I don't use IRC because I generally have to think a little or most of what I write comes out as gibberish. Some times it comes out as gibberish even when I do think about it. By the time I answer something it's...gone.
I've been reading about Syndie some. The world desperately needs something like this. The best of all worlds would be a distributed store like where you can male your own websites like Freenet, encryption like I2P, data centered store like Freenet and IPFS, a way to "pin" sites like zeronet so you could choose what you keep on your computer, serve all the forums, pictures, movies or whatever like BitTorrent and have a way like Syndie to control who post what where and to make web sites that are distributed. And yes I know I don't want much.
My understanding is that cjdns ran into problems with larger sets of users and that Yggdrasill has solved some of these problems. User reviews say Yggdrasill is faster. This is just what I've read I don't know personally. There also was a podcast where someone compared the two and he said that cjdns was bundling itself up with one of the digital coin systems which he didn't like and I agree is not ideal. I wonder if Yggdrasill's way of finding addresses in the network would be faster than I2P's?
I also would like all the software of Syndie to run "where I put it" not have it spread data in various places on the main drive. I know this is against the grain but slowly people are beginning to do this again. Appimage files, flatpack and other packaging formats are moving everything a program needs into one folder. The reason for this is fairly simple. I want to know where and pick where files are written without a massive amount of hunting around. Where temporary files are written can get fairly cryptic at times. If you surf the dark net then occasionally you run across something you don't want on your computer at all and if it's on your main drive this can be problematic. If all files are on a separate encrypted drive then you don't have that problem and you can move the software and drive where you wish. It's not tied to the OS drive.
Here's my very limited "possible" big picture of how things should be. You need two DHT's. One is a data store DHT hash and one is an Address DHT or the network. We already have a bit of that with I2P. I2P has addresses and you run BitTorrent over it which has it's own DHT. If you could search in both then I think it would be the best of both worlds. IPFS I think just looks for data DHT hashes. I'm not sure how an actual web page is linked or if there are addresses. Reading the data sheet it says it works over any network (ipfs-p2p-file-system_copy.pdf). I don't understand how the separate pieces of data re actually tied together as in say a web page with lots of different pictures for example with text describing them. The data part is good because you can have many copies. My thinking is you could have an address that is tied to a DHT piece of data. So let's say you look up an address and the original site is not up but there's copies. With that address is also a data hash so you could look at the data hash to find the page and as most pages are in reality a bunch of cat pictures then you could just link data and address hashes to the data of the cat pictures. So you could find both. I think I'm running into gibberish but it is a complicated thing.
Anyways I'm rambling but you asked so...