luis rodil 🇨🇺

fresh masto UI has popovers everywhere! had the world not agreed a while ago that these were a nightmare for accessibility?

The three dropdown popovers in the new posting dialog, each behave differently!? I wonder what the reasoning behind these decisions is.

https://blog.joinmastodon.org/2025/09/introducing-quote-posts/individual-post.mp4

blog post here:
https://blog.joinmastodon.org/2025/09/introducing-quote-posts/

en: You get Elon even in the wrapping for your traditional spring onions roast (calçots).

es: Con el Elon hasta en los calçots oiga.

cat: Et posen a l'Elon fins i tot als calçots ja. Malament.

GNOME has a maddening "developer experience". The theory is that if you learn GObject and survive the mental contortions of doing OOP/IPC in C, you can write anything in GNOME... except there are dozens of ways of interacting with the same API, each of them broken or unmaintained in some unknowable way. GNOME devs realized this, which is why, they created this syntactic sugar language called Vala that compiles to C, which is passable. Except when you want to write for the shell, you can't do that in Vala. If you want to add any functionality to the shell, you must do so in Javascript, using a bizarre runtime called GJS, that uses all kinds of non-standard JS ways of doing things, like includes using URI aliases, class constructors/initializers having a special naming convention used nowhere else `_init()`. And once you manage to do all that, when you are ready to run your Flatpak package, nothing works because Flatpak runs in a container that is not the same environment you wrote your code in. It's like combining a selection the worst of win32 COM objects, the worst of Docker containers, the worst of C#, the worst of vfs, the worst of C and somehow running a desktop on top of it. My deepest respect to the people that make it work on every release.

Codeberg has made it ultraconvenient for people without a credit card to donate in this page. https://donate.codeberg.org

I like how they prioritize the payment forms in this page, giving EU bank transfers more prominence than other forms of payment.

"public money, public code" is a great strapline. Access to code however is not the only requirement for useful software. "public money, public code, usable software" would be a more ambitious and useful objective.

@migratory@jorts.horse @toxi@mastodon.thi.ng yes, indeed "customization" (for lack of a better word) and programmability. In some way I find it interesting that things like vim are still around and thriving, or that text-based UIs are still relevant and reinvented periodically. It does however increase the chasm between the people who are comfortable in a terminal and everybody else.

@toxi@mastodon.thi.ng UX is a cargo cult and monoculture that thrives only by self-emulation, standardization of assumptions, fagocitation of trends and elimination of nuance. It takes traditions in interactive computing & graphic design and replaces them with nothing, a hollow void. Literally nothing, in some cases literal white space. Because as the cargo cult likes to parrot "less is more", because they manage to misunderstand two things in that one motto. The whole of modernism in design and the definition of cognitive load. As smartphones became the way that 7 billion people engage with computing, everything everywhere that has a screen must therefore be treated like a smartphone and that's it, that's the end of the debate as far as UX is concerned.

@toxi@mastodon.thi.ng The field of UX has a distaste for depth, any curiosity-driven research is disdained as minoritarian, or not user-centered enough, or art (!). And any in-depth engagement with technology is disdained as implementation and therefore beneath the expression of a "product vision", which they seem very comfortable in appropriating for themselves. I have taught coding to kids, to electrical engineers and physicists. I have taught medical doctors, to code in python for data analysis. I have helped scientists with HPC workflows and write shell scripts to send jobs to supercomputers. All of them understood that coding a little bit would be part of their jobs and learned easily and with the commitment they dedicate to any other subject. But not UX designers, most UX designers see code as somebody else's job. Or if they learn any code they commit the minimum effort, beause "I will not become good at this, but I want to learn so that I'll be able "to talk to engineers"". When so-called No-Code products where a thing, they all praised No-Code as the next paradigm. Now that LLMs have pretty much killed No-Code approaches LLMs are The Thing. But at the bottom it all lies the same disdain for "implementation", and experimentation.

@toxi@mastodon.thi.ng this thread is relavant to me, so am going in. I agree with every word you say Karsten, I feel a similar frustration. I started my career by building physical interfaces and at that time my obsession was to design for the senses, do away with screens. Now many years later I teach in the field of "UX", a notion I resist but my University employer embraces with glee. Your observations are true to reality, the Overton window of possibilities in so-called UX has been getting narrower and narrower the last ten/fifteen years. Effectively becoming a monoculture. A lot of people that years ago would have done interface design have been Figma-brained. I refuse to use the terms UX and HCI intechangeably because I think that UX is what happened when industry threw a ton of money to "design" and hybridized it with marketing. I have students these days that have been practicing UX for five years professionally and do not know what HCI is. Can't blame them. HCI has also not been very successful at staying relevant.

@goto80 it's gorgeous!

@l03s@post.lurk.org gefeliciteerd!