I have always thought Named Data Networking (a fut...
# thinking-together
k
I have always thought Named Data Networking (a future infrastructure project older than IPFS) is really promising: https://named-data.net/ Think instead of naming locations, naming the data you're requesting, so it can be pulled from any host as efficiently as possible
i
Ah, neat! I feel like I've never heard of this project before. But then looking at the list of papers, I probably have heard of it, but just missed the fact that "Named Data" is a proper noun in this context, haha. I'm curious if anyone here has worked with this, or has any ideas for how they might use it.
d
I understand it to be under the broader umbrella of "Information-Centric Networking". It's possible to characterise my own Object Network concepts as under that umbrella also, so this is of interest to me 😁👍🏼
I have a recollection that, like the Web, a fundamental omission of this architecture is not having a mechanism to continue to forward data updates to previously interested consumers, to save constant polling?
s
Van Jacobson’s tech talk is one of my favorites:

https://youtu.be/gqGEMQveoqg

It’s about Content Centric Networking but I find it interesting because he talks about about how perspectives that change the frame of discussion (unspoken assumptions etc) can be powerful. The example he talks about first is going from circuit switching to packet switching. Packet switching did indeed “win” even though everyone was at one point trying to make circuit switching fast. I think CCN didn’t catch on with any open standard the same way as packet switching. However it seems CDNs also use principles similar to “named data”? It’s just implemented as an app layer on top of TCP/IP and seems to provide the same benefits for a subset of use case dealing with immutable media objects.