With the Xtensa Lx7 twin CPU cores in the ESP32-S3 running at a relatively zippy 240 MHz and accompanying PSRAM of up to 16 MB, you might find yourself wondering whether it could run Linux. As [Paulneja] demonstrates with Linux kernel 6.11, the answer is a ‘yes’, though with the usual caveats.

What complicates matters with the ESP32-S3 is that it lacks certain amenities that spoiled OSes like Linux have come to take for granted, such as a Memory Management Unit (MMU). To deal with this, the NOMMU Linux configuration was used, along with a custom fork() implementation. Although the previous 0.7 version sort-of worked, the current 0.8 release is the first that manages to actually boot reliably and has a usable amount of RAM available after boot.

You can see the comparison between the two versions in the header image, with v0.8 having a blistering 3.7 MB available after booting and with overall resource usage and performance having improved massively. Note that only one core is available to Linux, with the other used by the typical FreeRTOS ESP-IDF stack to provide WiFi and Bluetooth.

This was all run on an ESP32-S3 with the N16R8 configuration, meaning 16 MB Flash that’s also used for writable storage and 8 MB of octal PSRAM. As for practical applications, it’s noted by [Paulneja] that this is a research project, though one could imagine this being an embedded Linux project along the lines of a network router running something like BusyBox.

Apparently, people hate typing. As every movie and TV show suggests, the future is talking to computers. There was a time when speech recognition was complex and not very good. But these days, even our lowly phones can do a pretty good job of speech recognition. Of course, one problem is that your phone probably isn’t actually doing the speech recognition. It sends it to the big business of your choice to interpret. I’ve been using Handy, a speech recognition system that works well for me. I’ve also looked at some that didn’t.

After all, it is sometimes nice to dictate to your computer, and it would be even nicer if you could keep your data local. On Windows, oddly enough, there is a well-developed speech feature that, as far as I can tell, almost no one talks about or uses. One video estimates that 99% of users don’t use it. Linux, of course, has many options, but historically, these have been difficult to set up or finicky.

Of course, the good news is that many of the Linux tools are open source and the models are quite good. That means other people have had the freedom to fork the tools and make them easier to use, at least in theory. The licensing of the models themselves may be different, but those will be hard to modify, anyway and they generally work well. The biggest problems on Linux isn’t the technology itself, but the tremendous variety of systems and setups.

Suppose you want to write a speech-to-text program. Will it work on ARM? What desktops will it integrate with? Can it use a GPU? What kind? What about specialized instructions in some CPUs? Then there’s the forced input situation; typing into arbitrary programs once you know what the user said. On X11, it is easy, but Wayland needs different handling.

A Shortcut

I’ve thought about using my phone with KDE Connect, which is an excellent program. It can let you use your phone as a keyboard and mouse for your Linux computer. Unfortunately, it is aimed at character-at-a-time input, and I’ve never found a way to make it work with voice.

Besides, the phone is beaming all the data to “the cloud.” You probably type things you’d rather not broadcast to the ether.

I had looked at Speech Note before, but it is sort of a speech recognition notepad. I didn’t find it seamless, and it didn’t work well on my system anyway. Vocalinux looks nice, but a quick test kept complaining that my Intel extensions were not available. Makes sense, since I have an AMD CPU. Even though the documentation said it should work, I was never able to get it to work.

The Easy Way

Turns out the application that worked readily on my machine was Handy. Keep in mind, Handy is just another tool that uses one of several models out there, along with other open-source tools. You might need to install some tools to deal with your system like xdotool or dotool, but they are probably already installed anyway. That isn’t to minimize the value of Handy. It is — well — Handy. You don’t have to load and configure models, set up a bunch of system-level hooks, or install a bunch of libraries. You install it, and it works.

You can configure it. The best model for you, for example, may depend on your machine and the languages you speak. You can configure the hotkeys and how the app types into your computer. But it does all the work of downloading and configuration.

No Cloud, Unless…

The models do run on your computer and you can make sure it takes advantage of your hardware. However, there is an optional alternate hotkey that takes your speech, processes it to text, and then sends it to your choice of AI engines to clean it up.

Of course, you could be running your own AI engine, but normally you’ll have it sent somewhere else with a prompt. You can tune the prompt or create your own, but the default one starts: “Clean this transcript: 1. Fix spelling, capitalization, and punctuation errors 2. Convert number words to digits (twenty-five → 25, ten percent → 10%, five dollars → $5) 3. Replace spoken punctuation with symbols (period → ., comma → ,, question mark → ?) 4. Remove filler words (um, uh, like as filler)…”

You do need an API key, but there are free options available. For experimenting purposes, I went to OpenRouter, generated a key, and attached it to one of several free models they have. The nice thing is that you can experiment with different models while keeping the same key.

If you search for free in the models box, you will find a few choices including openrouter/free which just picks a free model that isn’t too busy. That can be important because some of the models will introduce long wait times into your transcription.

On the other hand, you can make a new prompt, copy the original one in, delete the part about keeping the language the same, and add instructions to translate the output to French, and that will work, at least most of the time. So there are a lot of possibilities.

Rather than tell you all about it, we’d encourage you to install it and try it or watch the reveiw video below.

Special Mention

Although Handy is my first choice for day-to-day transcription use, there is another open source project that’s worth mentioning. Nerd Dictation is a very lightweight wrapper around the Vosk model. It does take a little bit to set up, and then it provides you with a command line tool that can start and stop dictation. Of course, you can assign those to macro keys. However, there is also a switch that allows you to simply output to stdout. That opens up a lot of possibilities for writing programs or even shell scripts that respond to voice.

To see what’s possible, run nerd-dictation begin --help. This will show you how to output to stdout, set a timeout, and handle other options.

Of course, the obvious project would be a voice typewriter. Many of the tools mentioned here either rely on or can use OpenWhisper and, of course, you can use it too, if you roll your own code.

Between lawsuits from Apple, and Microsoft being Microsoft, Digital Research’s GEM desktop for DOS never really had a chance. It did have another life on Atari home computers, but it’s the DOS version that provided the code for [Tomaz Stih]’s Linux port of the GEM graphical desktop — which isn’t a WM or DE for X or Wayland, for the record. It is entirely it’s own graphical display that will live in the framebuffer of a minimal Linux installation.

[Thomaz] is leveraging DR’s original code — or at least what started as DR’s code before a series of acquisitions and open sourcing — via OpenGEM and FreeGEM. Sample applications include the clock and calendar, but [Thomaz] says the APIs are compatible with Atari ST applications; presumably given the codebase the it will match the DOS version as well.

Much like when it was originally crushed betwixt Macintosh System and Microsoft Windows, we doubt many will be rushing out to use GEM instead of Wayland or XServer on Linux, but there may well be some use cases. If nothing else, it’s got to be lightweight.

If you missed the Digital Research GEM saga, this might get you up to date. If the idea of it running on Linux tickles your funny bone, you might enjoy seeing GEM on an AlphaSmart word processor.

FTP isn’t exactly cutting-edge technology. These days, if you control both ends of a connection, you’re probably using scp, SFTP, rsync, or something even fancier. But FTP refuses to die, especially if you are perusing old public FTP servers or talking to retrocomputers. Every now and then, you still need an FTP client. Naturally, there are plenty of graphical clients. But some of us would rather stay at the command line. You could just type ftp, of course. It works, and if you haven’t used it lately, it is probably better than you remember. However, I’ve long been a fan of NcFTP. While some other FTP clients have caught up, it still has unique features that make FTP a lot more productive.

Not Your Father’s FTP

Before maligning the standard ftp command, though, we should point out that it probably isn’t the FTP client you remember from 30 years ago. For example, on openSUSE Tumbleweed, /usr/bin/ftp is really tnftp, a portable version of NetBSD’s enhanced FTP client. Debian uses it too; the ftp package in both Bookworm and Trixie leads you to tnftp. Since current Raspberry Pi OS is based on Debian Trixie, you’ll encounter tnftp there, too. That’s significant because tnftp has already fixed many of the irritations you might associate with old-fashioned FTP.

You get command-line editing, history, and filename completion, things that are also in ncftp. Both understand passive FTP and IPv6. The tnftp client can also retrieve HTTP, HTTPS, and file: URLs, so commands such as:

ftp https://example.com/something.tar.gz

aren’t necessarily typos, although ncftp lacks this ability. But ncftp does have some killer features.

Remember Me?

One of NcFTP’s nicest creature comforts is bookmarks. Connect to a machine, move to a useful directory, and save it:

ncftp /pub/micros> bookmark oldstuff

Then later you can simply type:

ncftp oldstuff

The bookmark can remember more than just the hostname, making frequently used FTP sites feel much more like named resources than anonymous servers you repeatedly have to navigate.

NcFTP also maintains a cache of remote directory listings. If you’ve ever used FTP over a slow link, you know how annoying it is to ask for the same directory listing over and over. NcFTP can often work from what it already knows instead. Neither feature sounds earth-shattering, but together they make an interactive FTP session considerably more pleasant.

Get All The Things

Another difference becomes obvious when you want an entire directory. NcFTP supports recursive transfers:

get -R foo

or:

put -R foo

That seems obvious if you’re accustomed to modern tools, but traditional FTP is fundamentally organized around transferring individual files. NcFTP does the tedious directory walking for you. It also handles resuming interrupted transfers more naturally, something particularly welcome when the file in question is a multi-gigabyte disk image rather than README.TXT. With tnftp, you have to explicitly ask to resume an interrupted file. NcFTP will detect it and, depending on configuration, either resume or, at least, offer to resume the transfer.

Go Away, I’m Busy

NcFTP also has a clever background-transfer system. Commands such as:

bgget giant-file.iso

Hand a transfer to NcFTP’s spooler rather than tying up your interactive session. There are corresponding facilities for uploads. That’s an interesting distinction from simply detaching a shell command. NcFTP knows that this is a transfer job and maintains a queue of FTP work that can be retried and processed independently.

Shell Games

But perhaps the biggest reason to know about NcFTP is that NcFTP isn’t just one program. The package includes commands such as ncftpget, ncftpput, and ncftpls. These perform FTP operations directly from the Unix shell without starting an interactive FTP command interpreter. For example:

ncftpget ftp.example.com /tmp /pub/widget.bin

or:

ncftpput ftp.example.com /incoming widget.bin

This is much nicer in a script than sending commands to ftp using, for example, a here document and automating login with .netrc. For example:

ftp <<EOF
open ftp.example.com
cd incoming
put widget.bin
quit
EOF

Sure, it works, but any time you send input to an interactive program it is, at best, messy. The ncftpput program expresses what you actually wanted to do in the first place: put this file there. That’s much more Unix-like.

Don’t Do This At Home

None of these conveniences change FTP’s fundamental problem: ordinary FTP is not secure. Usernames, passwords, and data can travel without encryption. If you’re designing a new system and control both ends, you usually have much better choices. But sometimes you don’t control both ends. If FTP is something you run into, ncftp is worth knowing about. Bookmarks, cached directories, recursive and background transfers, and especially the script-friendly companion commands turn an antique protocol into something that feels surprisingly at home on a modern Unix command line.

Of course, just as you can use FUSE to mount an ssh server, you can use ftpfs, to make a remote server look like part of your file system. You never know when FTP is going to crop up.