Something I've always liked about TiddlyWiki is that I'm never forced to decide where something goes or what it's named. I can simply click on the new tiddler button, type something, and hit save. Links, tags, or other organization can be applied later, but none are required.

The Howm Emacs package works similarly. By default, Howm^[1] notes are organized in YYYY/MM/ folders and file names are automatically created based on the current time, e.g. 2024-05-07-130712.txt. I don't have to think about them.

It's nice to just make a new note and not care where it goes. I sometimes get twitchy about not having useful file names. "What about accessing your notes outside Howm?" First, I never do that. Second, there's grep, Spotlight, or any other text editor that can search a folder full of plain text files. This is not something I need worry about. In fact, it's slightly better than with TiddlyWiki, since to find something in TiddlyWiki, I need a web browser to open the wiki and search. There's only one useful way in. Of course that's never a problem, but still.

This is fine:

I continue to be surprised by how quickly Howm has become an important part of my workflow. I should write about how the combination of Howm and Denote fit into my process.

In the spirit of /Now and /Uses and /Hello pages, I've created a /Nope page.

It's for keeping track of things I don't like or want to do. It's a work in progress, but it'll probably see more updates than my now page 😄

In a post from Andreas:

Well, of course it’s great, even fantastic - for Adobe. They can lure you in with low prices, then gradually make the subscription more expensive, and then move the features you use to the “premium” tier where you have to pay even more. Not to mention that they can remove features on a whim if they feel like it, or charge you even more if you want to keep them. And if you want to keep access to your files, you have to be subscribed until the end of time - “nice pictures you have there! It’d be a shame if you couldn’t open them because you decided to opt out of our subscription model, wouldn’t it?”

I dislike subscriptions. I rage-quit everything Adobe a few years ago in a huff about subscriptions and, well, I don't love Adobe. However, I'm once again paying a subscription to Lightroom Classic and Photoshop.

Why? First, because the product is, overall, more appropriate for me than the alternatives. And second, in seven years of subscribing to the Adobe "Photography" plan, none of the things Andreas mentioned have happened. The price is the same (I'd happily pay double). They've only ever added features (I don't remember a single feature being removed. Has there been any?) If I cancel my subscription, I still have 100% access to my files (I just can't make additional edits).

I'm not trying to single out Andreas. His post just reminded me that these "What if...?!" doomsday scenarios have guided way too many of my decisions. What if [SOME APP] stops being developed? What if Apple behaves even more badly? What if some "proprietary" (usually sqlite, so not really) database becomes corrupted? What if I can't read [FILETYPE] in 50 years?

Of course these things can happen, but how often do they? Basically never, has been my personal experience. And if they do happen, there's almost always a reasonable way out.

Why suffer using something we don't love, on a just-in-case, instead of something we do love and find immediately more useful, because "what if!?"

Previous

I love software. I love learning what software can do. I love setting up workflows using all the fun software I've learned about. The problem is that this leads to chaos and complexity. Every time.

You're probably wondering what made me think of this. Well, yesterday I was exhausted from fighting with Obsidian and Emacs, so I installed Bear. You see what I mean? I spent an hour this morning importing stuff from Obsidian and cleaning up tags, etc. Bear is so nice and simple and absolutely does not lend itself to tweaking.

Problem solved! Not really.

What about daily notes? What about templates? What about complex exports? Why is folding so cumbersome? Where are the backlinks displayed again?

Sigh. I really appreciate Bear, but can I live in it? Probably not. I love simplicity, but can't abide constraints. I'm screwed, I guess.

I still keep an actual calculator on my desk, and use it regularly. I find it easier and more "stable" than using either a calculator app or even the built-in calculator in Raycast.

I bought the Casio shown above years ago because it had a large screen and doesn't use batteries. It's fine, but there are a couple of things about it that bother me.

First, there's no "On" button. Ok, there is, but it's buried under the AC button. Bugs me.

The On button issue is bad enough, but that's not why I hate the Casio. I hate the Casio because when I turn it off, the screen reads "CASIO" for a few seconds before actually turning off. It's like watching a little ad each time. I hate it, so I bought a new calculator.

This is the re-issued version of the iconic, Dieter Rams designed Braun ET66 (1987). I recently watched the Dieter Rams documentary, "Rams" (2018), which reminded me of the Braun, so I thought this model would be a fine choice.

I like it. It's simple, clear, does what I need, and looks good doing it. And it doesn't shout its own name every time I turn it off. Much better.

Look what I found in the garage:

It's my old Apple Extended Keyboard II from 1990 or 91. I last used this one in 2015.

The AEKII uses an ADB port, so I had to dig out my ADB->USB-A adapter (save everything!). I'm typing this post on the keyboard right now. The Alps switches are as great as I remember, and might be my all-time favorite switches.

Beyond the great switches and nostalgia, using the keyboard leaves a bit to be desired. First off, it's enormous. It takes nearly all of my felt deskpad and leaves barely enough room for a mouse. And speaking of the mouse, the size of the keyboard means that the mouse is over 12 inches away from my right hand. Reaching for it is a whole thing.

Another quirk is that the little home row nubbins are on the D and K keys. On newer Apple keyboards, they are on the F and J keys. I can't tell you how many times I've had to delete a bunch of gibberish and reorient my hands while typing this.

I could get used to the size, but the AEKII's Caps Lock key can't be mapped to Control because it physically locks down when pressed. It thinks it's a manual typewriter, I guess. This is a deal-breaker.

Even though it'll probably end up back in the box soon, it's fun using a 34-year-old keyboard that still works and is, in many ways, better than anything made today.

I switched from using Lightroom Classic (LrC) to Capture One Pro (C1) "for good" back in 2021. It wasn't because of Adobe's subscription model, or because I had some vague aversion to Adobe, the company. It was because I felt like I was getting better images, faster, with C1.

I kept a few notes on Lightroom Classic vs Capture One but haven't updated it in a while, so here are a few notes about why I have moved back to Lightroom Classic.

C1 has become laggy for me. I'm working on either a MacBook Pro M1 Max or Mini M2 Pro. Nothing should be slow on these machines, but whenever I move a slider in C1, I have to wait a couple of seconds while it catches up. This gets pretty annoying.

C1 never seems to improve its cataloging features, which have always felt a bit lackluster. I like C1's "sessions", but then I need some other app for cataloging. I'm working on using less software, not more, so cataloging is important.

Aside from specific issues, the thing that has put me off C1 is that over the past couple years, Capture One (the company) has made some moves that have affected how I feel about C1, (the app).

Capture One now basically forces me to pay for a subscription, and it's expensive. I know I said that subscriptions weren't the reason I switched, but that doesn't mean I like subscriptions. I don't.

More importantly, they've become increasingly focused on professional, in-studio customers. That's great for pros, but I'm just a guy who likes to take pictures. I'm paying a lot of money for software that is becoming less relevant to me.

The latest beta update of C1 introduced a "Studio" version and includes things like Client viewers and "Live" for Studio. These are not just less useful to me, they are useless to me. I started wondering if I'd be better served by Adobe. That was a surprise.

So I forked over my $10 for the Photography Plan subscription and installed Lightroom Classic and Photoshop (still one of the best deals going, regardless of how one feels about subscriptions). I resisted the temptation to install the new Lightroom, because I know how that ends.

I started by importing everything from 2020 through today. Since I'd used LrC for years, I was immediately comfortable. I made the few changes that LrC allows to the UI and settings, imported my old presets, set up my export templates, and was off and running.

It was nice to have nearly-instant response while spotting film scans again. It's good to have a catalog that feels robust. LrC feels tighter than C1 somehow. And the community is so big. Almost too big, honestly. It can be overwhelming. Adobe's huge ecosystem is valuable. I've never been great in Photoshop, but I'm better there than in one of the "other guys" like Affinity. And if I don't know how to do something, a dozen YouTube videos showing me how are within arm's reach.

I'll miss some things about C1, certainly. I'll miss the wildly-customizable UI. I'll miss that hitting "Auto" is often all I need. I'll miss the fancier exports. I'll miss the culling features during imports. And if I ever do get in a studio, I'll miss the flawless tethering.

If Capture One ever decides to cater to amateurs like me as well as pros, I'll take another look, but for now, I'm in Lightroom Classic. I feel a little dirty about it, though.

Originally posted on Scribbles: https://scribbles.baty.net/post/lightroom-classic-again

I know, I know, the reMarkable 2 tablet is designed as a digital replacement for writing on paper. It's not necessarily meant as a document reader. But that's what I've been using it for.

For years now I've saved web articles as Markdown files, converted them to PDF, and printed them for reading later. See My read-later service is made of paper.

I wanted to see how reading PDFs felt on the reMarkable, so I copied a few of my saved article PDFs to it. It turns out that I liked it very much. Highlighting text with the stylus works great, and I can even choose the color of the highlights, which don't display on the device, but they do on the actual PDF.

My only complaint was that the text was too small. To fix this, I made a new Pandoc template so that the font was larger, and the margins were smaller. I included some extra room on the right, so that the reMarkable toolbar doesn't cover the text. The script outputs the PDF directly to the Dropbox folder I have synced with the tablet. Here's how it looks with the updated template.

Nice, eh?

I create these by converting articles to Markdown using the Markdownload plugin. From there, I open the file in BBEdit and run a shell script which runs Pandoc to generate the PDF. Both the script and Pandoc template can be found here if you're interested. Note that they would need to be tweaked for your environment. This can easily be done via command line or just a shell script, but I've had it configured via BBEdit for so long that it's habit.

howm is an Emacs package for taking notes. It was recommended to me recently, so I thought I'd take a look. The project page says "howm: Write fragmentarily and read collectively." Worth a shot, right?

I haven't seen too many people talking about howm. The best introductions I've found are from Leah Neukirchen and Andrei Sukhovskii.

I installed it via use-package with the following

;; howm package config

(use-package howm :ensure t
  :config
  (setq howm-directory "~/Documents/howm/")
  (setq howm-home-directory "~/Documents/howm/")
  (setq howm-keyword-file (expand-file-name ".howm-keys" howm-home-directory))
  (setq howm-history-file (expand-file-name ".howm-history" howm-home-directory))
  (setq howm-view-use-grep t))
;;  (setq howm-view-grep-command "/opt/homebrew/bin/rg"))

;; Fix for help bindings
(define-key howm-menu-mode-map "\C-h" nil)
(define-key riffle-summary-mode-map "\C-h" nil)
(define-key howm-view-contents-mode-map "\C-h" nil)

;; Sensible buffer names
(add-hook 'howm-mode-hook 'howm-mode-set-buffer-name)
(add-hook 'after-save-hook 'howm-mode-set-buffer-name)

I couldn't get the rg settings to work, so I'm still using grep. It's fast enough for this test.

If you're interested in getting started with howm, you should read the articles linked above. Here's what I found interesting.

• Major mode doesn't matter. Howm works with everything from plain text files to Org-mode.
• It's date-centric. Most actions prioritize temporal nearness and files are named with just a date string.
• There are some clever built-in todo/scheduling features. Just put in dates like [2024-03-25]+ and howm figures out how and when to show it.
• Linking and backlinking are simple. Put >>> some phrase anywhere and it turns into a link which searches for that phrase when clicked. Put <<< term in a note and every other occurrance of that term will be underlined and linked to that note.
• howm hates evil-mode. If I were to try getting anywhere with howm, I'd need to figure out my own collection of custom bindings in evil-mode because by default they don't get along at all.

I created a handful of test files, and here's how howm looks so far.

And here it is browsing some notes.

Howm is neat, but I don't expect to move my notes into it just yet. I find Denote to be nearly perfect for note taking, so I'm all set. I do think a combination of howm in plain text files and Hyperbole to add some juice would be a nice combination.

When I ordered the M2 Mac Mini I opted for the smaller 512GB internal SSD knowing that, since the Mini would be always-on, I could hang as much storage off it as I wanted.

A hodgepodge of external drives scattered over my desk is not the prettiest setup, though. I wanted something cleaner. Something nicer-looking. I went with an OWC miniStack STX.

The miniStack can be ordered pre-configured with a variety of storage options, but I figured I'd save a few bucks and set it up myself. I ordered a 2TB Gen 4 M.2 NVMe stick and a Seagate IronWolf 8TB HD. It took about 10 minutes to install both drives into the miniStack and I was ready to go.

The whole point of the thing is that it is the exact dimensions of a Mac Mini. It sits underneath the Mini and so takes up zero additional desk space. This is way better than normal external drives.

An important bonus with the miniStack is that it adds three Thunderbolt 4 ports (actually four ports, but one is used by the Mini->miniStack cable).

There's a small catch. The PCI bus only uses one channel so top transfer speeds are limited to around 750MB/s. I don't know anything about NVMe drives or PCIe buses so I'm just taking their word for it, but my disk speed test showed around 790MB/s. That's plenty for my purposes.

I still need the CalDigit hub for the SD slot, Optical audio, and extra USB-A ports. Here's what the stack looks like:

Setting all this up gave me an opportunity to tidy up the rest of my desk. I think it looks great.

So far I've not heard any noticeable fan noise, so that's good. I do hear the HDD grinding away but that's probably due to both Spotlight and BackBlaze doing their things.

The 2TB SSD will be used for working files and recent photos. The big HDD will contain the rest of my media and photo archives, along with anything else I want near-to-hand but don't need fast-transfer access to. That leaves a 1TB SSD (not shown) hanging off the back for Time Machine and the vertical 8TB HDD for nightly backups using Carbon Copy Cloner.

It's a nice upgrade.

My life would be simpler if I could convince myself to stop worrying about things like^[1]:

• How should I rename photos during import?
• What if the app goes away?
• How will I find this later?
• Are my notes are in the right app or format?
• What if I switch to Linux?
• Are file names consistent everywhere?
• Kebab case or Camel case?
• Has anyone at the company done anything problematic recently?
• What will people think of me if I start/stop using this app?
• Will this be readable in 100 years?

Almost none of these are truly important, even though they all seem very important at the time.

I tried, albeit half-heartedly, for 18 months to grab a Fujifilm X100V at a "normal" price. The X100 series are awesome cameras. I still have an original X100, and had an X100T for a while. The original X100 is just so very slow that I don't often grab it. When the new X100VI was announced, knowing they'd probably be in short supply soon, I pre-ordered one at both Adorama and B&H.

After missing out on the first batch of cameras, it was clear that I was in for a long wait. This gave me time to think about cameras.

I don't need a new digital camera right now. I'm barely using the ones I have, which are two very nice cameras. See for yourself...

The little GRIII is great for carrying everywhere and I love the files I get from it.

The SL2 is for when I'm being "serious" about photographing something.

Where would the X100VI fit? I suppose it would probably replace the GRIII, but the more I though about it, the clearer it became that I didn't need it, so I canceled my orders.

I would love the built-in flash, optical viewfinder, and 35mm FoV, but until I start actually making photos again, I don't need a new digital camera.

I don’t like the way Glass shows images in a desktop browser when the browser window is wider than around 1,000 pixels. I prefer the layout in narrower windows, but I never have mine that narrow. This means whenever I’m browsing Glass, I have to shrink the window.

The Arc browser has “Boosts” that let me easily adjust the CSS of any website, so I created one for Arc. This is it:

.lg\:flex-row{flex-direction: column !important;}

Here’s the result:

This will do for now. I’ve written the Glass folks asking if there is a reason for the side-by-side default layout because I’d like to never see that layout.

This morning it took over 30 minutes to copy a 70MB file from my MBP to the Synology over WiFi. The wait resurfaced my thoughts about having an always-on computer on my desk with some fast, attached storage.

I just ordered an M2 Mac Mini (Pro) with 16GB RAM and a 512GB internal drive.

Since 2021, I’ve had an over-spec’ed MacBook Pro (M1 Max) with a 2TB internal drive and 32GB of RAM. I don’t do much that requires all of that oomph, but I figured it was nice to have anyway. With the Mini, I went with the Pro version mostly for the additional ports. A smaller, 512GB internal drive should be fine, since I’ll have a number of fast SSDs always attached. I’m not worried about not having enough room for my stuff. The thing I’m most worried about is “only” 16GB RAM. I’ve had 32GB for so long that I don’t remember what it was like working with less. I’m almost certain that 16GB will be plenty for my purposes, but it still makes me a little twitchy. Plus, $1,299 still feels like relatively cheap compared to the $3k+ I spent on the MBP.

The hardest part of all this will be that, since I’m keeping the MBP for mobile use, I’ll now have two Macs again. Consolidating to one machine a few years ago was such a breath of fresh air. I no longer had to worry about keeping configs updates or which folders needed syncing, etc. Now it’s all back.

Keeping things synced means worrying about which apps I use, or don’t, and if they sync by default or need to be set up to do so. My org files are all in ~/Documents/org now so iCloud should take care of that^[1]. However, my emacs configuration is in ~/.config/emacs. I’ll once again have to manage my .config files, which has always kind of sucked. I really don’t want to symlink everything and I really don’t want to go back to finding some “clever” method of managing them. The answer, I think, is to use fewer things that need configuring. Good luck with that, Jack.

The next step will be to figure out my storage options. I’m thinking I’d like some sort of Thunderbolt RAID thing with some SSDs inside. And also my existing backup drives. Oh, and Time Machine, for good measure. If you’ve found something you like, I’d love to hear about it.

Hirayama seems utterly content with his simple life as a cleaner of toilets in Tokyo. Outside of his very structured everyday routine he enjoys his passion for music and for books. And he loves trees and takes photos of them. A series of unexpected encounters gradually reveal more of his past.

Perfect Days was just beautiful, and exactly what I needed. I loved every simple, slow-moving moment of it.

"Nuke & Pave" typically means wiping a computer's hard drive and installing everything from scratch. That's not what I'm doing (although it is tempting). What I'm doing is more like a precision strike. (I'll stop using military analogies, now. They're gross.)

Things get out of hand, and when they do I need a reset. Right now my blogs, email, file-storage, and note-taking are all very much out of hand. I'm writing this post as a way to sort it out in my head.

Blogging. I've been posting to four websites for a couple of weeks. It starts innocently enough, but then I get to a space where I want to write something but don't because I can't decide where it should go. I'm dropping everything but the wiki and this blog at baty.net.

Email. OMG Mutt is so great. So is Notmuch and Mu4e and even Apple Mail. I've been switching between them for a while now and it's done nothing but mess with my muscle memory and cause weird sync issues. I don't get that much email, so I'll be dropping back to Apple Mail in the near term. I'll still be running mbsync daily just to have a nice local Mailbox copy of all my mail.

Note taking. Good grief, it's been a rough couple of weeks on the note-taking front. Face it, I'm an Emacs guy. I've been an Emacs guy for more than a decade. And yet, I found myself spending many hours installing and configuring NeoVim and LazyVim for some reason. It's not completely crazy, as I still use Vim as my $EDITOR, but that's only for popping in and out of simple text or config files. I need to stay in Emacs. I'll keep the Vim stuff around because I've done all that work, but there's no way I'm going to live there. And Obsidian can fuck right off.

I got mad recently because I wanted to jot something down and didn't know where to put it. I launched Obsidian because Daily Notes work great there. Stop it! To help avoid that urge, I've added some quality of life improvements to my Org-mode/Denote setup. Let's hope that works.

I don't know where this leaves Tinderbox. Still noodling on that one.

Otherwise, my notes go in Denote files using Emacs or in TiddlyWiki.

File Storage. This one I haven't dealt with yet. I have files scattered everywhere and it's crazy-making. There's the Synology with my archives, but there's also the Mac Mini with what was going to be my Archives until I chickened out. Now there's some files in both places and some in one or the other. Not cool. I'm currently leaning toward punting on the Mini and putting it all back on the Synology.

There are a few more areas that need a reset. I'll get to them later.

In a post from 2013, Digital Recordkeeping, I summed up my tools like this:

Tinderbox is my notebook. Evernote is my junk drawer. DEVONthink is my filing cabinet.

I was onto something back then. It has occurred to me is that I no longer have a dedicated tool to use as a Junk Drawer. Why is that?

I guess what happened is that DEVONthink became both my file cabinet and junk drawer. I'm now thinking this was a mistake. DEVONthink is for carefully filing things I know I want to keep. It's for organized, longer-term storage. Things like tax documents, journal scans, etc. It shouldn't be used as my save-as-pdf-for-(maybe)-reading-later system. That way leads to chaos, which is what I have now.

What does a junk drawer app need? I'd say fast, easy capture along with good search are the most important. It should be lightweight and allow for different ways of getting stuff into it, based on context.

Some might say that a simple set of folders in the filesystem should work. They're right, it could, but that's still too much friction. I want to grab web pages and images and snippets of text from everywhere and just pour them into something. I'm probably not going to organize anything, so a way to see what's new and a search would be great.

At some point I standardized on DEVONthink rather than the simpler EagleFiler for my archives. EagleFiler is great, though, and I'm thinking it would make a terrific junk drawer. I can capture things using its built-in F1 universal capture shortcut. There's also a dedicated inbox folder that gets sucked into it automatically. And EagleFiler is really just a wrapper around a set of folders, so it's easy to back out.

I miss having a dedicated junk drawer. I think EagleFiler is worth a try.

I’ve been tinkering with keeping offline copies of websites (mostly mine), and have always used either wget or httrack. I wasn’t aware of the WARC format until recently, so I thought I’d try creating a few WARC archives.

wget, as it happens, has WARC support built in via the –warc-file option. I added that to my usual set of switches and put it all in a shell script, like so.

#!/bin/sh
# warc-archive.sh https://example.com warc-file-name

wget \
	--mirror \
	--warc-file=$2 \
	--warc-cdx \
	--page-requisites \
	--html-extension \
	--execute robots=off \
	--directory-prefix=. \
	--wait=1 \
	--random-wait \
	$1

This creates a compressed, self-contained WARC file along with a mirrored set of files comprising the entire site.

Actually reading the WARC files is the tricky part. As far as I can tell the easiest way is using Replay Web.page. Drag the website.warc.gz file into the browser and from there you can search for documents, images, etc. and browse the site completely offline.

It’s a little convoluted and I’m still confused about what goes where, but it seems pretty handy having a single-file, self-contained, offline archive of an entire website.

Derek Sivers posted about how he handles backups and it got me thinking about how I handle backups.

I feel like I'm mostly covered. I use Backblaze on my MacBook Pro for continuous, off-site backups of both the internal SSD and the attached "Media" drive containing my photos, videos, etc. I clone "Media" to a separate external drive once a week. iCloud syncs my ~/Documents and ~/Desktop folders, so that should be covered. The headless Mac Mini is also using Backblaze. The Synology is synced nightly to Backblaze B2 storage.

I don't worry too much about losing stuff. And yet, it sometimes feels a little abstracted. A little too magic. This is where Sivers' piece resonated. I dig rsync and use it for pushing local website changes out to servers all the time, so I decided to also use it mimic parts of Sivers' simple routine.

For remote storage, I spun up a 5TB Hetzner Storage Box. Cost is around $10/month. They really are just dumb servers you can access via (S)FTP or SSH. Then I wrote a couple of tiny shell scripts that wrap rsync.

As an added measure, the script can also rsync to an attached thumb drive, so I'm doing that, too. The scripts back up slightly different things to the storage box vs the thumb drive, depending on "importance".

Now I can type "bkremote" or "bkstick" and I have a complete snapshot of (most of) my stuff, one local and one offsite. The nice part is that I can easily get at all of it. No need to log in to a service or unpack weird backup files. It's an exact mirror. The down side is that it's an exact mirror. No history or versioning. If I delete something locally, then sync, it's gone everywhere. I'm ok with this since it's just an adjunct to my current system. But I'm happy to hear about anything that I might be missing.

For many years I’ve put every new folder full of anything into a new Git repo. I never questioned it, I just did it because that’s what you do.

I’m thinking about no longer doing this. This morning I was daydreaming while waiting for a folder to finish rsyncing to a server and I was mesmerized by page after page of lines like “.git/objects/fb/70e546350cc4106caf1225706b44c85087ed27” scrolling by. I checked a few of my projects and was surprised by how much space all those .git/ directories use.

Do I even need them? I’m thinking no. I can’t remember the last time I went back and looked at earlier commits or needed to diff anything that wasn’t actual code. My ~/org files don’t need precise version control, they just need good backups. Static websites probably need Git because it’s nice for tracking template/CSS changes. Also, I’ll need it if I ever decide to use Github Pages for hosting, etc.

So I’m considering going through my project folders and removing .git/ directories unless I actually need Git.