Joshtronic

The Dumb Home

2026-04-12T00:00:00Z

For years, I've been chasing the illusion of the so-called "Smart Home". Thermostats and vacuums. Doorbells and cameras. Light bulbs and switches. Garage door openers and bird feeders. Stoves and grills. Sleep mats and of course, bathroom scales.

Fucking bathroom scales.

Oh, and the stove I've never been able to get setup.

Sure they add a level of convenience, but also, how necessary is any of it?

Considering how long humans have been around, without any of these things, best as I can tell none of it's necessary. Even factoring in the convenience of it all, I've been getting more and more paranoid about having these devices on my network.

I'm also getting more paranoid about where the data these devices collect is being sent to. iRobot being bought by a Chinese company doesn't help me sleep at night.

So the bathroom scales. We've owned two, one from Fitbit and another from Withings. The Fitbit one had a lot of issues connecting to Wi-Fi, especially after a battery change. After fighting with it multiple times, to the point of tears, I swapped it for a Withings scale.

The Withings scale was a smidge better, but also had issues connecting to the network. Most of the devices use really cheap Wi-Fi chips that don't play nice with dual-band routers. This has all led to me having yet another cheap device to create a dedicated 2.4 Ghz network in our "smart" home.

The bigger issue with the Withings scale is that we somehow managed to have a mystery 4th user profile. To make matters worse, at some point my wife stopped showing up on the scale. And to compound interest, evidently her phone stopped syncing the correct profile to her phone.

The nail in the coffin is that trying to edit the profile names to shore things up just didn't take. I'd make a change, go back into the app, it would reset. Looking through the online forums, it didn't seem like Withings was going to be much help.

Reaching a level of frustration that had me considering taking a hammer to the damn thing, I decided to reassess some things. Simply put, this wasn't where I wanted to be when Jesus comes back.

My buddy also said something to me the other day that made a world of sense. If the device needs you, then it doesn't need to be smart. Well I can't weigh myself without being materially involved.

So I did the unthinkable, I bought a dumb scale. Not too dumb, it's still electronic. Less features than we're used to, like body fat guesses (they are anecdotal at best) and definitely no Wi-Fi to contend with.

Setup was a breeze. Take it out of the box and put in the batteries. Step on the device, it tells you how much you weigh. It was quick too, the Withings scale always took a bit to settle on your weight, then you're prompted to perform a dance routine to select your profile.

Speaking of profiles, there are none, and that's just fine. Wife steps on, she gets her weight. The kid steps on, she gets her weight. And the data doesn't go anywhere.

Seriously, glorious.

I've been punching my weight into the Apple Health app, but even that seems like it will be short lived. I generally don't have much use for long term weight trends. Nobody else in the house cares either. Hop on the scale, get your weight, call it a day.

This whole exercise has gotten me thinking about what else we can dumb down in the house. At the very least, I'm sick of the 100 subscriptions that effectively means I'm leasing the hardware. I want more control of my data and my wallet.

WeeChat Using Wrong Nick After SASL Auth

2026-04-05T00:00:00Z

Slack is great... for work. Discord is sufficient enough. Sadly, I've never really found any community servers I have enjoyed. It also doesn't help that age verification is now a thing.

I've been of age for longer than I haven't, so that's not really the issue. My issue is with possibly needing to provide sensitive PII to a third-party company. Breaches are more frequent than ever, so I've started to be extra careful with what I share.

I'm also fortunate to have a buddy that reminds me that IRC still exists.

I already have some history with IRC. I frequented Undernet because it had an "elite" sounding name. I've jumped between EFnet and DALnet, and even Freenode before it evidently went to complete shit.

Said buddy lurks on Libera.Chat, safety in numbers, I opted to hang out on there. I've run through the gamut of GUI IRC clients as well. They are fine, but nothing GUI has ever felt as awesome as mIRC. Legitimately considering running it in Wine just for kicks.

In no particular order, here's some notes on my brief foray:

Textual - Great and modern, not open source, also not cross-platform
Halloy - No Debian package, didn't vibe with the first class splits
Quassel - Struggled a ton setting up a connection, probably a "me" issue
HexChat - It was fine, cross-platform but not for macOS

So after a week of setting up ZNC and fighting with all of the GUI clients, I decided to switch gears and try out some TUI options. No stranger to BitchX, because it too sounded "elite", I was sad to see that it's been over a decade since its last update. I mean, it makes sense, IRC isn't dropping features that need to be adopted, but also, I have to imagine it'll stop working eventually.

There are a few modern TUI IRC clients, but I went with WeeChat. Seemed like it was a bit more actively maintained than Irssi. Also, I don't like the "RSS" part of the name of Irssi. Odd, I know, but still my personal preference on the matter.

WeeChat was a treat to setup, thanks to an awesome guide on the Libera.Chat website. I went through the steps and BOOM! I was connected.

But then I noticed that my nick was "josh2" rather than "joshtronic". I checked my settings, I was authenticating correctly with my account name. Much confuse. I dug around in the configuration files a bit, and found that in ~/.config/weechat/irc.conf there were a few peculiar lines, most notably:

nicks = "${username},${username}2,${username}3,${username}4,${username}5"

Seemed that ${username} was sourcing the username on my machine, which is in fact "josh" and not "joshtronic". Going against the wishes of the configuration file, I made some changes directly and eventually figured out how to connect with the correct nick.

Note: You don't want to edit the file directly because it doesn't hot reload and exiting the client will overwrite your changes. You can be methodical though, exit the client, edit the file, start the client.

Then I took a bit more time to figure out the correct way to do it inside of WeeChat. From within WeeChat you can set the default nicks, or set the nicks for the server you're using. I opted to do it for Libera.Chat:

/set irc.server.libera.nicks "nick,nick2,nick3,nick4,nick5"
/save

Next time you jump in and connect, you should have the proper nick set, rather than your system's username.

How to Install dig on Debian

2026-03-29T00:00:00Z

It's been over 3 months since I posted about finally embracing Flatpak. This now short-lived proclamation coincided with me moving back over to Debian. Fortunately, I went into this particular distro hop in the right head space and things have stuck. So much so that I've been purging my dotfiles of Arch Linux references.

What's been really fun about this recent change is realizing just how bare bones Debian Stable is. Sure, if you install a desktop environment you could argue that it's bloated because of that. But in terms of convenience packages that are installed, I've found it to be a bit more minimal than I remember.

One such package, that I would have expected to be installed already is dig. The DNS lookup utility, is something I use quite regularly. So much so, that I already blogged about how to install it on Arch Linux. As a matter of form, I felt like it deserved its own blog post as I was on a machine the other day and couldn't remember the name of the package.

Similar to Arch Linux, the Debian package for dig isn't named dig. As dig is a DNS utility, and BIND (Berkeley Internet Name Domain) is the common DNS server, things are typically bundled together. And since you typically don't have BIND running on a desktop rig, you wouldn't have those lovely packages installed.

To get dig on your Debian system, you need to install the dnsutils package. Said package is a meta package that links to bind9-dnsutils. BIND 9 was a ground-up rewrite of the architecture used in BIND 4 and 8. Considering it's been in play since Y2K, there's probably not a BIND 10 in the near term.

That said, I just stick to the meta package as such: apt install dnsutils. You'll be advised on the meta status with a nice Note, selecting 'bind9-dnsutils' instead of 'dnsutils' and a list of all of the additional dependencies you'll need. Once you confirm and everything is installed, you'll have dig ready to do your DNS dirty work!

Switching from Neovim to Vim (Again)

2026-03-22T00:00:00Z

Okay, so already I did this dance back in 2017. Looking back, my observations were somewhat shortsighted, but felt decent enough at the time.

A year later, I contemplated the future of Vim once Bram Moolenaar dies. In 2023, we sadly found out what happens, with Christian Brabandt taking the lead.

Since Bram's passing, I've bounced between Vim and Neovim regularly. Frequently enough that I've kept the configuration for both in my dotfiles rather than planting a flag on either side.

Death by Configuration

That was until recently, when I decided to overhaul my very modest plugin list. Most of what I use is stuff for Vim that I like, and continue to use within Neovim.

My thought was, if I'm ever going to commit to Neovim, I need to embrace the Neovim ways. For me, that meant leaning into nvim-treesitter, nvim-lspconfig and the like.

I ended up spending an evening removing my old plugins, and building back up around these modern options. I found myself needing to throw more and more configuration into the mix as I debugged each new issue.

As the LOC stacked up, I started to have PTSD about the last time I attempted to do this. The plugins I use just work and this wasn't working out for me at all.

Vim, but make it Lua

After I rolled back my dots, I asked myself an important question:

What problem is Neovim solving for you?

I came up blank. I favor plugins I can use in both since I switch between the two regularly enough. While it's not perfect, Vimscript/VimL has never really been a problem for me.

So for me, Neovim really boils down to the heading of this section. It's Vim, but make it Lua.

Nothing against Lua, but this doesn't solve any problems for me. Because of that, I dumped Neovim and spruced up my .vimrc a small bit and called it a day.

Ring Around the Editor

While I don't want to say I'll never use Neovim again, I'm probably not going to use it again. I did formally drop my Neovim configuration from my dots, but it's all versioned, so I can recover it if need be.

If I do come crawling back, I'll have an answer to my question about what problem it's solving for me. Until then, there's no reason to deviate.

That said, with all of this bouncing around between Vim and Neovim I'm starting to wonder if it's just time to hang it up and give emacs an honest go.

Scandalous, I know.

What Not to Pack

2026-03-15T00:00:00Z

I like to travel intentionally light. I also like to make sure I bring enough of the right gadgets. Economical clothing and toiletry decisions aside, I tend to overpack technical items as well as downtime gadgets.

Sadly, I have to remind myself that I'm not a toddler that needs all his toys. I'm also not the type of person that needs constant stimulation or screen time.

Overpacking tech for me is partly preparing for a worst case scenario, sprinkled in with some convenience. That's topped off with devices to fill all this free time I probably won't end up having.

On a recent trip to Chicago with the wife, I pared down a small bit, but still found myself with a Nintendo Switch 2 that never left its case. I omitted my travel router because it didn't get any use the trip prior. A trip or so before that, I stopped bringing my Kindle.

A lot of my overpacking also has to do with redundancy, which really isn't necessary. Things like cords and power adapters can be replaced. Hell, they can be replaced at a drug store or gas station, as these things are now more ubiquitous than ever.

Another version of redundancy is packing for the entire family. Extra cords, ear buds and such. Everybody in our family is grown, so we're at a point where I give "friendly reminders" rather than packing like an over-prepared father.

The Unpacking List

I tend to make a packing list, but never a list of things I really shouldn't bring with me. Intentionally putting constraints on things to help reduce the load a bit. As I've already reduced quite a bit, I think I'm at a point of refinement and establishing some rules around tech that may or may not be adjusted over the next few trips.

Worth mentioning, I travel with a carry-on and a small crossbody bag as my personal item. The crossbody is already my every day carry (EDC) so not much changes from my day to day. So, in addition to clothes and just enough toiletries to fit in a 3-1-1 bag in my carry-on, I've whittled my list down to:

Laptop computer, mandatory.
Single power adapter that can satisfy every device that needs charging. Pretty easy if you only bring USB-C devices.
Single charging cable per device charging type. Another easy one, as nearly everything is USB-C these days, so just a single cable.

I'm not going through everything in my EDC today, but there is one little optional bit to it. I already tool around town with a small retro game console in my bag. It barely gets used, but whatever. For travel though, I swap out the console for a slightly larger one.

The retro game console could be subbed for an eReader (Kindle) but I usually don't read on trips. Also, if I try to read on a flight, I will fall asleep, like nearly immediately. It's one or the other though. At this point, I wouldn't pack an eReader and a retro game console.

I also toss in my AirPods as I typically don't carry them on me at all times. Maybe I should, we'll see.

Notable omissions are multi-adapter cables, wired ear buds and airplane adapters. Only the single recreational device, no Switch or Steam Deck stowed in the carry-on. Power banks are great, but phone battery life is great these days and outlets are everywhere.

The only gotcha is that retro game consoles, while USB-C, usually need to be charged with a lowly 5V adapter. In my experience, I can get through both legs of a round trip flight on one charge, as my play time is minimal.

I've Been Sleeping on YouTube Music

2026-03-08T00:00:00Z

Every time I try to get further away from Google, they find a way to dig its claws in. This time it's with YouTube Music, a service I've always been reluctant to give much of a chance.

I've been a bit of a lost soul the last few years, and I decided I was going to drop Spotify. This was some time in 2024, and I dedicated the entire year of 2025 to Apple Music.

My year devoid of discovery

Apple Music, by and large, is pretty great. I didn't feel like I was missing out much in terms of selection. Where I felt the burden though, was with music discovery. An entire year went by and I felt like I hadn't discovered any new music.

Music discovery was what I loved the most about Spotify. But also, the rumblings that they were seeding lists with AI slop made it hard to stick with them. Then there was the matter of Spotify possibly engaging in payola (pay for play) which made it easier for me to drop them.

Doesn't help that I have personal beef with them. I had an album pulled from the service for fake streams. Streams that were all coming from a playlist for some "pay for plays" service that I absolutely didn't sign up for.

My convictions haven't been strong enough to flat out abandon the platform.

There and back again

That all said, the year of no discovery with Apple Music came to an end. I moved back to Spotify because it's what I knew. Having field stripped my Spotify account before leaving, it was like I was starting fresh. I thought this was going to be a good thing.

Turns out, coming back and starting from next to nothing was pretty abysmal. I expected to run into so many new artists as I fed the algorithm my likes and dislikes. Turns out they kept just serving up the same bands, nearly all of which were larger / major-label acts.

Here comes a new challenger

Then I had somebody mention YouTube Music to me. Wasn't the first time it's come up, but I liked what they said about the music discovery aspect of things. Turns out the other person I knew that used it was quite the audiophile with eclectic tastes. I probably would have listened to him back then.

Fast forward a couple months into my foray into YouTube Music. I'm still getting used to things, and it's not all rainbows and unicorns, but it's pretty damn close.

Music discovery, my top priority, has been fantastic. I think mostly due to Google's algorithmic prowess. A lot of the new artists I'm finding are on the smaller side, so anecdotally, it doesn't seem like Google is prioritizing major labels or anything like that.

This is also the small caveat I've had with the platform. I do occasionally run into major-label artists' songs posted by random usernames or artist names that aren't the original artist. Unsure if that's just the same flea market magic as the Android app store, or just legit piracy or redistribution of copyrighted works on the platform.

Back to the positives, no complaints related to the breadth of artists available. I've yet to punch in some obscure artist and come up empty-handed. Also, since I uploaded my CD collection to YouTube years ago, I have all of my extremely underground tunes available too.

A square peg that actually fits

One of my favorite features is being able to make a playlist from a prompt. I hate every app having "AI" bolted on, but this feature makes sense. You can toss in something rambling like "Looking for deep cuts, reminiscent of a rainy day in some city, featuring bands like this one that I already like" and you'll get a pretty solid playlist from that.

Most of the time the playlists are about an hour in length. I was able to coax a 14 hour playlist out of it with a pretty vague prompt though.

As I've been singing the praises of YouTube Music to my peer group, the matter of "yet another subscription" came up. It is true, you may need another subscription. I hate ads, and my family loves YouTube, so we've had a YouTube Premium plan for a while now.

YouTube Music comes bundled with the YouTube Premium plan. This was also the case during my year of Apple Music, as we already have an Apple One plan for Apple TV and Apple Arcade.

Turns out Spotify is the "extra" subscription, as I can't seem to get the wife or kid off of it. Regardless of their reluctance, I've been enjoying YouTube Music and all of the new music I've been discovering.

Whitelisting AWS NAT Gateway IP Addresses

2026-03-01T00:00:00Z

Yet again, I was deceived by the Amazon Web Services (AWS) console. Now it doesn't help that my notes never seem to marry up with the current state of the UI. This time the issue was with a small oversight on my part when looking at the NAT gateway list.

The task I set out to accomplish was whitelisting the IP address from our NAT gateway over in another service. I went to the NAT gateway list, grabbed the Primary public IPv4 address and punched it in.

After waiting impatiently for the third-party service to do whatever magic they needed to, things were finally working. Easy peasy. That is, until the next deploy. With a new batch of changes in QA to test, I was hit with new errors regarding a different IP address than the one I had added to the whitelist.

But, the NAT gateway has a static IP address, right? RIGHT? I was able to confirm that, but I was still blocked. After some head scratching and soul searching about whether or not I should hang up computers, it dawned on me.

Herp. Each subnet has its own IP address.

Derp. Each subnet's IP address would need to be whitelisted.

Clicking the NAT Gateway ID from the list takes you to a page that shows all of the associated IP addresses. The word "primary" didn't quite click for me, clearly.

With both IPs loaded into the other system, and another round of waiting impatiently, things all started to work again.

Of course, you don't need to worry about a second IP if you don't have a second subnet attached to the NAT gateway. Amazon's little VPC wizard creates 2 subnets by default, so most folks are probably in the same boat.

I'm just happy (and fortunate) this didn't turn into hours of slamming my head against the keyboard.

Is that AI?

2026-02-22T00:00:00Z

overly long exhale...

I hate right now.

It's not even because I'm going to officially be mid-40s tomorrow. It's because I feel like life in tech is a perverted version of Groundhog Day. But instead of living out the same scenario over and over, it's being stuck in purgatory being asked one or both of the following damn near every day:

Can AI do that?
Is that AI?

Just typing it out has my eye twitching.

I wasn't sure if this post should be split into two, as I think I could probably write a tome on each one separately. I've touched on the first a few times now, as I think the bean counters love the hammer that is "AI" because it justifies a sick arousal some of them have about reductions in force.

Look at my spreadsheet, the numbers look good. The skeleton crew that's left should feel lucky to have jobs. God, I'm glad I didn't get into finance.

The latter question is one that gets posed in response to any and every little creative thing I approach now. Not by everybody, mostly by people that think "AI" is some sort of actual "intelligence".

I've spoken openly about how I use generative AI, agentic coding, and other "LLM-based fuckery" on this blog and beyond. When I use genAI for content, I try to be open about it.

Friendly reminder: I, a human being, write the content for this blog. Any em dashes were part of the conversion from WordPress to Eleventy, nothing more.

My peers in tech are a mixed lot, but most don't default to "you didn't do this, a robot did" mode. That's what pisses me off the most. The question isn't "did you use a robot for this", it's "I don't think you did this yourself".

I like to do creative stuff like making music. I've dabbled with just about every form of art thanks for being an IB dropout that lacked elective credits senior year. I've been harassed by the cops on multiple occasions while trying to capture a moment on film. I once manipulated a feedback loop at a fucking art gallery opening.

Yeah, I do stuff.

That's why this is such a gut punch every time I'm asked. If nothing else, it's a question of my integrity as much as my own insecurities about skill. Doing stuff and putting it out into the world is hard, so it just sucks the joy out of sharing stuff when you're immediately questioned.

The point of this rant? To share a new project, of course ;)

My wife procured a massive sack of dried parsley, and I decided to do a thing. Said thing is a website showcasing my culinary adventures, with and without parsley.

It's called Parsley Makes it Better.

It's dumb, and I love it so so much. What I don't love is being asked if the food photos are AI slop. They aren't. It's real, imperfect, food.

The boring work of setting up the site? Mostly the product of Claude Code.

The dumb little Drake emojis, that's all me. The words on the about page, also me. The laziness in coming up with a logo and utilizing the "herb" emoji. You guessed right! It was me.

This response from people is important to take note of. We, as humans, need to be creating more, and putting ourselves out there.

You could argue that this just trains the robots. Maybe so, but there's already a ton of content already available. New content isn't making these things "smarter", it's how they are being improved internally.

If we just decide to stop self-publishing, the models will continue to improve. What doesn't improve is having the availability of alternatives to the AI trash sites that are clogging up the search engines right now.

Ignoring Files with Eleventy

2026-02-15T00:00:00Z

I love Eleventy, but I think I made a mistake in how I structure my sites. I don't consider blogs and other static sites the same way I think about coding projects. I don't toss everything inside of a src/ directory so my site content lives alongside files like README.md and I never thought much of it.

That was until the other day when I noticed one of my sites, that I openly utilize generative AI on, had my CLAUDE.md published for the world to see. The file itself lived at /CLAUDE/index.html on the open Internet, and I'm pretty sure nobody even noticed.

Side note, I was actually thinking of publishing this file for anybody that maybe wanted to run a similar site to leverage.

Anyway, not a big deal, no secrets leaked or anything like that. In fact, because I thought I had discovered some earth shattering thing, I went on a side quest, scanning over 1,000 sites advertising that they ran on Eleventy. I just knew I was going to expose a bunch of sites written by robots. I was going to be the John Connor of my generation.

Turns out I didn't find a single one. Maybe a bug in my scraper. I lost interest at some point, no big deal.

I ended up checking a few of my other sites, and definitely found I had a lot of README.md files out there, which I wanted to clean up next week. Also seems I have solved this issue on a few sites already. To take it a step further, I seem to have solved it a couple of different ways.

Eleventy Ignore File

Both methods are pretty easy, but I'd argue this is easiest. The file naming and the structure of it is very familiar.

Add a file named .eleventyignore in the root of your project, and add the files that you'd like to ignore, like README.md or TODO.md. When Eleventy builds, it'll ignore said files.

Eleventy Config

Also easy, and maybe preferred if you want everything in one place. Open up eleventy.config.js and inside of your default function add a line or lines like this:

export default function (eleventyConfig) {
  eleventyConfig.ignores.add('README.md');
  eleventyConfig.ignores.add('TODO.md');
  // Other stuff...
}

Same deal, when Eleventy builds, it'll ignore the files you listed out.

Ideal Fix

The better fix here would be to decouple my site content from the root of the project. Doing so will eliminate any need to ignore files. This reduces my cognitive load, and further simplifies my dumb little static site adventures.

But that's for another day.

Finding a New Hairstylist

2026-02-08T00:00:00Z

I got some unfortunate news recently. My hairstylist of the last half decade is moving away. Not only is she moving, but she's moving soon. Like, "this is your last haircut with me" soon.

Honestly, this isn't too shocking, as she's always had a mercurial streak. Her globetrotting has always been pretty spontaneous. Closing her shop and moving to renting a chair somewhere was also quite knee-jerk.

I'm on a fairly rigid schedule with my hair. Every four weeks, not some "monthly" cadence. Accounting hates me on those "two haircut months" as it can throw off my budget column when I'm being a bit too frivolous.

Even though I'm supposed to be getting an intro to another stylist, I'm not holding my breath. I know the game, when you get dumped, you can't expect to be top of the other person's mind or demand follow through. Nothing personal either, I'm just prioritizing myself.

So now that I'm displaced, I'm on the prowl for a new hairstylist. At the top of the year, I decided I'm going to grow my hair out, and I'm being quite particular about avoiding a mullet as I enter my Ellen DeGeneres era of hair.

While I think I've already made up my mind on who I'll be pursuing for this next chapter, I thought it would be good to document my thought process a bit. I say this, but I'm actually down to two people, that are at different phases of their career. Funny enough, because of this, they both carry drastically different social footprints.

I've mentioned frequency and mullet aversion, but there's a bit more nuance than that. I'm not chasing any particular hairstyle at the moment, so I'm looking for someone that can manage a grow out. I look at this like an omakase experience, trusting the person to not do me dirty.

This is hard because I'm starting over. I was okay trusting my old stylist, because I had a rapport with her. With somebody new though, I'm taking a gamble regardless. This is where I think experience is going to count.

A person fresh out of school may be technically proficient, but I expect them to need a script. If I rolled in wanting the X style, they would probably shine, but that's not what I'm looking for. The hard part is figuring this all out before you're actually in the chair.

So now the matter of vetting ahead of time. I fucking despise social media, but being able to look up and see what they've done is great. Except when you are stalking someone with little to no social presence, and one with an enthusiastic presence.

I love the youthful exuberance, but also, a thick social presence is just marketing to help build up a business. Lacking that presence isn't inherently bad. Hell, my footprint's purposefully small. They just aren't out shaking the tin cup as much to drum up business.

Reddit comes in clutch here, as you can find people talking about a hairstylist, rather than trusting the stylist directly posting their best work as a form of marketing.

The other dilemma with browsing socials is, as I already mentioned, is that they are posting their best work. Social media is a highly curated experience that oftentimes purposefully omits the rough edges. I've had some bad cuts, we all have. Oftentimes, the worst cuts are by the person that's given me the best cuts. It's something that you can't figure out until the rubber meets the road.

Something else I'm weighing out is what it means to go to someone not only at the beginning of their career, but also being extremely young. Like, less than half my age, young.

This isn't ageism, it's acknowledging that in my mid-40s, a youthful trendy cut may not land. Hair's forgiving, but I don't want to be running around looking like I'm trying too hard.

That's also not to say that a young person can only do younger cuts, but it's something I'm thinking about. Most of my own experience has been with hairstylists that are younger, but still closer to my age. This is also where a strong social media presence can sell the skillset.

Worth mentioning, from my experience men generally go to barbers and not hairstylists. By proxy, most social media content is going to be of women. And to take it a step further, it seems like most people really only like to showcase their color skills more than their cut / styling acumen.

I get it, the money's in color and processing. This just happens to make it a bit harder to vet on social media alone, unless those are the services you're after.

Maybe something I should have mentioned earlier in this post, but I've been focusing my search on people that are running their own spots. Nothing against folks that opt to work at a bigger salon, but I've always had great experiences with upstarts.

All of this being said, I have a week or so before I reach out to anybody. I think my mind's made up, but documenting where my heads at has been good. Or it's been bad, as I didn't actually get me over the line.

You could just learn to code

2026-02-01T00:00:00Z

The generative AI wave has made it very apparent to me that the world is absolutely out to get software engineers. I don't remember feeling this way over low code and no code solutions when they emerged to take my job.

Right now though, I feel like a witch and everybody is carrying a match.

The thing is, every potential extinction event I've been through has had a commonality. The option to learn how to code has always been there. It's arguably the best option as the world continues to throw code at most problems.

Learning to code gives you the correct language and experiences to work with tools that have convinced the bean counters that cutting a team size exponentially can somehow still yield the same output.

Sure a model can generate a bunch of code, but if you yourself have no experience coding, you have no idea how good or bad it is.

A large swath of people are seeing the magic of a pile of code minutiae being generated as the end of developers. In reality, they lack any sort of reference point to be able to know if what they saw was a work or not.

I may feel like I have a target on my back, but I know I'm not a mark.

So why do so few take these opportunities as evidence that maybe learning to code could be beneficial? Same reason a pill to help you get skinny seemingly put the body positivity movement on mute.

The easy path is the preferred path by most, even when they don't take the easy path for other things. Unsure how people can train to and run a marathon, but then think learning to code is some sort of unattainable thing.

Oh that's right, because the perception that an LLM just generated a barely functional app in 14 seconds. All while making it seem complex because it told you how many tokens it had to consume to do it.

I love the tools that we have today, but I wish more people would take the time to learn to code. If nothing else, use the tools to help you learn how to code, rather than blindly accepting generated content that you legitimately wouldn't know is a hot mess or not.

I'm not Joshing you

2026-01-25T00:00:00Z

I've long been fascinated with how many people there are out there with my name. Not my first name, but my full name. It's not even that there are too many to count. It's that while most don't physically look like me, on paper many of them look exactly like me.

The common theme tends to be "makes music, works in tech". Coincidentally, this is one of the common tropes where I live too.

Inspired by the gathering of the Kyles in Kyle, Texas, I've always wondered if one of the Josh Shermans out there would do something similar. Even before the Kyles did their thing, I've long been collecting my name twins on LinkedIn. Even though I've yet to meet one of my name twins, it seems like we're all decent enough people to accept friend requests from strangers.

Not having enough motivation or confidence to start a gathering of the Joshes, I decided the next best thing would be a weekend project. Given that we're in the middle of the great snowstorm of 2026, this weekend seemed as good as any.

Armed with my trusty non-Josh named LLM, I got to work. An hour or so later, I now run the premier web directory for humans named Josh.

Since I like to shoot low, I doubt I'll get many submissions, so I'm probably going to try to aggregate a bunch of Josh-adjacent sites to include in this portal. Next time I have some free time, I want to add in screenshots of the sites, but I also like the text-based old-school vibe.

If nothing else, the Josh that browses the web via the command-line won't feel shortchanged. Lord, I really hope that Josh is out there.

While it was a decent little time waster on this cold Sunday afternoon, there was an interesting takeaway. Of the small number of Josh sites I've already added, only one had an actual meta description.

I know that a lot of this doesn't matter as much anymore, especially now that robot summaries are all the rage. It's interesting to witness an evident extinction event of something that was once considered essential.

This also made me realize that this site doesn't have a meta description, so that'll be on my TODO list for next week.