@vovsyy@colmtuite@base_ui We do the same. To add another level of complexity: our items are hierarchical. So instead of a flat list, we display those in a tree view. Haven‘t found an async pageable hierarchical combobox out-of-the-box anywhere
@StatisticsFTW I do the same. My suspicion is that working on (multiple, sometimes huge) markdown files is part of the problem. There must be something better. This is in my backlog of things I need to look at. Maybe it's time
https://t.co/RZzFyXaEVW
@poteto I've been working on sth similar. Roughly like this, which I found later https://t.co/Ym43yACDyi
You think markdown scales well enough compared to sth agents can query via SQL?
Does it work as a "team knowledge base" that's tracked in git that auto-synthesizes on pull/merge?
@awilkinson Got it looking for potentially interesting Vinyl releases based on my listening history and Discogs collection and message me with weekly findings
I think part of Clawdbot’s success is from being something a big co would never make.
Too much liability, messy business model, risky to deploy, ecosystem comparability…
That’s why it’s so great: it’s just useful instead of trying to be those other things.
@hasufl I'm running it on a Framework Desktop with a local MiniMax-M2.1. Works okayish. I find myself using my $17 Claude subscription until I hit its usage limits, then fall back to local. Kinda hope that Quality+Speed of local models improves with time
@notbrent@expo Got it. Is autolinkingModuleResolution the solution to have e.g. current React version for a react router 7 workspace and 19.2.0 for a Expo 55 workspace in the same monorepo? Or is that not really possible?
Installed @openclaw. It immediately realized that chromium was missing on my server which it would need to perform some tasks. Found my git repo where I track my package install scripts. Made the necessary changes. 15mins in, so much left to explore, it already feels like magic
@mjackson Always found it strange to harvest humans for energy. Seems inefficient. Apparently the original idea was to use human brains as CPUs for the simulation. This still seems a bit strange but a bit more interesting.