December 2021 Status

Early this month, I read Drew's article about productive IRC usage.

I decided to give it a try in the #sxmo channel on libera.chat. I started by asking which of two projects I should prioritize developing. This led to more swipe-typing work.

Swipe-typing

My wvkbd setup got a lot more responsive, with a highlight trail, case-insensitive matching, and a smaller word list. Staceee suggested some changes to the input typing which could fix the known edge-cases, and Proycon has some ideas about language processing. I think if we add ngrams context-awareness, that could improve the quality of the suggestions.

I added some patches to support my commonly-used programs: VTE3, git, micro, and a new workflow pattern for switching workspaces (without using the super-key or the num row). I'm using a much shorter keyboard, to save that precious screen real-estate.

My app store script was merged into the sxmo-userscripts repo. That will be the home of my emoji and unicode scripts too, when I get around to submitting them. There's no convenient way to install them, but that could be good for the future.

In future-work, there's some necessary pgrep/pkill fixes, and an exciting improvement to the proximity-lock (shout-out to sourcehut for teaching me `git send-email`, very helpful for sending patches to the kernel tree).

go projects

I use several projects written in go. This month I patched senpai (an IRC client primarily for use with IRC bouncers) to be more usable on a narrow display. I also started to look into better mouse-support for lf (alternative to ranger).

aerc... I still use, but I think my effort there might be better placed into a notmuch wrapper. For UI-design, acme's plumbing rules come to mind, but as far as I know, there's no touchscreen friendly software with that functionality.

gemini projects

After reading nytpu's article about fragments on gemini, I think I may try to update Grande and Gerbil to support it.

The scope of gemini-browsers is quite large, so there's always something that could be done there.

I think there's some idea about the open-source application lifecycle in here: we create projects and add features and users until it gets to the point where either it needs to be rewritten (e.g. ranger => lf) and the cycle restarts, or it's too large to be replaced (chromium, kernel). Feel free to disagree. Note that suckless tools aren't 'applications' the way a gemini browser is. Maybe acme-plumbing is the way to go: develop a generic platform atop which simple email, mastodon, gemini, and xmpp tools can sit. That's enough philosophizing for now though.

I promptly updated my VPS to alpine 3.15, which came with PHP 8 and some issues for nextcloud. I believe they've been worked out and backported. I guess the lesson is to wait to update (unless you like contributing fixes to alpine :).

This gemini-capsule is still on a debian buster machine, but may add more redundancy if any open SD cards materialize.

life

I'm doing a little bit here and there, which could be sustainable while leaving more time toward open-source work, but still hoping for something more substantial.

This time of year is usually my vacation (where I spend *more* time on Free&OpenSource software), so it's a little strange to have no change in that respect.

Thanks for being a great audience. Living in "Christ's Era", it's always a great time to reflect on his work; I wish you all the best.

-Zach (earboxer)