Has there been research / discussion on nesting cred instances? I did a quick search in Discourse and Discord and didn’t return anything that seemed to be exact match. Forgive me if this has already been discussed or is already in someone’s mindeye (@decentralion ).
A specific example might be more helpful.
SourceCred (SC-CRED) and Commonwealth (CW-CRED) are two projects in the “Governance” (GOV-CRED) space. So there are three possible cred scores:
SC-CRED
CW-CRED
GOV-CRED
They both have plugins running separately on Discourse and Github. So in total, there are four plugins:
CW-Discourse
CW-Github
SC-Discourse
SC-Github
Currently, it’s clear how CW-Discourse and CW-Github influence CW-CRED, but a few questions:
how does CW-Discourse and CW-Github potentially influence SC-CRED? There may be contributions on CW-Discourse that influence SC-CRED, so we can tag and can “grandfather” some cred to them that they can “claim”. Possibly, another node can be introduced so that another project or contributor once someone gets mentioned so that they can get cred flowed to them.
How does CW-Discourse and CW-Github influence global GOV-CRED? This may be as simple as creating a global GOV-Cred and then shifting the weights between the four nodes, so that GOV-Cred can be allocated equitably to all four plugins.
If there is a GOV-Cred, how is cred shared between nodes that are a project (CW and SC) and user? Perhaps another node can be specified, individual and project. And the GOV-Cred distribution will have another weight to adjust.
In going down this thought experiment, the value of a Credchain becomes pretty apparent. As new plugins are created, they could act as an oracle flow the history of their graph to chain. Each plugin, user, and node gets represented as an address with a human-readable name. Any front end could pull from this identity graph and that node could be attributed even if they aren’t actually signed up using the plugin’s login system. Note, the case around @protocol not really having so much cred. As more projects start to use the CredChain, they will continually reference other project’s plugin’s, users and more. Later on, meta-cred can arise from this. The graph and embedded reputational value that is captured that presents a really, really strong open platform with cred computed by a set of trusted validators.
Zooming out, one could imagine this type of attribution happening an arbitrary number of levels down, and we could get to a global cred score, representing all the separate domains of scientific research, artistic expression, and more.
On a more serious note, however, I think what you’re describing is kind of the idea that Colony is taking on with reputation weighted voting that’s split into departments. Here however, you’re imagining different organizations that have their own Cred, then there’s categories that have Cred, and so on all the way up until a meta/master-cred level? Is that correct? If so, that would kind of be like an org with different departments, but in a non-linear always changing map - awesome
Also, and this might be a wild leap here, but are you saying that in this model everything would have a “Cred address” that would be a first class agent in the system that can have it’s own SourceCred instance determining it’s Cred? This way CW-Discourse could have an address and run it’s own SourceCred instance, but then things within it like categories would also have addresses and SourceCred instances that would run over threads in each category, then same for threads running an SC instanace over comments in threads, etc… So everything would have it’s own SourceCred graph that connects to a larger graph? (kind of like how on Notion every page/task/thing is a first class item that can then have more pages/tables/things inside of it). Feel free to ignore this if that’s just crazy talk lol
If I’m not imagining what you’re imagining, however, then please correct me and/or provide another explanation. Turtles are cool, and so is Cred, so I really want to get this lol
This is basically 100% of what I was getting at! I think it would be really cool. I think with a “meta-Grain” token this can be a really powerful way for meta funding of research, public knowledge. Something kind of like a cross between Synthetix and Colony. Where the Proof of Work is the cred that accrues.
I think the point around Colony is really astute as well. Have always thought that Colony is really cool, but thought that the implementation felt a little too top down. Although, I haven’t kept up with the latest work so I may be greatly mistaken. When thinking about Colony, I had always imagined Rep being allocated by some “boss man” in a manual process. The SourceCred automation through plugins is a really elegant solution that can automatically provide a lot of value
Really think these meta constructs will be important, because not every project is equally able to obtain “benfits”, like funding. Sometimes simply because the nature of the project is more low-profile, but still very valuable.
A good example of this is comparing code libraries, with a finished app. The attention that openssl and firefox get, is very biased towards firefox. But which is more valuable, that’s probably much more evenly matched.
Here’s a question about this though.
Flowing cred between projects is much more objective when you’re considering raw activity. What we’ve learned so far though, is that having humans in the loop and to debate and set their values is a crucial part of making Cred match up well. So Cred is a lot more subjective, it has culture and opinions embedded.
