Thomas Cannon

If you’re using puma-dev (having .test domains and automatic HTTPS is super nice!) but want HTTP2+ support + all the benefits of caddy, I’ve started work on a small script/setup that provides a puma-dev like experience, but using caddy!

And best of all, there are no intermediary steps/tools, it’s just a setup that strings together 2 out-of-the-box programs.

https://github.com/practical-computer/caddy-dev

This is a super quick post, as I’m mired in a few projects. But! The web is a living document, and I want to get these thoughts out there; I can always update this later.

I’ve spent a fair bit of time over the past 3 weeks trying to chase down how to get a good Lighthouse performance score, particularly as I refine Little CRM and hone in my Web Awesome usage. Lighthouse was constantly dropping the score to the ~80s, leading me to:

• Build a whole static site in 11ty that was just the markup + asset compilation
• Figure out a Caddy setup that would give me HTTP/2 with full compression locally
• Burn even more time after that spamming the “analyze page” button.

Ultimately, the lessons I’ve learned are:

• The Lighthouse tab is finicky; particularly the Cumulative Layout Shift. The other stuff it checks is good, though!
• What Lighthouse reports for Simulated Throttling might not actually reflect real-world loading. Switch to DevTools throttling to check those results as well
• For layout shifting and overall loading details, check the “Performance” tab, it’s much more insightful.
• Make sure to throttle CPU & Network to get reasonable information when testing locally
• Building a good Web Awesome bundle (or really, any framework-layer bundle) is really important for first load performance. a Cloak can work, but for slow network connections it can make things unnecessarily worse. General rule for building the Web Awesome bundle: page, callout, card, icon, avatar, divider, tab-group. Then, pick the things you really really want to be available ASAP. But everything else (eg: tooltips) should be autoloaded
• Break up your applicaton bundles from your framework bundles. Combined with preload directives on Font Awesome’s CDN, that improves caching and load performance thanks to HTTP/2.
• Most importantly: with a slow network connection, there is always going to be some rearranging until SSR gets to be stable. So, aim for getting content on the page for users ASAP. They already know they’re on a bad network

These have been a long time in the making; and they’re barely the initial alphas, but the Practical Framework is officially open-sourced.

They’re 3.0 releases because they’ve gone through 2 internal revisions. Next up is:

• Updating existing apps to use 3.0
• Refinements from there
• Tons of documentation, because the framework takes a very different approach than others on the market

If you’re curious:

• The Ruby repo
• The JS repo

After months of research and development, a new guide about how to build maintainable view layers in Rails (and overhaul an existing UI without stopping feature development) is live.

thomascannon.me/guides/th…

This is the largest thing I’ve ever seriously published, clocking in at about 8.2K words, with an open-source snapshot to boot

After a short break, I’m going to extract the framework-level code into libraries, then start the process of migrating Little CRM 💪

I’m currently in the middle of re-tackling the view layer for Practical Computer. Now that there are 2 apps + a framework, the complexity is high enough to actually tackle the problem comprehensively

So, from my view of the Rails View Layer landscape, you have the following paths:

• continue as-is, as we have for 15 years
• get partway there with ViewComponent, having to make concessions & hacks. This is a glaring flaw; table-stakes: viewcomponent.org/known_iss…
• go with Phlex all-in, acting as a clean-room replacement. Basically; nothing aside from Form Builders & explicit helper uses, no ERB files
• roll something on your own, like Garett proposes in some of his musings: garrettdimon.com/journal/p…

Am I correct here? Am I missing something?

A new mini-tool from the Practical Computer labs: an Extism plugin to process Mustachio templates: github.com/practical…

Overall, I’m impressed with how Extism is structured. Being able to use WASM to deal with a thorny issue (running Mustachio’s official C# library in a Ruby codebase) with a small, cross-platform binary that can be embedded into the repo itself, dramatically simplifies the problem at hand.

The goal is to hand this off to Postmark directly, since this gives their entire ecosystem a huge boost in usability.

Hey there!

I wanted to give a quick status update on what Q1 2025 is looking like for Little CRM.

Tl;dr:

• It’s not going anywhere! I’m loving working on it, and seeing the path to 1.0
• I’m taking some time to get past a hurdle in my other craft, making music
• I’m still going to be tackling immediate bug fixes and hard blockers
• If you’ve got things that you’d really need for the first 3 months of 2025, let me know ASAP and we’ll come up with a plan
• If you need to pause your subscription, or hold off, that’s A-OK! We’ll figure out a plan
• I’m telling you this because I value transparency (especially in alpha/early stage products), and because the focus on sustainability and juggling multiple skills that Little CRM was designed for is a core part of the workflow here.

A bit more context/details

The back-half of 2024 has been a slog. The short list includes:

• One dog requiring double ACL surgeries, which was a 16-week intensive recovery period
• Hurricane Helene grinding my city to a halt for over a week
• Breaking my wrist cleaning up from said hurricane
• ✨Life✨

So, not a lot of free time for myself.

If you’re like me and have learned a new skill during adulthood; you’ve likely had to deal with the very long tail of getting into the zone. For example: it takes me approximately 2-3 hours to get out of my own head and “past the suck” when I take time to make music. For comparison, I can get into the zone ~15m, on average, for code.

My working theory is that since I’ve been programming for 18 years, the base skill set is just there.

As you can see, music had to be put on pause. Thankfully, I could continue to chip away at Little CRM, but the deal I made with myself was:

When this is all over, you swing the pendulum the other way. The work’s not gonna stop, so you got prioritize the things that got deprioritized to balance things out, then find the path forward.

That skill gap has bugged me for years, and I know the only way out is through (sheer practice hours clocked). I’ve gotta rip the bandaid and do that.

I want Little CRM to work for you; and things to be done fairly. I’ve seen too many alpha projects wither because the team got burned out, with customers left wondering what’s gonna happen to them. If you desperately need specific features in the next 3 months, let me know. It’s also totally okay to hold off signing up, or pausing your subscription. Little CRM ain’t going anywhere.

Because it’s me, I likely won’t be able to cold-turkey stop work on Little CRM during this “break”. It’s simply too much fun, and I’m eager to keep things going. Obviously, bug fixes and immediate blockers will get fixed ASAP! This is just a short-term, intensive focus change.

If you’ve got any questions, need anything elaborated, or want to chat more; I’m here! And thank you so much for your support of Little CRM. It has been affirming & the shot in the arm I needed.

This great thread came across my desk on Bluesky, so I wanted to muse about the ways I’m trying to avoid the pool of rakes it points out that surround Little CRM.

Because to quote my response at the end:

The eternal question: “How can I build a system to try to help folks who aren’t just plain weird, when the space is filled with just plain weirdos?” 🥴 https://bsky.app/profile/thomascannon.me/post/3ld3ewavtrk24

note fetishism and why it’s greedy & anti-human https://bsky.app/profile/amyhoy.bsky.social/post/3ld3bxzlplt2t

like notes are fine, complex notes, lots of notes all great — but that’s not actually what they were about. they weren’t wanting to learn. they wanted to POSSESS

and not in a pdf goblin way https://bsky.app/profile/amyhoy.bsky.social/post/3ld3bzqirqb2s

Note fetishism is definitely real, and adjacent (IMO) to falling into the Project Management Hyperfocus Hell. Both are focused on the desire to force the universe to bend to your will by documenting everything; as if brute-forcing knowledge into digestible bites will enforce order. In essence, possession as Amy pointed out. 😬

The ways I’m trying to address this in Little CRM are:

• Deliberately keeping notes short, by keeping the restrictions of microposting (240 characters)
• Likewise, threading has proven to be a great way to chain ideas together. It automatically provides conversational cadence, encourages wrapping ideas up, and orients people towards writing (more on this later).
• There are not going to be todo list/project management/activity streak features.

In my design docs, the exact quote is:

We have so many fuckin' todo list options The tool trusts you to be a professional

There will be prompts to follow up with folks, refresh your notes about people/teams, and a way to track actions you’ve taken; but I think the key difference is that they will be opt-in (you have to decide to engage with them). It’s important to refresh your notes, and these prompts will be designed to help you think about what’s next.

They were super into esotericism and pretty sure they thought the process of connecting bits of knowledge would magically uncover knowledge that was previously hidden

https://bsky.app/profile/joshuawood.net/post/3ld3ccsf7hs2y

they’ll do anything if it will reveal things to them, except live and learn like human beings

https://bsky.app/profile/amyhoy.bsky.social/post/3ld3d3frblj23

OOFA DOOFA. I’ve been grimacing & figuring out how to avoid this since I figured out what the flows would generally look like.

I think the biggest ways to avoid this are to avoid too much retrospection (the next quotes cover this!) and by actively encouraging folks to spend minimal time in the app. Two big ways:

• Having the UI use minimal JS, be performant, and optimized for information density & adding notes quickly.
• There will be an egg timer on every part of the app, with the option to dim the page or play an alarm. Let people know they’re not meant to have everything logged in the app, just the stuff that matters most.

I’ve never understood why people care about the graph view. Now I get it - it’s like looking at your porcelain figurines all lined up. https://bsky.app/profile/lukas.blue/post/3ld3fckyonk2b

Yuuuup. I, like all engineers, had dreams of having a graph view eventually, but this thread convinced me it’s not a good idea.

Honestly 60% of the value of note-taking is writing, 40% at best is retrieval https://bsky.app/profile/lukas.blue/post/3ld3fckyonk2b

One of the reasons why I didn’t ship the alpha with micro post searching or any reporting. Writing is rubber-ducking!

I definitely blogged recently about how a lot of my note taking often turned into just data hoarding, it’s easy to fall into that trap 🙈 https://bsky.app/profile/cassidoo.co/post/3ld3dsbhvhk2m

I haven’t talked about it much, but Little CRM is going to have limits in how much data you can have! It’s a design choice. A napkin number: only 1,000 active teams/people total. I might even drop that number lower. Think about it! 1,000 people is a staggering amount of data to consider important. To quote Dale Carnegie: “You unmitigated ass!” You can’t realistically expect to keep that many relationships fresh; you need to pick & choose where your focus will lie.

I don’t know what the limitations are on microposts/events/purchases/redemptions. I might punt on that until there’s better insights into usage. Maybe the focus on microposts/threading/thinking about “what’s next?” Will automatically guide folks away from this thinking.

the note fetishists don’t even do anything with their notes so it’s inverted productivity cult, in a way https://bsky.app/profile/amyhoy.bsky.social/post/3ld3cm7hsuf2i

Again, having prompts that focus on:

• proactively reaching out to folks
• figuring out “what’s next?”
• quickly refreshing a small part of their notes

Will hopefully mitigate this. TBH it’s going to be impossible to truly avoid inverted productivity, but I can try to design to avoid it.

This was a fun thread to dig into! As always, if you’ve got questions, ideas, or want to push back on it; reach out!