Starting writing an experiment AI agent in Rust. It is a long-running daemon. There is a CLI that connects to it via a local socket. The CLI can either be run directly in the terminal, or it can be even set as a user's shell. This means when SSHing in to that user you get immediate access to the agent.
Just released v0.9.1 of Acuity Index. It helps dapps on Polkadot/Substrate consume blockchain event data through a real-time indexing API.
Since v0.8, we shipped config-driven index specs, explicit query keys, hot reload, reindexing across spec changes, composite/multi-field keys,
Prometheus metrics, event proofs, JSON-RPC 2.0, better throughput/reliability, and stronger production safety. v0.9.1 adds simpler subscriptions,
cleaner event payloads, top-level timestamps and WS heartbeats.
https://t.co/XnrskoGL6x
What happens when you run a @Polkadot node on hardware it was never designed for?
A Parity engineer tried it on a RISC-V board over a weekend. Not for production but to see where things break.
@Leo_Glisic@yacineMTB I don't use office software very much, so I'm not the best person to advise you on that. LibreOffice and Google Docs are probably the best avenues to explore on Linux.
@jonmyers You can order many things on Amazon direct to Vietnam without a forwarder. They have an arrangement with customs so it goes straight through.
I asked GLM-5.1 if Helix (Rust neovim clone) correctly handles unicode rendering / editing. It said this:
Short answer: partially — grapheme-aware editing works, but rendering has known issues.
The core problem is issue #6012 (https://t.co/TEfOwSyb5r) — unicode_width returns column widths that don't match what terminal emulators actually use, causing misaligned rendering near emoji and special characters. unicode_width is even pinned to v0.1.12 in Cargo.toml with a comment that its newer versions "break our logic and disagree with common width definitions in terminals."
A fix using the widecharwidth crate (same one used by fish shell & wezterm) is in progress via PR #15599 (https://t.co/uMJcGtWhJo) but hasn't landed yet. RTL/BiDi support would require implementing UAX #9, which is a much larger undertaking with no current plans.
If any publicity is good publicity, then the reported Hyperbridge exploit (a third-party bridge) has done more for Polkadot awareness than the marketing bounty ever managed at a better price point.