DailyNote” hides the explorer.) Another nice tweak is that if I add a “refexplorer-hide” field to any tiddler and set it to “true”, the explorer is not shown on that tiddler. Nifty.

TiddlyWiki comes with a button for copying a permalink to each tiddler. The problem with that for me is that when I’m running the wiki locally, permalinks look like this

http://localhost:8080/#CommandLineInterface, which obviously won’t work. Soren’s version of the button replaces localhost:8080 with the live hostname, e.g. https://rudimentarylathe.wiki/#CommandLineInterface . This saves me a ton of copy/paste/edit hassles.

Putting it all together.

When I’m ready to publish, I open a terminal and type prl (for “publish rudimentary lathe”)

prl is a script…

#!/bin/shcd ~/Sync/rudimentarylathe./scripts/publish.sh --push

That’s it.

I wish more people would spend time getting to know TiddlyWiki. It’s amazing. It’s a Quine , which makes it ridiculously flexible and powerful. And yet it’s very simple. It’s also a free, local-first, easily-distributable, storable, backup-able single HTML file.

TiddlyWiki is fun, fancy, and
future-proof. I live there now.

Are automatic backlinks useful?

Posted on April 10, 2021 by Jack Misc

When I started using Roam, I found the way it handled backlinks to be a revelation. Other software does backlinks, but Roam’s implementation made it feel new. Suddenly, backlinks felt necessary.

I started writing everything in Roam’s Daily Notes, and I’d link things by putting brackets around each word or phrase that I thought I might want to review later. I made lots of links. After a while, I noticed that many (most?) of these linked words and phrases would end up as empty Roam pages containing nothing but backlink references.

In effect, what I was doing was creating saved searches.

I noticed something similar in my TiddlyWiki at rudimentarylathe.wiki . The automatic backlink references at the bottom of each note were in most cases links from one of the daily notes, and this ended up as a collection of backlinks like “2021.04.10 – Daily Note”. Not very helpful. Would I be better off just searching for the topic in question? I think so, so I recently changed the note footer on the wiki to do just that. It now shows a list of “tiddlers” with the most mentions of the current tiddler and also those with titles containing the same word. You know, like saved searches.

I’m starting to question the value of automatic backlinks in my notes. I still want them, but I’m not sure I need them the way I thought I did. They no longer feel necessary, but are they useful?

I started thinking about this again after re-reading Sascha Fast’s post, Backlinking Is Not Very Useful — Often Even Harmful . I had an adverse reaction to the article when I first read it. I thought it was mostly sour grapes because Roam was eating The Archive‘s lunch. I read it as, “The Archive doesn’t have automatic backlinks, so they must be bad and you don’t need them!” There was this right in the first paragraph:

Automatic backlinks are not only automatic when there is software that is showing them for you. If you create a backlink apparatus by habit it is still automatic. The automatization software would then be in your head

That felt like some rationalization gymnastics right there. I looked up “automatic” and it said, “done or occurring spontaneously, without conscious thought or intention”. A habit of manually creating links still involves conscious thought and intention.

And then the article went on to try and cast automatic backlinks as “linking notes” vs manual linking as “connecting knowledge”. I thought that was a bit of a semantic crutch.

Here’s another stretch…

Just think a moment about how difficult it really is to use the internet and its web in a productive way. The single most productivity-destroying problem with using the internet is the temptations link provide. The same is true for your Zettelkasten if your link structure is not well-groomed.

To compare the distraction of links on the internet at large to those within my own writing in a curated Zettelkasten doesn’t seem at all relevant.

I was looking for problems with the article going in, and I felt that I found them straight away, so I skimmed the rest with a jaded eye and a bad attitude.

I may have overreacted.

Now that I’ve spent a year building my notes using tools that make backlinks easy and automatic, I’m coming around to Sascha’s point of view. I have hundreds of empty “pages” containing nothing but backlinks. There’s no context, no color. No knowledge. It’s just “linking notes”. Here’s Sascha’s closing comment:

Backlinks are a perfect example on how features of software not only can be useless but actively harming you work by redirecting your attention towards to the superficial belief that you need to place links, instead of trying to connect knowledge.

I still feel that “harmful” is an exaggeration, but I better understand his point now that I’ve had some time with it.

The answer to the title of this post is, of course, “It depends”.

Many people use the term “zettelkasten” for any old pile of notes. But for a true zettelkasten, one containing notes specifically intended to help garner and build one’s knowledge over time, automatic backlinks aren’t as useful. Explicitly linking between ideas and notes and providing context for the links is much better.

For other collections of notes, though, automatic backlinks can be very helpful, even though they aren’t functionally much different from saved searches. For example, I keep notes about people I know. Having backlinks created automatically whenever I link to “Mom”, for example, is a nice way to see all of the times I’ve mentioned her, right there next to my notes about her. This, for me, is very useful. Hi mom!

TiddlyWiki is more fun than Roam

Posted on November 22, 2020 by Jack Misc

I fell in love with TiddlyWiki almost exactly 2 years ago . I wrote in it almost daily until late August, 2020 , when I moved full-time into a public Roam database.

Roam is great and I love it. I’ve tried everything else, and nothing beats Roam for easily taking, linking, and re-using notes. I’m still using a private Roam database for work projects and CRM-type stuff, and it’s great for that.

Roam is efficient, fast, clever…and boring. Easy isn’t the same as fun.

TiddlyWiki is fun. It’s playful. I can’t really explain it, but creating new “tiddlers” and messing around with customization and finding new organizing principles is actually enjoyable in TiddlyWiki. I am probably one of only a handful of people who actually prefer the separate view and edit modes.

So, I’m going to return to using TiddlyWiki instead of Roam for my daily notes and scraps. My wiki is still at rudimentarylathe.org , which is a thousand times more fun as a URL than https://roamresearch.com/#/app/jackbaty am I right? I expect Roam will offer custom domains at some point, but so for they don’t.

I feel that TiddlyWiki’s local-first, single HTML file, free and open-source approach is better suited as a place to do “public self-modeling” for the long term.

Plus, it’s more fun. I hope to see you there !