# The web as a system capability <!-- One deck for the whole series: each slide carries one part's core claim, and the narration says why it matters. The numbered parts make the full argument; this is the shape of it. --> --- ## Publishing was direct, and not naive <!-- The early web made publishing direct, and direct was not the same as naive. Understanding that difference is what explains why AFS is not a plan to turn websites back into folders. That is part 1. --> --- ## Then a URL started calling a program <!-- CGI let a URL trigger computation. It did not end file-based publishing, but it forced the web to deal with resources, actions and state at the same time. Part 2 traces that turn. --> --- ## A website became a system you operate <!-- Content management systems, databases, browser runtimes and hosting platforms each solved a real problem. Together they made publishing inherit the full operational weight of running an application. Part 3. --> --- ## The web lost the ability to answer back <!-- From weblogs through TrackBack to Webmention, independent publishing kept looking for a way to receive responses. The hard part was never the comment box: it is the identity, verification, moderation and presentation behind it. Part 4. --> --- ## In ARC, the web is a Device <!-- Calling the web a Device is not a fashionable rename. It puts publishing, live interfaces and application-specific logic back inside the system boundaries each of them should carry. Part 5. --> --- ## Shared semantics, open boundary <!-- Sharing system semantics can make it easier for sites to interact, but that convenience cannot be paid for with a closed network. The boundary is native cooperation and open protocols running in parallel. Part 6 closes the series. -->
