dev at @asana • big OSS enthusiast • team space 🚀 • pls try @analogdotnow
Poland
Joined December 2023
- Tweets3.8K
- Following417
- Followers194
- Likes20.5K
Wait, Kotlin/Compose Multiplatform people are still reimplementing parts of iOS UI for Skia? 😦
Have we really not learned anything from Flutter? It’s 2026, people, why are we spending time on this stuff. Big companies posting articles like “agents allowed us to go native” and then you see this ⁉️
Isn’t one platform where we have to deal with the fake UI materials and a11y hacks enough? Leave Flutter alone, it’s a weird creature in the Google zoo. I like it, yes, if I have to write cross-platform widgets I’m sure as hell staying away from the Java land. But it will always be worse than native.
The demo is cool, I’m happy someone is trying. But @imadetheseworks’s version is just better😶🌫️
Haze 2.1: Shatterproof Glass
🪟 Glass is now stable
🫧 Merging container surfaces
🎨 Blur and Glass styles that respond to animated state
🌍 Jetpack Compose and Compose Multiplatform
🔗: chrisbanes.me/posts/haze-2.1…
One prompt for my "OSS tools inspector" Grok bot (worked for 2-3 mins) costs 1% of usage.
So if we count actual tasks for bots it will be roughly 40-60 a week. That's on Premium (not Super or Ultra)
Well 🫠
Cool attempt, but I'm wondering - is that the reality we're living in rn? The one where we spend tokens on recreating smth that has been attempted by others before (and gained traction)?
I think the past 3 years have been the rise of open source and its downfall simultaneously.
Before AI subscriptions this was a common pattern:
> hmm, I would love to have this popular <thing> on <platform of choice>
> should I try to build it?
> nah I'll spend like 2 months on it, I have to ship projects at work, spend time with the family, it's too much
> let me find if anyone had the same idea
> oh wow this dude has almost nailed it, still needs some polish, will suggest fixes or file an issue
> everybody wins
> OSS is amazing
Now people are paying $200 for Claude:
> this <thing> is so cool I need it for my next <platform of choice> project
> hey I saw someone posting about it, I can use their thing
> oh, well. it's cool but I really need one missing piece and kinda don't like their approach
> damn I have a lot of tokens left in my sub, I can make Opus build it
> the clanker is done, damn that's nice it even made docs. I gotta publish this
> hey everybody, try my package! worked for me
> the project is dead after 3 months (unless there's monetary incentive / will)
I'm not shaming anybody for forking (@nank1ro 's repo in this case), I think OSS exists to try different approaches and tinker. But don't you get a feeling that people are more & more hesitant to help each other? PRs and issues and hitting someone up in DMs is no longer "fun". It's annoying, slow. We don't like slow. Agents showed us fast. We like fast. And we become more isolated with our agents because of that. "Rebuilt it how I like", "cloned it in 2 hours", "used 1B tokens but now it's in Go"
Or maybe it's just a worldwide surge in adhd? We have dozens of things to focus on, so we think stopping on one of them is a waste of time. And we're spreading thin instead.
"forui" was also built by 2 guys who needed it for their app btw. same logic, I assume
Niche tweet but very on point
I tried Paper, Wonder design and Doop design over the weekend. Here’s my hot take: none of these tools are made for designers.
They are made for AI to design for you. And if you want to design, you do it in Figma and then bring those to these tools and let AI continue.
These apps feel more like rectangle tool with mcp than a design tool. Except rectangle tool, type tool and in some pen tool, that is all you got.
Now, a lot of people swear by these. I honestly feel like those are the first to be replaced fully with Claude or GPT.
I also feel that people like these tools because they give them a false sense of doing so much. You iterate so fast, you produce so fast. But with that much speed quality is what suffers.
That is why on X all we see is the same thing over and over again, same dashboards, same visuals, same styles, same effects. Everyone talk about taste but that is where it ends.
Majority is obsessed with tools, not the design.
Back in late 2024 I found this thing called OpenHands and was so pumped I wanted to make a video about, even made a cover.
I have no idea what that tool is. Or what I wanted to say. “Vibe coding” wasn’t even a phrase then. Feeling old rn
Asked Grok 4.6 Extra High for a Go rewrite of our internal CLI which is a single .mjs file with 0 dependencies, about 2k lines
> Spits out 10k
> dude i thought LOC would go down
> wtf
(i have zero Go knowledge)
If they were serious about the PHP bit, they would leave that famous CVE unpatched
We've been carefully thinking about a new design direction for Cal, and today we're pleased to announce old.cal.com
Our design team have been working really hard on this one
People using Haiku even for small internal automations are delusional (it's also dumber than Deepseek)
> Their five largest customers went from 36% of revenue in Q1 2026 to just 21% in Q3
Let me spoil this for you (don't fall off your chair):
- Cursor (SpaceX)
- Microsoft
- Google
- NVIDIA
- Amazon
Some interesting stuff from digging into @AnthropicAI's S-1 filing today:
- Claude Enterprise users who don't use Claude Code spend less than $75/month on average, but the top 5% who do use Claude Code spend roughly 20x as much
- They recently delayed pre-training a new frontier model for more than a month because they needed to improve security measures against model weight theft
- Over 80% of Anthropic's revenue in Q3 2026 came from model versions released within the preceding six months
- Anthropic expects to pay @SpaceX roughly $3.1B per month for compute once all contracted capacity comes online. They could effectively be paying $37B/year, up to $120.5B in total, but the contracts can be cancelled with 90 days' notice after early 2027
- Over 90% of contributions to Anthropic's own codebase in September 2026 were written by Claude. The average daily volume of code contributed per engineer in Q3 was 45x what it was in 2024
- Their five largest customers went from 36% of revenue in Q1 2026 to just 21% in Q3, but every single one of those five customers actually spent more in Q3 than ever before
- They recognized a $2B operating expense in Q3 just for matching employee charitable donations. It issued roughly 28 million shares in October to settle the obligation
- Their $42B Broadcom financing facility has an interesting clause. Normally, the notes carry 9% interest and convert into shares at a 10% discount, but if Anthropic misses certain revenue targets, the interest rises to 13% & the conversion discount increases to 15%
- Google has the option to purchase up to $10B of additional nonvoting Anthropic shares at the IPO price
- More than 11,900 customers now spend at least $100K annually with Anthropic. Over 1,900 spend $1M+ and roughly 250 spend $10M+
- Claude Cowork is seeing faster adoption than Claude Code did at the equivalent point after its launch