Lets rename this project to a thing that better represents what it is

[Gozala]

As per ipfs/notes#436 I would love to use ipfs-lite as foundation for IPLD with persistence and (IPFS) network. However I think ipfs-lite is a controversial because it sets up wrong expectations for API (compatibility) and inter-op.

I think it would be a good idea to recognize this as a separate layer in the IPFS stack and give it it's own name & attempt to design best API for the DAG storage & replication without being tied to on that IPFS exposes today.

[carsonfarmer]

I am 100% on board with this. It is essentially a DAG Service so something along those lines makes sense. IPDS perhaps? I have not real preferences here, so happy to take suggestions/comments/recommendations from your perspective.

[Gozala]

If I recall correctly @mikeal had a suggestion of:

Inter Planetary Dag Persistence (IPDP)

I have another one:

Inter Planetary Data System (IPDS)

Or you can swap Data with Dag

However I'm personally not a big fan of acronyms as they prone to conflicts. And would much rather pick something else here's few

• DataNet
• DAGNet
• blockscope
• datascope
• blockspan
• dataplot
• blockplot

I'd like to capture that it represents part of of the larger network of content addressable linked data / blocks.

[Gozala]

/cc @lidel

[carsonfarmer]

DAGPeer or PeerDAG might also work?

[Gozala]

I also just tough of DAGDist or BlockDist as this a service that distributes blocks across the network.

[lidel]

No strong favorite on my end, but I agree we should try to come up with something with "DAG" in the name,
as it automatically sets proper expectations about the abstraction level at which it operates.

Inter Planetary DAG Service (IPDS) – boring, but also really clear for folks in our ecosystem
OR write it in a style that indicates its a lower-level plumbing-thing: dagservice → dagserv

[mikeal]

IPDP - Interplanetary DAD Persistence?

[mikeal]

hahaha, I had spent so much time reading comments I didn’t see that @Gozala already mentioned this name in the original post, my bad :)

[Gozala]

My reading of this thread seems to indicate that DAGService or dagserv suggested by @lidel is most popular ? If there are no objections, I suggest we pick on of the two and call it a day.

[mikeal]

Sorry for the late entry, but I just realized that I’ve been calling the commands I create to put dag’s on the IPFS network as “seed” commands and we hadn’t considered that here yet.

DagSeed? DagSeeder?

[carsonfarmer]

While I think DagSeeder is actually pretty representative of what this library ultimately will be, I'm not sure its the right name for an end-user. The seed or seeder language (to me anyway) infers something more akin to pinning or providing. But this library could be used in a "client-only" type mode, where it can fetch but doesn't provide. Service seems like the most robust name, so that's still my vote.