4 ms·
I completely agree. I've seen similar ideas get shot down here on HN; but I think we need two separate protocols. One for purely documents, and a second for the
by bitshiffed 9y ago
I completely agree. I've seen similar ideas get shot down here on HN; but I think we need two separate protocols. One for purely documents, and a second for the universal application platform.
Among other implications, it should be much easier for regular people to create content, and applications should be free of the document-focused legacy.
- michael_fine 9y agoIf we did that, how would you propose implementing upvoting in HN without js?
- robin_reala 9y agoUpvoting in HN works without JS already, it just has a page refresh.
- zeveb 9y agoAnd it doesn't even need a page refresh; it could use a 204 No Content response. Potentially an yet-cleverer static browser system could also use 206 Partial Content to replace a portion of the page.
- krapp 9y agoBut at that point, it's no longer "purely a document," it's still an app, just one that runs entirely on the server instead of partly or mostly in the browser. For the web to be just documents, you have to go all the way and remove state entirely. Idempotence all the way down.
- yjftsjthsd-h 9y agoSo how about Gopher for documents, and HTTP(S)+HTML+CSS for apps? I think we have the tools already.