gemini://mntn.xyz/posts/2021-10-11-short-thoughts-on-making-a-modern-pda/

*Edit:* This ended up being not so short. Hopefully someone finds some use from it.

There have been a number of great posts about PDAs lately, from rejuvenating old Palms to building a modern version of a PDA. (There is a roundup of recent posts on ew0k's gemlog.) I am also increasingly interested in this idea. But first I wanted to respond to idiomdrottning's post on PDAs vs books.

Idiomdrottning enumerated a few reasons that PDAs might be interesting to modern users:

> Ultimately, as far as I understand it, the primary purposes of using a PDA instead of modern stuff is:

> nostalgia (or the "nostalgia for what you missed", which I totally get)

> hipster/weirdo cred (having a "unique thing", like "cyberdecks" a few months ago)

1. Unlike a smartphone, it should be possible to easily build or assemble your own PDA from components rather than buying it off a shelf. This can be a good learning experience in itself, it allows you to understand the hardware better, and it gives you greater personal control over the device's final form. It also contributes to repairability, although this depends on the complexity and integration of the components used.

2. A typical PDA (if we're going with the traditional definition) is a low power device, so it should both be more appropriate in a future where energy is more precious, and it should be less capable of distracting you with flashy videos, animations, ads, and such. This also makes it less capable of spying on you, especially if the camera and microphone are omitted.

3. The oldest PDAs didn't even have network capabilities, but instead relied on syncing in a charge cradle. In today's always-on world, I consider this a plus. I can envision a wireless-enabled cradle that performs synchronization of email, contacts, documents, maps, and reading material only when the device is inserted. This would be an excellent way to forcefully limit your own connectivity.

I am in full agreement that books are better than PDAs for many things, including relaxation and deep learning. But the PDA can serve a purpose for those who want a portable distraction-free tool they can use to address practical concerns, alongside carefully selected content for casual reading and reference.

I did a bit of research and discovered a few good options for building a modern PDA that others might find interesting. First I'll list my personal must-have constraints:

The platform should be fairly open and common enough to ensure that core libraries are available

It needs to be small, even tiny by modern phone standards (pocketability is key)

I should be able to run my software directly on the device, without hacks like running Linux in a Kindle chroot

There are tradeoffs with each of these, and I'm willing to bend a little in most areas if needed.

In general, power consumption issues in particular led me to e-ink devices. These have some added benefits, including being pleasing to look at and appearing more natural, as opposed to the glowing screens that so often rule our lives.

First, and perhaps most interesting to me, there's a couple of e-ink screens with built-in ESP microcontrollers, with integrated touchscreens and battery support. There is the M5Paper from M5Stack and the LILYGO T5 from LILYGO. Both are 4.7", fairly small compared to modern smartphones, which in my opinion is a much better size. I have read that the current version of the M5Paper has some power consumption issues, so maybe the LILYGO is a better choice for now.

For a slightly larger option, there is the Inkplate 6, the Inkplate 6 PLUS, and the Inkplate 10, all built from recycled e-ink displays. These don't have built-in touch controls, so you will have to source an appropriately sized resistive touch screen overlay from eBay or elsewhere, as well as a touch controller that could be connected to the I2C or SPI pins.

One big downside of ESP-based devices is that PDA-style software is not really available yet, as far as I know. UI libraries are available, and I found at least one epub reader project, but everything else needs to be built from scratch. A fun project for someone, perhaps?

Another issue with these devices is the fully integrated design. Individual components are probably replaceable with some high-level soldering skills, but especially on the smaller boards the components are packed in tightly. Still, they are affordable options and are very friendly for people who would rather mess around with software than assemble hardware.

This sounds intimidating, but it's not as bad as you'd think. There is a massive array of e-ink screens out there, many with built-in SPI or I2C interfaces and/or pins for Raspberry Pi connections. Waveshare in particular has many options in a variety of sizes, some with grayscale and some with color support. The big downside is that you will need to source your own touch panel for most of these screens, and it can be hard to find exactly the right size (but keep looking--it's probably out there).

The biggest advantage of this approach is that you can use a common Linux-compatible single board computer like a Raspberry Pi Zero. These are cheap and ubiquitous enough that you can stock a few extras, power consumption is reasonable, and software support is stellar. With common interfaces, it is also likely that you can easily upgrade or replace either your board or screen in the future as needed.

The downside of a Pi Zero compared to ESP devices is that Pi Zero power consumption, while better than a smartphone, is still high enough (80mA) to deplete a small battery over a day's time. (Note, the Pi Zero has no sleep mode.) With a small OS image, boot time may be short enough so you can just turn it off when not in use.

You could also build your own device with an ESP, Arduino or compatible, or other microcontroller. Software support will again be an issue, and you may be left to figure out your own path for displaying to the e-ink screen--not my idea of a good time, but some may enjoy the challenge.

Some small e-readers may work as a PDA if Linux can be installed on them, though this often requires jailbreaking. PostmarketOS supports a small but growing number of e-reader devices. I did not look into this option as thoroughly because I don't like using devices that have to be jailbroken to be useful, but it is a valid option and it could give you a working device very quickly. Downsides include lack of repairability and lack of transparency into the internals; on the other hand, these devices may be more durable and lightweight than many homebrew solutions.

Short thoughts on making a modern PDA, and Re: PDA vs books

Potential hardware

ESP e-ink touch devices

Build your own e-ink touch device

Repurpose an e-reader