4 ms·
I think at the base of the development the protocols should be based. Where they are consumed is then up to experimentation.
by rv-de 6y ago
I think at the base of the development the protocols should be based. Where they are consumed is then up to experimentation.
- TeaDude 6y agoFair enough. I'm just spitballing stuff for potential future client software developers should someone ever go ahead with this idea.
- rv-de 6y agoSame here. I wouldn't be surprised if there are already many projects aiming into that direction. F.x.: https://prism-break.org/en/all/#mesh-networks https://prism-break.org/en/all/#mesh-networks, there's IRC, XMPP, ... But those projects mostly seem to try to establish alternative internets / protocols mostly based on the conventional internet. The active enforcement of low-bandwidth-use would in my opinion be the core concept. As soon as media can be transferred it's going down hill in many ways. I think usenet died this way. c/p, content sharing, viruses, spam, ... all that crap
- auggierose 6y agoI don't know what spitballing is but it sounds disgusting.
- TeaDude 6y agoA common term used in software development for "throwing crap at the wall" My instincts tell me it was coined in the 90's amid the disgusting/edgy phase after the long "dad jokes" phase of the years before. (Unix, byte, nybble ...the list continues)
- 0xdeadb00f 6y ago> A common term used in software development for "throwing crap at the wall" it's quite a common term outside of software development too. Edit: googling the term, it seems to have been coined in as early as the 1930s: https://grammarist.com/idiom/spitballing/ https://grammarist.com/idiom/spitballing/