#uservt — Public Fediverse posts

Just for fun I created a Shavian QWERTY keyboard map for #FreeBSD.

The hard part turned out to be that I had to work from pictures of layout diagrams, and tiny ones at that, cross-referencing them by eye against a Unicode chart. No-one has a text list of what Shavian alphabet characters go with what keys.

I named it gb-shaw, of course.

#ShavianAlphabet #nosh #uservt #KeyboardLayouts

The #nosh Guide has a whole chapter going into a lot more detail on HID configuration; which can range at the administrator's discretion from explicitly choosing accept-all-HIDs to a strong stance on supernumerary, misbehaving, or downright malicious #HumanInputDevices.

Given that I was one of several people who proposed going this way in the 1990s, it is long past the time that recompiling one's kernel was the way to choose what #HumanInputDevices to trust.

http://jdebp.info/Softwares/nosh/guide/user-virtual-terminal-configuration.html

#uservt

Just so that anyone coming across this #OpenBSD idea has any questions about the #nosh user-space virtual terminals:

In nosh uservts, USB HIDs are opt-in; whereas with most kernel VTs & X11, USB HIDs are opt-out.

The administrator has to explicitly choose (in the ways laid out in user-vt-realizer-configuration(5) which allow various combinations of address, class, and ID matching) to have a #YubiKey realized as a keyboard HID on the #uservt.

https://undeadly.org/cgi?action=article;sid=20250822064253

#FreeBSD #NetBSD #Linux

Right.

I've got Debian and FreeBSD systems with both the system and service manager running things, and a NetBSD system with just the service manager; all of the services under service management; a sprinkling of many of the #djbwares services running, including regular clock synchronization with sntpclock, dnscache, and tinydns private roots; and my test HP USB 104-key keyboard auto-connecting to the #uservt system when plugged in.

I'm running out of release blockers. (-:

#nosh #redo

@cazabon

My code (because of the way that #USB #HumanInputDevices work) has to store button/key press information as an array of boolean flags in two distinct places (1 local and 1 shared), so of course it has array bounds checks.

The Elecom Deft's report descriptor bug had me looking for quite a time for where I'd accidentally put the wrong bounds check on mouse button values.

#nosh #uservt

Here's what I invented:

It's the #ECMA48 FNK control sequence with leading parameter characters for private extensions:

USB keyboard page keys have a leading '?' (modelled after DEC extensions).

USB consumer page keys have a leading '=' (modelled after SCO extensions).

Key numbers are the USB usage codes in those pages. Modifiers are encoded as a sub-parameter, DEC values minus 1, allowing multiple keys per control sequence. Sub-parameters default to 0.

#nosh #uservt #TerminalEmulators

@cazabon

I had a moment sort of like that the other day when I put myself into the input method editor without the Japanese keyboard plugged in, and I couldn't immediately remember without reading my own manual page what the old IMLIB way of exiting an input method editor in ASCII control sequences was. (-:

I had forgotten because normally I just mouse click on the TUI widgets that I get over SSH; but I was locally using the machine's physical keyboard.

#nosh #uservt #InputMethod

@cazabon

Amusingly, it wouldn't be able to do anything at the end of such a countdown. All of the parts of the system apart from the login processes run unprivileged.

#nosh #uservt #TerminalEmulators

There are a good 80 potential Application Command/Launcher keys in the USB HID doco. Plus USB has enormous scope for future expansion, here.

Reasonably, these can only fit as either FNK or DECFNK.

FNK is tempting, because it's both ECMA-48:1991 standard and no-one has ever used it. But it uses SPC as an ECMA-35 intermediate character, and I suspect that that will give bad TUI input parsers the willies. Time for some experiments.

#nosh #uservt #TerminalEmulators

It's 2025. It should not be the case that I'm the first author of a terminal emulator to wonder what control sequences the "multimedia"/"office" keys on my Windows keyboard should send.

But I have just looked around at TeraTerm, XTerm, and several others; and there's nothing. Apparently it doesn't help the GUI emulator writers that "multimedia"/"office" keys are not really a thing when it comes to #X11.

It seems that I have to pioneer something.

#nosh #uservt #TerminalEmulators