Journal Entries By Tag: #computers

Assorted journal entries with the tag #computers.


My Cyberdeck

TL;DR — My custom "computer" from a future that never was...

👓 6 minutes

Posted: August 14, 2043

After spending the last few months collecting parts, I was finally able to piece together my new cyberdeck over this weekend, and I must say, I’m pretty pleased with the result. Most of the equipment I used is vintage (or, as some might call it, “outdated junk”), but it supports a number of different data formats and interaction modes, making it handy for a variety of uses.

The main unit is a Tec<Net Walkabout T4 portable terminal with an upgraded Sino-Logic 16 processor (replacing the original 12-core version). Additionally, I ripped out the old port interface module and replaced it with a new one from OdaCom that supports USB-6X, SimStims, about 12 different kinds of ISO-chips, TriD, and even HDMI-Classic (so I can plug it in to any of the old displays in my workshop). Unfortunately, the original display on the Walkabout was cracked, and since I wanted it to be portable, I had to replace the screen with a 20-year-old (pre-merger) Samsung Android that I hardwired into the display adapter. I mean, it’s only a Super AMOLED screen (so, only 2D content), but it’ll work for now (maybe I’ll have better luck the next time I go to the E-Cyc center).

The cyberdeck, running a shell.

Software-wise, I decided to stick with what I know, and that was EncomOS. I’ve been using that particular flavor of GNU/Linux since the Meta / Microsoft merger and the Zuckerberg Affair, and since I already had root access to the Walkabout, it was an easy update to make.

As I said, I’m very happy with the end result, but I honestly I don’t know if I’m finished yet. I was going to put a GPL Stealth Module in it, But I may wait until I actually need it (especially since the crypto-cops tend to hassle anyone carrying one anymore). Likewise, I could replace the display with a short-throw holoview, or even plug a set of Thompson Eye-Phones in to the Hub, but I’m comfortable enough in both shell and 2D GUI to get by without VR for most activities (plus, since the optical data cord is hot swappable, I can always plug in the Eye-Phones in when I want the full XR experience).

I’ve embedded some more images below, in case you want to see more. As I said, I’m quite happy with the finished product, and have already started thinking about what to add to the next version.

I’ll keep sharing updates on any future improvements I make to it.

The cyberdeck, booted into self-test mode.

OK, it obviously isn’t 2043 (yet), but the images above are real, and I really did “build” a cyberdeck (several years ago, in fact).

At present, the “brains” of the device is a Samsung Galaxy s23 smartphone, connected via USB-C to a hub. The hub, in turn, is connected to a TeckNet Heavy Duty back-lit keyboard via a USB cable and is physically attached to it via silicon and Sugru. A 2600 mAh power bank that I picked up cheap a few years ago is also glued to the keyboard, and a metal brace is attached to both the keyboard and power bank, giving it some stability, as well as a place for the phone mount to attach (via magnets).

Middle view showing power bank and USB modem.
Left side-view showing the modem’s phone jack.

The hub has 2 USB-3 ports (one of which is dedicated to the keyboard, but that’s OK), a TF card slot, an SD card slot, a USB-C charging port, and an HDMI port. Overall, the device is lighter than a notebook but more tactile than a glass screen, and sits very easily on my lap.

Right side-view showing USB hubs, one with an HDMI out, and the other with an ethernet port.

I had originally intended to attach both the USB hub and phone mount to the keyboard via some kind of tab-and-slot sliding mechanism (not unlike how Joy-Cons attach to the Nintendo Switch), but I couldn’t find the hardware I would need to implement it. Still, if I do another one, I’d like to explore that as an option, making the whole device more modular (being able to swap out different USB hubs for different needs, and maybe alternate mounts, so I could use a tablet instead of my phone).

I built it over the past couple of years, and actually went through several updates along the way (improving the hubs and phone holster).

Early prototype build.
WIP on the 'deck.

I’m sharing it now because I’m entering the Hackaday Cyberdeck contest (my entry). This post is mostly the same info that’s over there.

Of course, it’s not perfect - it’s not as durable as I’d like it to be, and it’s not exactly easy to carry. My hope had been to mount the whole thing to either some kind of metal frame or plate (a la a hiking backpack, but smaller), providing some much need structural support (and stable grips to hold on to), but I could never find what I was looking for. Plus, I built it before my current obsession with mechanical keyboards, so while the keyboard is nice, it doesn’t have quite the desired click.

But, all-in-all, it was a fun project to put together, and it’s come in handy more thana few times (when I was between machines, or waiting on repairs).

How well does it work?

Overall, I think it works well. Although the small screen limits some of its functionality, the relative simplicity of a phone-based system does lend itself to certain tasks, like journaling and shell-based interfaces (like MOSH), two things I like to use it for. A previous iteration of this design was powered by an S9 which even ran a web server (a virtual machine running nginx and nodeJS), and the keyboard was useful for direct access to the shell.

At the end of the day, the phone is a very powerful device in-and-of-itself, and the added functionality that comes fromt he hubs (whether for extra memory, peripherals, or even an external monitor

Plus, because it’s a Samsung phone, plugging it into a monitor activates DEX mode, a Desktop-like EXperience (see what they did there?) with multiple windows, background apps, and touchscreen controls (or support for an external mouse, if that’s your thing).

[picture of the keyboard hooked up to an external monitor, running dex]

Running DEX on the 'deck.

Of course, it still has some practical issues - running the external monitor drains the battery from the phone, even when plugged in.

But the overall experience, as far as I’m concerned, is quite #cromulent.

I even wrote most of this post on it.

Cyberdeck as writing machine.
WIP on this blog entry.

Does it support VR?

It does, or it did, sort of, but not for long.

Given that the phone is the brain of the “device”, any USB-C compatible phone can be plugged into it. The previous brains for the device were a Samsung Galaxy S9 and S10, each of which could plug into a Samsung Gear VR.

Unfortunately, Samsung discontinued it, so it doesn’t work with the s20 (the current brain) or later. I keep hoping that these devices will somehow get “opened up” with later non-standard firmwares and enable something like the failed Project DayDream to live up to it’s full potential.

Conclusion

In the end, I haven’t used it much - it’s too unwieldy to take anywhere, and if I’m honest, I don’t do alot of mobile computing where it would be useful. I had planned to address the first issue by mounting the device on a metal frame, and maybe I will if I ever work on a v2, but for now, it remains sans handle or reinforcing structure.

And so it mostly sits, collecting dust… just a souvenir from a future that never was.

Atari BASIC Colleen (an 8-Bit emulator) running on the cyberdeck.
Termux (a shell emulator) running on the cyberdeck.

Case In Point

TL;DR — A short post about the recent site outage and the new fan I acquired for the server.

👓 2 minutes

Last Sunday evening, quite out of the blue, the web server that hosts this site started making a horrible whining noise. A quick finger-check determined it was the fan making the noise. I tried to clean the unit (and the Pi4 beneath it) with some compressed air, but alas, the damage to the fan had been done.

My Pi4 webserver, heimdallr, in its original glory.
Top view of heimdallr, so you can see both the pins and the fan that would fail.
A slightly-blurry bottom view of heimdallr, from back when it had 4 screws AND 4 little rubber feet.

Thanks to the magical delivery gnomes, a replacement case arrived the very next day (well, technically it was the following night, which was unfortunately about 2 hours too late for Monday Night’s session of Planar Vagabond, but close enough). So, Tuesday morning, I started swapping the old case for the new one, and 15 minutes later, it was back online.

After some looking around, I went with a “ICE Tower Cooler” by GeeekPi. It was relatively inexpensive, easy enough to put together, and the LED in the fan adds a distinctive, #cyberpunk vibe to the machine (particularly compared to the somewhat professional-but-pedestrian original case).

The updated heimdallr, showing off its new LED powers.
Another view of the updated heimdallr and its color-changing fan.
Rear-view of the updated heimdallr, proving it looks just as cool from the back.
Video of the updated heimdallr booting up.

The Wave of the Future

TL;DR — A brief history of "The Wave of the Future".

👓 less than 1 minute

While watching Weird Science for the hundredth time today, I noticed this amazing image hanging on Wyatt’s wall:

Catch the Wave!

I don’t know how I’ve missed it all these years (although I suspect it was because I’d never watched a high-def version before). I’ve long been a fan of the original Great Wave off Kanagawa woodblock print, and have enjoyed the many images that parody and/or were inspired by it, but I’d never seen one with such a great “digital” look to it, and I was curious to find out how it came to be.

Apparently, it was made by a design agency called Grafik in 1981 for a computer trade show. The amazing thing is that, despite its appearance (and intended use), it wasn’t actually computer-generated at all, and was instead created by meticulously coloring each square by hand… with pencils.

“The Wave of the Future”, indeed.

Fortunately, I was able to find a larger version of it, and now, thanks to the power of the internet (and, in particular, reddit user acoolrocket), I’m happy to have this image on my laptop as my wallpaper. Too cool!


On William Gibson and Cyberspace

👓 2 minutes

I’ve been on vacation for the last couple of days, and have used some of the time to finish reading William Gibson’s excellent “Sprawl” series.

I actually read the first book in the series, Neuromancer, some 14 years ago, and always meant to get back to it, but just never did. Then, about 2 years ago, I re-read Neuromancer and dove straight into the second book, Count Zero, before again losing momentum and abandoning the series. While packing for our vacation, I happened across my copy of the third and final book in the series, Mona Lisa Overdrive, and I made the decision to use this vacation as an opportunity to finally finish reading the series, a feat that I accomplished just yesterday.

First, I have to say that I loved the book. You can tell that Gibson’s style got more focused as the series went on, making each book better than the last. Also, the series fits his style well – he has a habit of creating apparently unrelated strands of storytelling, featuring characters that don’t seem to have anything to do with each other, and bringing them together in the climax. In that way, Mona Lisa Overdrive serves as the climax of the series itself, bringing apparently unrelated characters and story elements from the first two books together (along with some new ones) into an explosive ending.

Much of what I like about the series are the background elements, like the way he describes the sprawl and the histories of his characters. But, most of all, I love the idea of cyberspace:

Cyberspace. A consensual hallucination experienced daily by billions of legitimate operators, in every nation, by children being taught mathematical concepts… A graphic representation of data abstracted from the banks of every computer in the human system. Unthinkable complexity. Lines of light ranged in the nonspace of the mind, clusters and constellations of data. Like city lights, receding.

As a computer nerd whose been into networking information and virtual worlds since the days of BBSing (and through into MUDs, the web, and even OpenSimulator, for a little while), the notion of connecting to digital realms directly via ones own mind has always appealed to me. In fact, one of the most depressing things about the books, to me, is that in the nearly 30 years since they were published, very little of that technology has come to pass.


Two Hard Things

👓 less than 1 minute

Came across this little ditty today, via Martin Fowler:

There are only two hard things in Computer Science: cache invalidation and naming things.

– Phil Karlton

Personally, though, I prefer the corollary:

There are only two hard things in Computer Science: cache invalidation, naming things, and off-by-one errors.

Too true.