"If our generation, uses the seabed for storing carbon that we shouldn't have emitted in the first place, then the generations coming after us won't be able to use the seabed to store their emissions."
Кадр: Telegram-канал «Ирина Волк»
。新收录的资料是该领域的重要参考
I'll admit this is a bit idealistic. The history of open formats is littered with standards that won on paper and lost in practice. Companies have strong incentives to make their context files just different enough that switching costs remain high. The fact that we already have CLAUDE.md and AGENTS.md and .cursorrules coexisting rather than one universal format, is evidence that fragmentation is the default, not the exception. And the ETH Zürich paper is a reminder that even when the format exists, writing good context files is harder than it sounds. Most people will write bad ones, and bad context files are apparently worse than none at all.
44 src: *src as u8,