Not a Spring Onion

I put bugs in software. Master of half-finished projects.
Irony is my default setting.
Hate factories, but love conveyor belts.
Buzzword-Driven Development.

Edible.
Totally not a robot.
Definitely not a mouse.

Pan-Demi-Sexual, _not_ Pandemic-Sexual.

// Your Plastic Pal Who's Fun to Be With
#YesBots

// Cats of all sizes also welcome
#YesKittens #YesCats #YesTigers

Languages: German, English, Lisp
Location: Deep in the woods (Kaiserslautern, Germany)
Special Powers: AuDHD, synesthesia, incoherent in several languages
Pronouns: Any are fine :3
Repo of half-baked ideas: https://codeberg.org/wakame/unformed_ideas

Last Notes

2025-10-05 00:30:24 UTC

- reply

On the other hand: If a piece of software would derive a "proper" world model from its input, it might be interesting to read and I see no ethical problem with it.
(Not an LLM. A program that would learn similar to a human.)

2025-09-11 16:38:42 UTC

- reply

Maybe separate chambers? /j
Like a revolver sniper rifle.

2025-02-17 04:22:55 UTC

- reply

Hiding can take a lot of forms:<li>Tabs</li><li>List-ish things with scrollbars</li><li>Menus</li><li>Modal dialogs</li><li>List boxes and combo boxes</li><li>Toolbars</li><li>The little question marks that show you a tooltip if you hover over them</li>
Aside from saving screen space, this can also save the user time (if done right) and reduce the amount of information the user is bombarded with (also if done right).
Hopefully your users know that your GUI is not just a room, but a whole house with different rooms (including a broom closet where uncommon settings are stored).
To turn a bucket of input and output elements which would cover several computer screens into something that fits on one screen, we have developed a boatload of metaphors (which are often not really metaphors IMHO) that have to be learned and understood by the users.
And no, there are no "natural interfaces".

2025-02-17 03:38:22 UTC

Why GUIs suck: A rant
Current GUIs are hard to create. Current GUI tools suck, and the resulting GUIs also suck most of the time.
IMHO:<li>GUIs should be easy to create and easy to iterate.</li><li>They should be editable by the end user.</li><li>The changes made by the user should be stored in a way that they can easily be backuped and applied to another installation. In short: These changes should be in configuration files.</li><li>The editing process should be graphical. Moving or deleting an element should be done by selecting an element e.g. with the mouse.</li><li>The user should be able to add text areas (and maybe images) directly to the GUI.</li><li>The user should be able to replace GUI elements with other elements that have the same function.</li>
To achieve this, we need better tools. Well, actually a better approach to building GUIs and integrating them with an application.
It has been 57 years since "the mother of all demos", and little has improved compared to then.
#GUI #UX #Programming

2025-02-07 22:26:47 UTC

- reply

A lot of conversations are just rituals to signal others that you are currently not in the mood to stab them.
My attempts at being more efficient by wearing "I am not in the mood to stab you right now" shirts were met with mixed results, since some people started avoiding me while others were suddenly asking more questions...

2025-02-07 19:09:55 UTC

- reply

Driving a car: Weird, assumes "being in control" mindset.
Taking the train: Wholesome, being encased in a body of steel that takes care of you.

2025-02-07 00:11:24 UTC

- reply

The thing is: With this kind of discussion, the question is never "could this be a useful building block in a larger concept?".
More like: "Can't we have virtual trucks on the information superhighway loaded with real goods?"

2025-02-06 23:00:20 UTC

- reply

That's just hype buzzword allergy.
Often the result in sitting in meetings or spending time with friends, then having to explain why using the hype of the day is a bad idea.
"Can't we use the information superhighway to deliver our organically grown vegetables?" - "No."
"Maybe in five years, we can travel the information superhighway instead of flying." - "No."
"I think people need the information highway more than food or shelter." - "No, they definitely need food and shelter more."

2025-01-29 20:20:43 UTC

- reply

Would something like "The Incredible Machine" count?
Generally, sure, if you want some complex thing to happen, you have to define a state that makes this possible.
In case of computers, that can be a list of instructions that tell the computer how to move data around.
If I want to simulate how rivers are created, I could use a sandbox (a real one) and (real) water that flows down a sand hill.
In many cases, you have interactions between things, and if you more or less know the rules that govern these interactions, you can try to create a "program": An initial state that (hopefully) results in a desired state.
In the real world, we have all this pesky physics that makes it hard to create a ball that rolls along a street forever.
In a simulation (or VR thing), this doesn't have to be the case. You can have magic springs and balls that only roll over pink surfaces and shadows that push objects around.

2025-01-23 20:51:40 UTC

- reply

With BrandEmotions®️, you can ensure that your advertisements hit potential customers directly in the amygdala.
Using our technology in combination with A/B testing, you can combine paid volunteers (aka self-employed microcompanies) with AI pattern generation to derive the optimal Advertolution™️ for your business or brand.
(Advertolution, BrandAddiction and BrandMeth are registered trademarks of Torment Nexus LLC.)

2025-01-16 17:38:30 UTC

- reply

"Yeah, we are following a solo.
No idea who they are.
Does the license plate 'V&JS4evah' ring a bell?" :blobcatgiggle:

2025-01-15 21:49:03 UTC

Cyberpunk 2077 is a game where you own multiple vehicles, even when public transport is faster and easier.

2025-01-11 09:48:21 UTC

- reply

> it was like groping around in a giant bin of tiny broken drones for the one that was still intact
Welcome to fedi :blobcatgiggle:

2024-09-07 19:07:06 UTC

- reply

301 Moved Permanently -> fbi.gov

2024-08-19 15:48:17 UTC

- reply

So I can have a repository of "reasonable features" for an application: Spellcheck, Load/Save, connecting to a remote API, etc.
By adding some of these features, I can show a user my current interpretation of what they want, which then allows them to make a decision and add it to the specification ("no cloud stuff please", "saving is good", "I want unicorn icons", etc.)
So basically I can "talk"/"write" at an abstraction level that I care about (in contrast to a level that my code editor understands).
And... from a technology view, my approach is almost "boring". I simply add markers that describe features to a spec. So no "text-to-GPT-to-bad-comprehension" pipeline.
I think something like this has been missing in many computer-related applications/tasks:
Comprehensive domain models that allow communication with the machine on a higher abstraction level.

2024-08-19 15:41:40 UTC

- reply

I am currently working on something similar, but different.
My basic idea is to generate user interfaces based on a written (formal) specification.
(This is not AI, not machine learning, not genetic algorithms.)
When I talk to a human about part of a UI (e.g. when designing something, or even when I provide support over the phone), there is a certain level of abstraction:
We talk about "opening a file", not the concrete steps where I click "File", then "Open...", etc.
So we are clustering/aggregating sequences using a term we both know.
(What actually happens when I click "Open..." is another aggregation, since it involves several steps in the code.)
If I tell a human developer to write me "an application", they will ask a lot of questions (because that is simply too generic, too much effort would be spent adding features that will likely not be relevant).
But with a code generator, that information is enough to write a thing that is very likely not what I want.