You might get tired of reading about floppy disks way quicker than I’ll get tired of sharing floppy disk content, I’m almost sure of it.

Since I’ve been working on my Failing Sectors project, a website that’s entire purpose is to be slow, breakdown, and wear out floppy disks and drives, I’ve been building the media and drive back stock collection. I came across these Verbatim Datalife Color 3.5 MF2HD floppy disks.

To me, these things are a work of art, the colors, the transparency, the contrast between that and the black metal shutter, with its “murdered out” style embossed logo.

I bought 2 packs of them, with the “one to stock, and one to rock” strategy, basically meaning I can keep a fresh box, and use the other ones, but they’re so nice, I don’t really want to use them, so I’ll buy another box that will be my usable floppies.

Now I’ve got hundreds and hundreds of floppies, very few are this photogenic.

Now here’s the thing, since I’m a “content creator” taking photos of my computers, there’s a desire to see that there’s actually a floppy disk in the drive. Think about a black floppy drive, with a black floppy disk in it, is it really in there? It’s almost invisible, and the 2nd most popular color? Beige. And that has its own challenges.

Now, data reliability? That’s going to be tested in the future, from the reading and research I’ve done, there was a curve of reliability that peaked in the 90s, and as the medium phased out, the less reliable they became, and these were produced in 2003, for sure a last push, and seemed to be targeted towards students and youngsters. I guess we’ll find out the hard way, that’s kind of the point.

So I’ve been messing around with this floppy project for a while, and it’s been an up and down adventure. I put up my first test project, and it failed to do what I wanted, to read the floppy every page load. And that was one of many show stoppers, I’ve said many times I learn more through failure than success, and let’s just say, I’ve been learning a lot lately.

Forcing a floppy read every time a page loads, for every user? Sounds simple, but that’s just not how the internet is built to work, every single piece of technology between you and a website is built for speed, and a lot of that speed comes from caching. From the server’s RAM, disk, the connection between the website and your computer, your browser, and even your own computer. Every single step had to be told “ignore how you’ve been built, and wait for this machine to spin up a floppy disk, and load the content” and I had to figure out a way around each roadblock, and I learned, computers don’t want to be slow in 2026.

Next, I couldn’t just tell you that it was happening, I had to prove it to you. So I got to work on learning how floppies worked, where they cache, how fast a real read should take, and how to prove the reads were real. Not just to myself, but to the visitor. That’s the whole point of Failing Sectors, which is still sitting on the back burner, but I did build out an MVP proof of concept that works, and it’s provable. You can check it out here floppy.er1c.win and when it loads, I’ll most likely hear it spin up.

Now, if you are anything like me, and a little mischievous, you’ll sit there and hit refresh and destroy the floppy, well that’s the point, that’s why the project is called Failing Sectors. Luckily I pulled up my analytics tonight, and noticed there’s absolutely no one reading this blog anyway, so it feels pretty safe, but in the future I’ll be creating some redundancy, and some security filters that limit single IP addresses, and hopefully have enough audience to burn through some floppies! (but don’t test that yet)

So here’s the project update: The technology works! (Or maybe I should say, I’ve broken it, successfully) I still need to build the server, which is a whole story on its own, and the first server is already going to be replaced before it even launches, but I’ve gotta shoot the video about it first, because that’s all a part of the project.

Since part of this large project also involves read floppies on a floppy <> USB controller, I’ve tested this as well, and built a bunch of cool tools, and understand more that I ever wanted to know about floppy speed, sector reading, and logging!

The journey continues…

Still deep in the homelab adventures. Learning more every week and keeping these machines, some new, some old, some really old, humming in the background has been its own kind of reward (and pressure)

I’ve also been working on the other machines out in the garage. Even the ones already running, some are web servers half-configured with half-built sites and half-built ideas. Others are Proxmox boxes running in the background, waiting for me to spin up some VMs.

I’ve been pulling everything into Obsidian, wiring it up with notes, tasks, and project lists. That’s the part of the hobby that actually feels like work. I always think a hobby’s going to be fun, and then somehow it turns into work. There are lists, deadlines, and all the stuff that doesn’t feel like it belongs in a hobby, but it totally does. Just surprising that that’s where I’m at with these projects.

Lately I’ve been thinking about why I do that, why I can’t stop turning even the fun projects into systems to manage, I keep finding myself trying to push harder on it, make more progress, get things done. If I’m going to spend the time, I want it to count. But then I stop and think, isn’t this supposed to be fun? Why does a hobby have deadlines? Part of me still says, of course it does. Let’s get things done, learn something, use the time and energy while it’s here.

By the time Friday night rolled around, I’d just wrapped up a couple of long weeks at work. I knew some free time was coming and was ready to get back into the fun stuff, the homelab, the projects, all the things I hadn’t touched in a while. I was heading to bed, kind of excited about having a full Saturday to do whatever I wanted. That’s when I started thinking, what would a perfect Saturday actually look like?

The weird thing is, when I thought about it, it didn’t look that different from a workday. I’d still be building something, fixing something, learning something. The lines between work and hobby have pretty much disappeared. A lot of the stuff I’m doing at work started as something I learned while homelabbing, and a lot of the processes and documenting in the homelab use tools or structure I pulled from work. It’s all connected, just different projects, different screens.

That’s probably where this weird kind of hobby pressure starts to creep in. I’ve never really thought about it before, but I’m starting to feel it, that sense that even the fun stuff comes with its own expectations.

The more I think about it, the funnier that sounds, hobby pressure. Hobbies are supposed to be fun, the things you do because you want to, not because you have to. They’re the projects you love doing, the stuff that makes the time disappear.

Even with the stuff I actually enjoy doing, I still catch myself thinking, why do I always want to push it forward? Why does every idea have to turn into something to finish, share, or document? I guess that’s just how I’m wired, I like building things, making progress, seeing stuff move.

Funny enough, when I pictured my ideal Saturday, this post was on that list. So here we are, Saturday night, crossing one more thing off

Pressure relieved for tonight.

I’ve been using Obsidian lately. They call it a “second brain.” Maybe it is. What I know is this: it’s a Markdown editor, which means everything is raw text. No hidden formatting, no bloated files that only open in one program. Just text I can open anywhere, edit anywhere, and keep forever.

Here’s how I use it.

I started with one little web server, an HP EliteDesk 800 G3, the same box you’re reading this on right now. I didn’t spin it up with some big master plan. I just wanted a reason to mess around with Linux again. Putting a website online gave me that excuse. You need a destination before you can enjoy the drive, and for me, the website was it.

From there, projects started stacking up. Backup scripts, little utilities, network tweaks — each one felt like progress, but I had no system. Notes were scattered across random files and half-remembered commands. And it wasn’t just simple stuff. One day I’m looking up a long backup string for the hundredth time, the next I’m piecing together my floppytop script, or trying to remember where I saved the code that spits out my web stats. The projects were fun, but the way I was tracking them was holding me back. That’s when I realized I needed to start documenting for real.

The bench is a mess. I’ve been buried in the P6T build, finally making progress, then of course I got sidetracked watching homelab videos on YouTube. Next thing I know, I’m thinking… maybe I should make my own. Not just a quick update—an actual build video.

That’s when I realized, filming a build adds a whole new layer of planning. So yeah, I just made this project way more complex… and probably more expensive too.

A lot of this stuff only happens once: the first boot, the first POST, that first shot of the old board actually firing up. No re-takes. So I’m just rolling with it: cleaning here, untangling cables there, grabbing shots when I can, and keeping the build moving.

Script is written. Camera setup is halfway there. Shooting starts tomorrow.

This box used to be a serious gaming rig. Now it’s moving on to its second life, the way a lot of old gaming PCs do, running as a nerdy little server in my homelab. The plan is to turn it into the floppy server for Failing Sectors, my retro web project built on old hardware and slow pages just for the fun of it. This ASUS P6T is from one of the last generations that shipped with a floppy controller.

Enough planning. Let’s actually build this thing.

Since I’ve been spending the last few months with all my free time at a command line, pretty much all text mode, I knew I needed to spice things up with some old ASCII art. I actually planned on requesting it from a still-producing/releasing artist, but everyone’s old and everyone’s busy. So I decided to bust out some of that old 80×25 text mode art myself. I’ve run into some old scenesters along the way, and I guess I might “be back.” Who knows. What I do know is I feel like I’m back home.

I’ve been in and out of the text mode “art scene” since I was 14 years old, and while this might not blow your socks off, I’m super happy with it. Most importantly, I’ve got an identity, so I can keep building on my floppy web project.

It’s been great finding out there’s a whole crowd of people doing what I’m doing, stepping away from social media and carving out their own corners of the internet. The indie web is back (no, really, it’s a thing, the #indieweb). I’ve also been diving into the Gemini protocol, a group of enthusiasts pushing the web backwards in the best possible way. No JavaScript, no graphics, just raw text and ASCII.

Ok, I got tired of how ericserv.com looked, so I gave it a quick visual overhaul. Since I have netstats running, the next version will be much geekier. Tonight was just a quick fix because this page was one of the first things I put up when I started building my homelab.

I keep learning, advancing, improving, and upgrading.

My failing sectors floppy project is in full swing, currently building redundancy.

Optiplex GX520 tower
Optiplex 760 DT
Optiplex 755 SFF

It started simple enough. I wanted to host some personal websites because I was tired of paying for WP Engine, and honestly, I just wanted an excuse to dive back into Linux. I wasn’t even aware of the word homelab at the time. I just wanted to run a simple web server.

I got my first real “server,” an HP EliteDesk 800 G3 Mini. Got it up and running. I was now online.

Once the site was up, I was like, alright, this is the production server now. I don’t really want to screw with it too much. But I still wanted to keep learning, messing with stuff, installing things just to see what happens. So I figured I’d spin up a second box. Something I could totally trash if I needed to. Reinstall, wipe it, whatever. That felt like the right move.

That second machine became OptiServ, my sandbox server. And yeah, I’ve got a soft spot for it now. There’s just something about SSH’ing into a machine you named, giving it a purpose, and watching it grow into something real. That’s when it clicked.

I was hooked on homelabbin’.

I’m nowhere near done. Every day I’m learning more, building more, trying things, breaking things, fixing them, figuring it out. I’ve even got some floppy-drive projects in the works. More on that soon.

Dropping a quick pic of where the lab stands today. If you’re a normie, this might look impressive. If you’re one of us, you already know. This is just the beginning.

We haven’t even scratched the surface yet.

I’ll keep you posted. Buckle up.

Fell into a nostalgia hole last week. Here’s what happened:

Sometime after a boring Steam Summer Sale, I asked ChatGPT if there was a way to play old Windows games on my Mac Mini M4 Pro. That one question kicked off a whole rabbit hole. You know how it goes. One minute you’re curious, the next you’re deep in forum threads and file formats you haven’t thought about in twenty years.

That’s kind of what I love about using ChatGPT. It never shrugs or says, “eh, maybe later.” It’s just always ready. Always down to figure stuff out. I’ve gone from random ideas to working setups in a single afternoon, just because it kept saying, “Yeah, we can do that.” This time, that yes led me back to Unreal Tournament.

I had ChatGPT generate me some old-school readme.txt instructions, like the kind you’d find on a sketchy FTP site back in the Unreal Tournament days.

File: UT-README.TXT

This was one of the first games I really played during the “video card” era of PC gaming. Software rendering was dead the moment OpenGL showed up. And with the right setup, Unreal Tournament felt like the future. We weren’t lugging CRTs to LAN parties anymore. We were online. It was 3D shooter warfare over cable modems.

I’ve since learned Quake 3 can run on Mac silicon too. Adding that to the to-do list.

The GE FM/AM Electronic Digital Clock Radio 7-4612B recently made its way around the internet as a meme. The joke was that everyone had the same one—but I didn’t realize how popular it really was until I saw it. I instantly recognized it. Most people were talking about the 90s, the annoying beep, and the fake wood grain. And yeah, I remember all that… but it also unlocked a different memory for me.

To me this clock wasn’t about waking up. It was about staying awake, late at night, exploring the world of AM talk radio.

I don’t remember exactly what made me flip over to the AM band at 14. But if there was a button, a switch, or a dial, I was going to mess with it. That’s just how I was. I’d push it, scan it, maybe even take the thing apart to see how it worked. Until it no longer did.

Let’s just say in 1992, not a lot of kids were hanging out on the AM talk radio band. Maybe that’s what drew me to it. What I found between the static was the Larry King show. He became the voice that kept me up. The format was simple. A guest would be interviewed, and then the phone calls would start coming in. I think that’s what hooked me. It felt interactive. Real people calling in with questions, or sometimes just to be heard. It was like late-night cable access, but on the radio. The audience? Night shift workers, truckers, and other nonconformists listening in the dark. I knew I wasn’t the intended audience. And that made it all the better.

I stayed up many nights listening to Larry King and Art Bell. But in Houston, Sunday nights were something else. They played old-time radio shows. Stuff from the 1940s, before television. Dragnet, Sherlock Holmes, The Whistler. Even back then, it felt like time travel. These were the same broadcasts sent to soldiers on Armed Forces Radio Service, complete with the original commercials. That’s what made it feel real.

krrrrrkkkkkkkkkkkkkk… ssshhhhHHHHHHhh… bweeee-OOO-bweeee-OOO… KKKKSHHHHHHHHHHH-eeeeeeeeeee…

I remember getting excited about that modem connection noise. I knew that sound meant I was about to be online. It felt like stepping into something much bigger. I was connecting with smart people, other geeks and nerds, pushing boundaries, learning, and doing cool stuff. (Ok, maybe it wasnt that cool)

Today, everyone’s connected, and the internet doesn’t feel like an exciting place to escape anymore.

I need to start seeking out those corners of the internet where geeks are having geeky conversations and building and creating things.

I’m going to admit it, this image that chatGPT made, was so good, I’m going to have to come back and write more about this topic, just to make it worth the cycles it used to create it.

Update: Just for fun, I decided to refresh the post by asking Gemini’s Nano Banana 2 to redo the image. It serves as a perfect tech demo of just how far AI image generation has progressed in less than a year.

I haven’t been excited about a full tower PC case launch in maybe 20 years, and it’s funny that this one has me paying attention. Turns out I’m not the only one.

SilverStone is releasing the FLP-02, and it looks just like the PC cases from the late 90s that we all used to hate. Those boring beige boxes with the turbo button. Back then they were everywhere. Now that nothing looks like that anymore, nostalgia has kicked in and made them desirable. So desirable, in fact, SilverStone thinks theres a market for them. 

This new ‘sleeper’ style offers modern GPU support, power supplies, proper cable management, and of course, airflow. While I’m sure the target audience is mostly gamers, I’ve been searching for a full tower to build a server with plenty of space for spinning hard drives. This might be the one.

Back in the day, I used to take those boring beige cases and mod them myself. I’d cut out windows, add lights and extra fans. Eventually those started being sold in stores and we called them “pre-modded” cases. Did that take away some of the fun? Yeah, maybe. But being able to buy a case that looked cool and was designed for modular builds? That changed everything.

Heres a link to a youtube video about it Silverstone FLP-02 launch

I’ve been doing my daily walks, listening to the history of people creating shareware games (see previous post), and the way my brain works, I thought, “Hey, I want to make a game…” So I did. Honestly, today there’s almost no reason you can’t learn to create whatever you want.

Here’s where we are in life right now: I went to ChatGPT, told it what I wanted to build, and it was ready to roll. I didn’t just type “make me a helicopter game” (though you probably could). I started with an idea, mapped out a plan, and started creating the project with AI as my co-pilot. Along the way, I learned a ton about game design, building environments, creating graphics, setting up collision logic, debugging, and working on playability. Most importantly, I gained a whole new respect for the original teams behind games like Choplifter and Chopper Strike — they were doing amazing things with way less tech.

My goal was to create a working game demo, and after about 20 hours, it’s up and running. It’s my first real game under my belt, and now I’ve got a ton of ideas and the confidence to keep building whatever I want. It’s a fun place to be.

Could you do this too? Absolutely. Did my years of web development help? For sure.

Want to play it? eChopper 0.03 is here to play.

I’m going to continue to add a few game play features, and make it feel a little more complete, then I’m moving on to the next challenge.

Later, I’ll write more about how to actually learn while working alongside AI — not just getting answers, but breaking things down, asking why and how at every step.

I’m currently reading (okay, listening to) the audiobook of Shareware Heroes.

First off, I’m just glad someone documented this era of my youth, the days of dial-up bulletin boards, sharing games, pirating software, staying up late, and making sure my computer was always uploading or downloading something. I’ll probably write more about it once I finish.

Update: I originally thought it would just be cool to hear about the old shareware scene during my daily walks. But, man, this book has been way more than I expected. It’s got me so excited and so hyperfocused that I actually started building my own game, I’ll be ready to share that soon!

You can check out the Shareware Heroes site here, it’s also really cool.

So here we are, the first post on ericslog.com.

You’re either early, or you’re deeply invested in the blog and have scrolled all the way down to the beginning. Either way, appreciated.

Let’s start with the why behind ericslog.

I’ve really been enjoying sharing projects and lessons learned over on my oneprojecteveryday site. It’s been a solid outlet, but it’s also got some limitations:

1. It’s only project-focused, which boxes me in when I want to talk about other things.
2. It’s hard to keep up with doing projects and documenting them daily, so updates don’t happen as often as I’d like.
3. It didn’t allow comments or likes, which made it feel a little one-sided.

And then there’s a fourth thing, the word “daily.” Once you talk about a “daily” project online, it starts to carry this pressure. Like you’re letting someone down if it’s not updated. That kind of weight can suck the fun out of it.

I’ve been working on this site daily, I’ll go back to my One Project Everyday and catch up, and break this site launch down, and catch up on all the other things I’ve been working on.

So, enter ericslog.com.

This site’s just me posting ideas whenever I feel like it. Could be music, projects, AI stuff, learning notes, Jeeps, whatever’s rattling around in my head.

I once heard Merlin Mann say your personal blog should be like one of those nerdy basement model train sets. You go downstairs and tinker, not because anyone’s watching, but because you enjoy building your own little world. That idea stuck.

The design here is meant to handle short posts, longer posts, and everything in between. Sometimes it’ll feel like Twitter. Sometimes like a full blog. It’ll stay structured, no matter what I throw at it.

Will I use AI here? For images, yes. For copy, kind of. I’ll always write my own posts. Sometimes I’ll ask a prompt like “help me refine this and check grammar, but keep it in my voice,” but I’ll never say, “write me a blog post about fishing on a Thursday morning.” The world doesn’t need more AI slop.

I’ve got some AI predictions coming soon, by the way.

I like writing, I like sharing, and I like having a place to collect my thoughts and file them away on the internet. There’s no deeper meaning behind “ericslog.” I just needed a short URL, and it ended up sounding a little like Star Trek. Nerdy, but not that nerdy.

Anyway, that’s the intro. Nothing fancy. Just a new little corner of the internet I’m carving out for myself. Let’s see where it goes.

So I needed to test embedding a YouTube video, and I’m posting one of my favorite series: Ronald Finger’s Fiero Revival. It’s a multi-part series, and honestly, I loved it.

Ronald Finger is an amazing creator, but he’s also just a guy filming himself tackling big projects, making mistakes, and learning along the way. I’ve watched him rebuild his garage, restore cars, DIY frame repair and beautifully documenting the all the projects in an amazing package.

I finally learned to love learning, and heres why:

(update: this is an old article from my blog, that I’ve moved over)

It took me a while, but I’ve finally come around to loving learning.

Funny, considering how much information we’re surrounded by these days. But it’s more than just having access to knowledge. For me, it’s about curiosity and the drive to keep growing, personally and professionally.

At 46, I’ve hit this cool moment where personal and professional growth feel like they’re feeding off each other. It’s not about learning a new skill to check a box. It’s like tapping into this endless well of information and knowing it’s on me to make the most of it.