>Think about some sort of publish/subscribe system, in which a web-page creator can regularly hit a publish command that makes it available for archiving, and various web archives can subscribe to receive updates
This seems like something we could (should?) have right now. Maybe IA should write a Wordpress plugin, if they haven't?
>Think about creating an archive of software as well, that perhaps may have to include emulations of defunct hardware and operating systems to make the Web always backwards compatible.
A site that archives software and runs it in js-based emulators sounds like a great idea. It would probably be illegal, though. And it almost certainly wouldn't work properly for everyone, as long as it depended on the browser. But still a great idea. That any runtime and software can have a URL is incredibly compelling.
Maybe we need to leave the browser model for documents and come up with something else for using what amounts to streaming software?
>Change the naming system, and stop thinking of the URL as a location—it’s a name, a format he picked to look like a Unix file name simply because people were comfortable with that.
>Think about some sort of publish/subscribe system, in which a web-page creator can regularly hit a publish command that makes it available for archiving, and various web archives can subscribe to receive updates
This seems like something we could (should?) have right now. Maybe IA should write a Wordpress plugin, if they haven't?
I was vaguely thinking more of publishing and updating directly to IA through some kind of API, (rather than having them scrape everything - that may actually exist, I don't know), something you could even add as a button to a text editor. But that is pretty close.
>Think about some sort of publish/subscribe system, in which a web-page creator can regularly hit a publish command that makes it available for archiving, and various web archives can subscribe to receive updates.
Sounds like RSS, which I assume a few sites still use. Most people seem to prefer Twitter.
This seems like something we could (should?) have right now. Maybe IA should write a Wordpress plugin, if they haven't?
>Think about creating an archive of software as well, that perhaps may have to include emulations of defunct hardware and operating systems to make the Web always backwards compatible.
A site that archives software and runs it in js-based emulators sounds like a great idea. It would probably be illegal, though. And it almost certainly wouldn't work properly for everyone, as long as it depended on the browser. But still a great idea. That any runtime and software can have a URL is incredibly compelling.
Maybe we need to leave the browser model for documents and come up with something else for using what amounts to streaming software?
>Change the naming system, and stop thinking of the URL as a location—it’s a name, a format he picked to look like a Unix file name simply because people were comfortable with that.
YES. No TLDs, just unique arbitrary strings.