2 ms·

The Successor to RSS Is Git Clone

6 points by 8organicbits 2y ago

jauntywundrkind 2y ago
I listened to the Oxide Computing podcast on sagas recently, and it's like 95% a story for state and where it's kept and how it's shared. "Manage state of vm instances that are actually running" I so so so strongly believe eventually git will be the thing. Computing will just unfurl into git logs. We will be able to share our spaces effectively & build tools stop that. Lovely to see this article. Very fun future.
builddeployrun 2y ago
Successor doesn't seem like the right word since its purpose is different from RSS. I like the idea of linking back like this, will likely include it on my projects in the future. If adoption were high, a browser plugin to highlight supporting sites would be great.
kevincox 2y ago
I really don't get it. > You get a copy of the entire works of your favorite authors Atom supports pagination. If it was a major issue people would support it. But the nice thing about RSS is that you don't need to download everything. It is your choice. Unlike if every post ever is at the HEAD commit of the repo. > that is blazing fast, works offline, is searchable, transformable, censorship resistant So is am RSS feed once downloaded. > auditable, more trustworthy How? > ad free, tracker free Just the same as RSS? > There is no comparison to the old ways of doing things. There really isn't because it doesn't even solve the problem that RSS does. How do I tell what the new posts are? How do I tell the difference between a post and the homepage? How do I tell if a new post appears? How do I know the URL based on the file name in the repo? The link provided doesn't even go to the Git repo, it goes to the web UI (although I think in most cases GitHub will redirect you if it notices a fit client). I think linking to the source of open source sites is quite cool. But I don't see how this is succeeding RSS. I could imagine a Git-based feed protocol such as each commit just containing a single HTML file. But just linking to a random source repo doesn't do it.
- breck 2y ago
  It might help if you tried out a working implementation of this (the WWS client): https://wws.scroll.pub/readme.html https://wws.scroll.pub/readme.html

6 points by 8organicbits 2y ago

I listened to the Oxide Computing podcast on sagas recently, and it's like 95% a story for state and where it's kept and how it's shared. "Manage state of vm instances that are actually running" I so so so strongly believe eventually git will be the thing. Computing will just unfurl into git logs. We will be able to share our spaces effectively & build tools stop that. Lovely to see this article. Very fun future.

builddeployrun 2y ago

Successor doesn't seem like the right word since its purpose is different from RSS. I like the idea of linking back like this, will likely include it on my projects in the future. If adoption were high, a browser plugin to highlight supporting sites would be great.

kevincox 2y ago

I really don't get it. > You get a copy of the entire works of your favorite authors Atom supports pagination. If it was a major issue people would support it. But the nice thing about RSS is that you don't need to download everything. It is your choice. Unlike if every post ever is at the HEAD commit of the repo. > that is blazing fast, works offline, is searchable, transformable, censorship resistant So is am RSS feed once downloaded. > auditable, more trustworthy How? > ad free, tracker free Just the same as RSS? > There is no comparison to the old ways of doing things. There really isn't because it doesn't even solve the problem that RSS does. How do I tell what the new posts are? How do I tell the difference between a post and the homepage? How do I tell if a new post appears? How do I know the URL based on the file name in the repo? The link provided doesn't even go to the Git repo, it goes to the web UI (although I think in most cases GitHub will redirect you if it notices a fit client). I think linking to the source of open source sites is quite cool. But I don't see how this is succeeding RSS. I could imagine a Git-based feed protocol such as each commit just containing a single HTML file. But just linking to a random source repo doesn't do it.

breck 2y ago
It might help if you tried out a working implementation of this (the WWS client): https://wws.scroll.pub/readme.html https://wws.scroll.pub/readme.html