Katja Lutz 🦋🇨🇭
banner
katywings.bsky.social
Katja Lutz 🦋🇨🇭
@katywings.bsky.social
Anarchist altruist, down-to-earth, independent hobby astronaut 🧑‍🚀 and amazingly lively, passionate web developer, in the stand against the technocrats. Creator of @nitropage.bsky.social, the FOSS visual website editor.

nitropage.org / lufrai.org
Pinned
Introducing 🦜 Loty - The tiny localization library

Tiny bundle size starting at 523 bytes
Type safety with autocompletion in editor
Customizable message formatter
Messages are parsed lazily
Loads missing messages from base locale
Open source and fully tested with 100 % coverage
Framework agnostic
I guess it must be the Linux Chrome build, as Firefox does not have this issue under Linux 😅
February 13, 2026 at 3:42 PM
Reposted by Katja Lutz 🦋🇨🇭
Update: this seems to be a linux specific bug. Not sure yet if it's related to the linux specific Chrome build, KDE, Wayland or some other subsystem 🤔
February 13, 2026 at 3:33 PM
Update: this seems to be a linux specific bug. Not sure yet if it's related to the linux specific Chrome build, KDE, Wayland or some other subsystem 🤔
February 13, 2026 at 3:33 PM
@alexreardon.bsky.social Do you know any good workarounds 🤓?
February 13, 2026 at 2:16 PM
So native drag and drop in the web does not allow to listen for parallel keydown events?

And Chrome for some reason does not properly update ctrlKey and shiftKey states in drag events..

Is nobody using native dnd?

How do you differentiate between copy and move, if the keys are not tracked 🤯🤌?
February 13, 2026 at 2:12 PM
@cursorful.com How would you do this currently (assuming that I use your cursor zoom feature)?

Am I right that visualizing keystrokes (and other overlays for that matter) is only possible in post-production 🙈?
February 13, 2026 at 12:08 PM
Then I rather trust a human implementation though 😅😂
February 13, 2026 at 11:59 AM
Do you know a trustworthy key logger/visualizer extension for Chrome?

I need to show the key strokes in some of my upcoming Chrome screen recordings 🙈.
February 13, 2026 at 10:49 AM
That said: These are just the concerns, coming from that inner doomer side of things 😅.

In parallel I am also incredibly hyped about v2 🙂, not having to worry about async will be a huge relief and the work on the stores and projections is chef's kiss 😋👌!

6/6
February 13, 2026 at 9:12 AM
Just yesterday I wrote an effect in Solid v1 that does result in a write, luckily the write happens after some async operation, so this hopefully still will work in v2, maybe.

I guess the new rules just trigger a general fear: what if I need to do sideeffects v2 wont allow me to do?

5/6
February 13, 2026 at 9:12 AM
But v2 with it's focus on derivation does not like behavior leaving Solids boundaries.

I remember Ryan saying something like ~Effects are not sideeffects~ (@ryansolid.bsky.social Correct me if I am wrong 😅).

v2 summed up: "Async wont be a sideeffect anymore 🎉, but please avoid other sideeffects."
February 13, 2026 at 9:12 AM
~In a perfect world everything is derived. How dare you write applications with sideeffects that not always nicely derive!~ 😂😜

This brings me to "effects": in my head they still essentially are *side*-"effects" - something that by its nature leaves and enters the framework boundaries.

3/6
February 13, 2026 at 9:12 AM
Then there is this general notion to restrict the user in what they are "allowed" to do depending on context, instead of designing the system around what the user "can" do within the language - e.g. no writes under owned scope.
2/6
February 13, 2026 at 9:12 AM
I would be very curious about the details 😅. What are your concerns?

Personally I worry a bit that the focus on async might ruin Solid for fps heavy duty, e.g. games. These usually do not need async, what they need is raw speed. Luckily I don't make games, so I can live with this. 1/6
February 13, 2026 at 9:12 AM
Solid v2s async capabilities look promising 🥰
Some people were wondering how you could build additional pending affordances or do error handling on top of optimistic data. I'm not a designer but I think this updated demo shows the possibility.

Every action randomly fails 33% of the time.

stackblitz.com/edit/github-...
Solid2 Todos Optimistic Data w/ Affordances - StackBlitz
Run Solid2 Experiments created by Ryansolid on StackBlitz
stackblitz.com
February 12, 2026 at 5:05 PM
Atleast it doesn't sink 😆, or does it 😳?

I start to think Adam Wathan and you should combine your efforts. Maybe the combined FOSS income would be enough for a little wooden skiff 😂
February 12, 2026 at 2:43 PM
1 million CSS enjoyers on this planet, not bad 🥳👏!

Waaait one second 🤔.. Now I understand your hidden evil agenda!! You advocate for new CSS features not to make our lives easier, but to gain lots of subscribers and buy a Yacht from ad money of course 😤.
February 12, 2026 at 1:50 PM
The abraception is real in this one 👻
February 12, 2026 at 11:57 AM
Since a couple weeks I slowly stopped posting on Twitter. What a relief to be away from this buggy, manipulative mess.
February 12, 2026 at 9:49 AM
Interesting 😮, I guess the advantage is that you don't have to camelCase all the properties and copy-pasting CSS just works. But don't you loose auto completion and syntax highlighting with the template literals 🤔?

Or does it work like this?
styled.button/* css /*`
color: hotpink;
`
February 12, 2026 at 9:13 AM
Juhuu 🥰, bin sehr gspannt 😁. Wichtigi Info: keep it simple! De Ryan arbeitet aktuell a Solid v2 wo einigi breaking chances wird beinhalte 😅 (insbesondere bi createEffect).
February 12, 2026 at 9:01 AM
Reposted by Katja Lutz 🦋🇨🇭
Sometimes the best things happen by mistake: I started today with the plan to move the dnd wireframe elements into their respective scroll parents. But the current setup actually is way better: You can see the drop target indicator even if it technically is outside the scroll parents boundary 🤯!
February 11, 2026 at 10:06 PM
Intelligence - Wikipedia
en.wikipedia.org
February 12, 2026 at 8:50 AM
Intelligence has so many wide range qualities and lots are still missing in the current wave of LLMs.

In my opinion once LLMs include all of these qualities, we could start describe them as AI. But for the moment LLM is the proper name - unless you try to sell something 😅.
February 12, 2026 at 8:50 AM
Reposted by Katja Lutz 🦋🇨🇭
Prisma ORM v7.4.0 is out!

This release introduces a new caching layer to speed up query execution. We’ve been testing and measuring it internally for a while now, and it’s ready to ship.

If you’re running Prisma ORM in production, upgrade today!

Full details in the changelog:
pris.ly/q9ckZ2x
Release 7.4.0 · prisma/prisma
Today, we are excited to share the 7.4.0 stable release 🎉 🌟 Star this repo for notifications about new releases, bug fixes & features — or follow us on X! Highlights ORM Caching in Prisma Clien...
pris.ly
February 11, 2026 at 5:55 PM