SourceCred: Future of work or dead weight?

[peth]

Before you click, know the title is a bait.
I think SourceCred had a solid chance of being the future of work despite its several shortcomings, if it had focused on solving the UX/UI issues.
As it currently stands, its a project that nobody besides us is using (!nor maintaining), its a piece of MetaGame we’ve been pulling along & keep on putting effort into, without gaining much if anything.
Delayed distributions, ever unhappy contributors etc. etc.

After using SourceCred for almost 3 years & using Coordinape for about a month, here are my 2 cents.
Let me break it down:

New problems

• Its an abandoned project that nobody is maintaining, we’re stuck with a huge deprecating codebase that constantly needs to get fixed & updated just to stay operational
  • We have very few active builders & their time is precious
• Integrating Coordinape into it means spending more time on something that doesn’t work & isn’t likely to survive

Historical problems

• Lack of proper UI for evaluating work & analyzing the contributions graph, leading to ever-present hacky solution of didathing, the need for a creditor & a visualization tool that never got built
• On the building side, its just quantification which is a crude measurement of value creation because not all commits & PRs are created equal
• On the other side, pushing everything through the whole-community-evaluation with 5 levels of rewards through emojis, crude evaluation + people with no context evaluating + disparity in XP weights with an appearance of emoji numbers
• Historically, people who were doing random stuff got over-rewarded, whether it was just writing a lot, writing a lot in didathing
• People who were doing things that carry the project forward have been getting underrewarded

Upcoming problems

• Just imagine the shitshow if we had more than 20-30 active contributors
• Plugging in Coordinape into SC & into didathing would mean builders getting double or triple dipped

Random solutions

• Forcing builders to all go through didathing would basically be forcing them to use a worse functioning system because theirs isnt available to others
• Increasing automatic tracking of everything by increasing weight on discord & discourse tends to reward people for talking. (PRs & commits aren’t all made equal but you can at least be sure they aren’t all talk)
• We can keep on talking about fine-tuning the weights but after 3 years still dont have anybody committed to doing it & ultimately:
• None of these solutions solve the core problems
• None of these solutions solve the fact SourceCred is an abandon project with an aging codebase

Upside

• SourceCred has been with us for so long & abandoning it would feel bad
• We’re the last carriers of SourceCred, if we had the resources to maintain it & take it to the next level, it could be a part of our product offering
• Doesnt require everyone to be added to a Coordinape team so it could be cool for community-wide rewards, as a base pay & as pay for people who aren’t active enough to join a team but still contribute from time to time

SourceCred vs Coordinape

• Posting in didathing feels like shit & induces anxiety because its like a huge community-wide scrutiny arena, whereas writing contributions in Coordinape feels natural & safe
• People with no context or understanding of different domains evaluating each other’s work ends up with bad evaluations & hurt feelings. In Coordinape, evaluations are team-based. Team members have most context into the work being done, how valuable or not valuable it is etc.
• Rewards being team-bound means we can more easily see who is not pulling their weight or not using the tool properly, who is being under-rewarded or over-rewarded & most importantly, we can more easily affect change
• Rewards being team-bound also means we can do some budgeting & stop the project from devolving into developing in a single dimension or disperse funds to meaningless contributions
• Coordinape’s UX/UI is really smooth & straightforward. It requires no debugging headaches & technical people to set it up. I can recommend it to people without 50 caveats.
• Distributing Seeds through Coordinape took a whopping total of 10 minutes & went without a hiccup.

My conclusion

I love SourceCred.
Its been an integral part of MetaGame for the past 2+ years & it definitely does have its upsides.
If we had millions of dollars in our treasury & dozens of builders, it would be cool to have a few builders dedicated to it fully time, build the creditor etc.

Coordinape works damn well & seems to be working on more integrations for automatic tracking of contributions. If we really want a tool that quantifies shit, there are other projects working on solutions.

I think we should fully transition to using Coordinape & pause (probably permanently) SourceCred distributions.

New problem: what happens to XP?
We maintain the XP system. Imo we should simply make it an abstraction of how many Seeds a person has received from the multisig. Number of Seeds * 36 or whatever random number.

Thoughts?

To me, honestly, it really seems the costs far outweigh the benefits at this point.
Did I miss anything? Redeeming factors? Reasons not to abandon SourceCred for Coordinape?

[dysbulic]

I really like SourceCred and would hate for us to stop using it.

I propose that we build an adapted version of Coordinape that natively supports multitiered distribution and builds a graph to be a part of the SourceCred graph.

We can diminish the weight of GitHub to mitigate the issue of double dipping. Ultimately, I’d like to connect the GitHub activity to the Coordinape contributions so that cred flows more thoroughly through the graph.

I think SourceCred is pretty close to a solidly usable product. The main things I think need to be done are to switch from using a text file as the data store to using Neo4j which will allow for interfaces that deeper investigation as to the sources of XP. Also, I want to add caching so that the system isn’t building the entire graph each time.

I’m going to try writing a Filecoin grant describing the functioning of this system and see if I can get them to let me switch from my original grant to this.

[peth]

@dysbulic

I propose that we build an adapted version of Coordinape that natively supports multitiered distribution and builds a graph to be a part of the SourceCred graph.

Sounds like a heavy lift side quest.

@dysbulic

I really like SourceCred and would hate for us to stop using it.

I really like SourceCred but I like MetaGame more. If we had $10m & 50 builders, I’d think its a great idea to keep building SourceCred & turn it into a great product that it could have been.
With $20k & like 3 active builders, I honestly think we should just call it a day & move on

There’s lots of costs related to maintaining it & keeping it running, then there’s a literal shitton more required to build on top of it & take it to the next level.

Disregarding the fact we love it & its a cool product:

• Whats the benefit of doing so?
• Whats the painpoint that it solves for us that Coordinape doesn’t?
• Why is it more important than improving MetaGame as a platform?

[vid]

I think we can make it work with just Coordinape. It’s simple and straightforward. It requires no maintenance. It will save us hours of work.

[Misanth]

Why baiting? This isn’t youtube to compete for views

its a piece of MetaGame we’ve been pulling along & keep on putting effort into, without gaining much if anything.

wdym without gaining much of anything? The xp and ranking system was at the base of our gamification and the seeds that came from sourcecred fed us the past two years?

Integrating Coordinape into it means spending more time on something that doesn’t work & isn’t likely to survive

Whether or not will it survive depends on the fact whether we abandon it or not.

Plugging in Coordinape into SC & into didathing would mean builders getting double or triple dipped

Plugging Coordinape into SC will result in Coordinape being plugged into SC. “Builders” or any other role double/triple dipping has nothing to do with coordinape/sc inherently since:

• Sourcecred is configurable
• It’s all a matter of agreement between relevant parties(read: a modus operandi is established) on how should contributions be tracked/reported

Random Solutions
…

I don’t see the point of this paragraph since you’re basically just listing random thoughts heavily biased towards the opposite side of the argument than the title. If you’re going to offer solutions, there’s no point in purposefully listing issues. You could’ve written those under problems.

Upside
{…}

I will just add my own perceived upsides:

• it’s a revolutionary aspect of tracking contributions and relationships between participants regardless of the difficulties
• It’s something different - My understanding was that metagame is trying to change how we collaborate and manage interactions by flipping things upside down, turning the powers at play on their heads (even our logo breathes that connotation). The appeal with sourcecred was that it didn’t care who you are, as long as you participated. I really do not know of any other system that said fuck you to the corporate way of managing staff like sc did.
• Flexibility - you literally tell sourcecred what you think is valuable and what isn’t via plugins (weights)
• Cred(XP in MG) can in no shape or form be replicated with any other known tool. If it can, someone please let me know asap

SourceCred vs Coordinape

• Posting in didathing feels like shit & induces anxiety because its like a huge community-wide scrutiny arena, whereas writing contributions in Coordinape feels natural & safe

What do you do in did-a-thing? You report, get evaluated and evaluate others.
What do you do in Coordinape? You report, get evaluated and evaluate others.
I don’t see how one is different than the other in that regard. Anxiety comes from within, not from outside->me, an extremely anxious person that took 30 mins for every report regardless of its length.

• People with no context or understanding of different domains evaluating each other’s work ends up with bad evaluations & hurt feelings. In Coordinape, evaluations are team-based. Team members have most context into the work being done, how valuable or not valuable it is etc.

I get the point about evaluating outside of one’s domain but it’s something that can be easily solved but it’s way more fine-grained than the black n white example. Among other things, it is important to consider we average about 8-10 evaluations(emoji reactions) per did-a-thing. How many out of scope evaluations can there be? I think this is more of a human/framework issue than sourcecred/coordinape.

• Distributing Seeds through Coordinape took a whopping total of 10 minutes & went without a hiccup.

I think you’re comparing apples to oranges here. Distributing seeds takes about 3 minutes after the calculations in SC too. (feed csv to gnosis disperse>approve>boom). The part where we struggle is rate limits considering our user db size and the calculation in general.

New problem: what happens to XP?
We maintain the XP system. Imo we should simply make it an abstraction of how many Seeds a person has received from the multisig. Number of Seeds * 36 or whatever random number.

Under the hood this means replacing page rank theory with a basic multiplication formula. Not only that but, the laws of logic dictate that if ( XP=SEED.received * Y) then that means we’re directly corelating experience with money earned. Therefore, if I am providing things for free, I’m not earning XP?
That’s a dangerous precedent.

I think there’s lots that we can do to achieve what we want without making this radical change. I don’t think we’re doing proper justice with these comparisons. Also, when I try to boil down the direction this would put us on, I can’t help but feel like we’re giving up on coming up with new solutions and frameworks and reverting to corporate ways of handling a company because it was “hard” and wasn’t raking in fat stacks of $$. I hope I’m wrong.

Conclusion:

Strongly against completely abandoning sourcecred in favor of coordinape.

[polimyl]

To be honest, SourceCred is a big reason of why i joined MetaGame. If MetaGame didn’t have SourceCred i likely wouldn’t be here now.
Why? I’ve been (and still mostly am) a rogue with social anxiety that just randomly does a very useful thing once in a while. SourceCred automatically picked up most of my contributions and made my XP go up I contributed and number went up without anyone having to do anything.
This is the power of SourceCred: being able to automatically convert contributions into a value number, from all possible sources, without any human bias or other shit.
So in my opinion SourceCred is an integral part of what MetaGame is and if theres any possibility of us continuing to use/maintain it then we should take it.
Now, we didn’t really explore what possibilities there are (at least to my knowledge).
Example Possibility: We could pitch SourceCred to some people for a grant to officialy become the maintainers of it. Then we could make it a part of MetaOS, integrate QuestChains etc.
Probably also wouldn’t hurt to tell the SourceCred discord that the last big user of SourceCred thinks about abandoning it, maybe they also got some ideas.
So we should at least explore more possibilities, SourceCred deserves this after how much it has done for us.

Coordinape is just a glorified #did-a-thing with a more casual format. People post what they did, other people value it. This alone is in my opinion not sufficient.
It also isn’t permissionless, so it’s not really suitable for newcomers.

[alec]

@peth

As it currently stands, its a project that nobody besides us is using

This is not true. SC has been and is still being used as an integral part of several DAOs’ value attribution system, including ones many times larger than us (see e.g. MakerDAO). That said, they are funding someone (less than FT) to maintain their instance. Sourcecred is quite complex and does indeed have significant costs in both manual maintenance time and daily server compute.

@peth

or maintaining

This is unfortunately true. The core product is receiving only very minimal community maintenance, which is unsustainable. I am doing most of the maintenance of our own instance, and understand it pretty well, but there are some pretty major limitations that make it much harder than it should be to maintain. That said, it’s probably less than 5% of my time spent on MG-related work.

I would love for SC to receive adequate funding to be self-supporting, but IMO it’s its own thing and should not get mixed up with MG.

To me, SC has always been a primary quantitative tool and Coordinape qualitative. I believe that having both is the best of both worlds (and I bet the SC community would agree!). It seems to me that we could solve most, if not all, of peth’s value attribution concerns with a bit of thoughtful design (in the architectural sense). Ditching did-a-thing for coordinape, adjusting the cred params, etc.

With that said, I don’t think abandoning SC would really buy us much, though we will need to put in a bit more effort into better integrating the two systems so it’s not a PITA for contributors.

[peth]

@Misanth

wdym without gaining much of anything? The xp and ranking system was at the base of our gamification and the seeds that came from sourcecred fed us the past two years?

And now that we’re using Coordinape, whats the benefit of pulling it along?

@Misanth

I don’t see the point of this paragraph since you’re basically just listing random thoughts heavily biased towards the opposite side of the argument than the title. If you’re going to offer solutions, there’s no point in purposefully listing issues. You could’ve written those under problems.

Thought 1) is the one we all most often recommended when we were talking about it & thought 2) was your proposed solution.
Also, I’m not acting like I’m trying to be unbiased here. I am very clearly advocating abandoning SC.
The whole point of this thread is for those who are against to voice their side of the argument.

@Misanth

I don’t see how one is different than the other in that regard. Anxiety comes from within, not from outside->me, an extremely anxious person that took 30 mins for every report regardless of its length.

Have you tried it?
Its very much different because its not a chat & its not written where 500 people can see it but limited to 8.

@Misanth

Among other things, it is important to consider we average about 8-10 evaluations(emoji reactions) per did-a-thing.

This is actually a great point against SourceCred. Whereas with Coordinape, it doesn’t matter whether someone allocates their GIVES or not, in SourceCred, if a high weight emojier goes on a vacation, everyone else will literally get paid less too.

@Misanth

I get the point about evaluating outside of one’s domain but it’s something that can be easily solved

How?

@Misanth

I think you’re comparing apples to oranges here. Distributing seeds takes about 3 minutes after the calculations in SC too. (feed csv to gnosis disperse>approve>boom). The part where we struggle is rate limits considering our user db size and the calculation in general.

Yes, the very final step takes 3 minutes but maintaining & getting it not to fail on calc has caused mints to be more than a month late at times & caused numerous headaches both to Lux & Alec as well as all of MG.

@Misanth

then that means we’re directly corelating experience with money earned. Therefore, if I am providing things for free, I’m not earning XP?
That’s a dangerous precedent.

In SourceCred, people earn XP & get paid in Seeds
In Coordinape, people earn GIVES & get paid in Seeds
In both SourceCred & Coordinape, there is no Seeds without XP nor XP without Seeds.
Its not a dangerous precedent, its not a precedent at all.

@Misanth

I can’t help but feel like we’re giving up on coming up with new solutions and frameworks and reverting to corporate ways of handling a company because it was “hard” and wasn’t raking in fat stacks of $$. I hope I’m wrong.

Whats corporate about Coordinape?

@polimyl

Why? I’ve been (and still mostly am) a rogue with social anxiety that just randomly does a very useful thing once in a while. SourceCred automatically picked up most of my contributions and made my XP go up I contributed and number went up without anyone having to do anything.

Eeeh, shouldn’t the goal of metagame be to help you battle & get rid of social anxiety rather than finding a technology that enables it?
Tho I must repeat Coordinape is easier on social anxiety.

@alec

This is not true. SC has been and is still being used as an integral part of several DAOs’ value attribution system, including ones many times larger than us (see e.g. MakerDAO). That said, they are funding someone (less than FT) to maintain their instance. Sourcecred is quite complex and does indeed have significant costs in both manual maintenance time and daily server compute.

Idk if MakerDAO is still using it but afaik the biggest user besides Maker was 1Hive & they stopped ages ago. MakerDAO itself only ever used it for rewarding governance posts on the forums, literally nothing else (at least when I last checked).

@alec

This is unfortunately true. The core product is receiving only very minimal community maintenance, which is unsustainable. I am doing most of the maintenance of our own instance, and understand it pretty well, but there are some pretty major limitations that make it much harder than it should be to maintain. That said, it’s probably less than 5% of my time spent on MG-related work.

Is it only 5%? Has it been? With all the problems over the past X months?
I wonder how much time @luxumbra spends on it monthly…

@alec

To me, SC has always been a primary quantitative tool and Coordinape qualitative. I believe that having both is the best of both worlds (and I bet the SC community would agree!).

I believe so too & I think having a proper graph of all the contributions etc. would all be extremely cool, but…
Edit: Actually, I don’t even fully believe so. I think some work will always be quantifiable whereas other won’t (eg. building vs design) & as such will always be double dipped if its rewarded both through quantifiable as well as qualitative means through Coordinape, no matter how we configure the weights.

@alec

I would love for SC to receive adequate funding to be self-supporting, but IMO it’s its own thing and should not get mixed up with MG.

Wouldn’t mind it getting mixed if we actually had the resources to do both SourceCred AND MetaGame

Honestly, I think you’re all missing the point here.
Besides Alec maybe

The Point

Resource constraints.
Instead of getting caught up in all the ways we love SourceCred, let’s try to do a sober analysis of the costs & benefits.

• We got literally like 3 active builders who make up a little over a single person working full time, when all added together.
• We got less than $20k left in the treasury

To get SourceCred to work properly:

• Maintenance costs
• Server costs
• Revamping the weights (this will take a significant amount of time & discussion by a bunch of people, its also a role we’ve been unsuccessfully trying to fill for 2 years)
• Building the Coordinape plugin
• Getting SC over to Neo4j
• Make it not have to run everything every time
• Building the actual UI for it & whatever the fuck else

The point is:

• Its a whole damn product of its own & a big, complicated one at that
• Costs related to getting it to run properly, let alone take it to the next level, are significant
  • We are talking more $ than we have & diverting like 50% of our building power to it

Imo both SourceCred & Coordinape are means to an end - accounting systems for building MetaGame.

• The cost of using & improving SourceCred is literally tens if not hundreds of times bigger than just using Coordinape (which works smoothly & requires no maintenance or improvement on our end).
  • Are the benefits 10 times bigger as well?
  • Is SourceCred as important to us as MetaGame itself?

I’m not saying we should never return to SourceCred but currently I think it would make more sense to pause it than keep on building on it.

[peth]

@peth

The point is:

• Its a whole damn product of its own & a big, complicated one at that
• Costs related to getting it to run properly, let alone take it to the next level, are significant
  • We are talking more $ than we have & diverting like 50% of our building power to it

Imo both SourceCred & Coordinape are means to an end - accounting systems for building MetaGame.

• The cost of using & improving SourceCred is literally tens if not hundreds of times bigger than just using Coordinape (which works smoothly & requires no maintenance or improvement on our end).
  • Are the benefits 10 times bigger as well?
  • Is SourceCred as important to us as MetaGame itself?

Since nobody is addressing these questions, I’ll continue to drive my point home & try to wrap this up.

The platform we are building (MetaGame itself) has a bunch of glaring problems:

• The weird hamburger loading bug
• Cards being spaced ridiculously apart on players list
• There’s no profile pictures in the profile creation flow
• The leaderboard takes like a minute to load
• Etc. etc. etc.

(These are only the problems, not even mentioning possible new features that could take MetaGame itself to the next level in terms of being a useful & well-rounded platform.)

Ultimately, we’re averaging a total of 30 visitors per day, staying less than 2 minutes.

But peth, what do any of these have to do with SourceCred & its coolness

Nothing & everything.

The point I’m trying to illustrate is quite simply, that we don’t even have enough dev power & money to fix the problems in MetaGame/MetaOS itself - let alone the problems of 2 other tools.

Fixing SourceCred’s problems & properly combining it with Coordinape would be epic, I admit.
But, to think we can build the most epic accounting system at the same time as MetaGame itself, while we can barely build MetaGame itself, is, in my most sincere opinion, delusional.

In the end…

If we don’t build a platform thats useful to people, that gets people to adopt MetaGame & join the DAO space - even the most intricate & epic value attribution system wont save us.

My proposal

Not worry about it for now.

• Pause SourceCred for Season 9
• See how well Coordinape serves us on its own
• Focus on improving MetaGame the platform itself
• Start gathering more user feedback so we have a better idea of what features & parts of MetaOS should be prioritized

Then, in Season 10:

• If proper value attribution is a problem & people are requesting improvements, revisit this idea improving SourceCred itself, revisit the weights config, do a proper Coordinape integration etc.
• (Also, maybe connect with topocount (prev. SC core dev) & see how Coordinape’s automatic value attribution system (inspired by SC) is progressing, look into Govrn, into BitDAO’s SC-like attribution system, whatever other solution there is, follow up with @METADREAMER about MetaCred & overall do more research before plunging ahead)

Good enough?

Is this a good enough conclusion for now? Do we need to vote on it?
Thoughts? @dysbulic?

[s_ben]

Noooo! Don’t leave!!

As part of the skeleton crew at SourceCred keeping the lights on (just barely), this is a tough thread to read. My heart is stirred by those defending SC. I also see where @peth is coming from. A few thoughts and suggestions that hopefully move the discussion forward

@peth

As it currently stands, its a project that nobody besides us is using (!nor maintaining)…

Well, not quite…Maker recently decided to stop flowing rewards, but is still running the instance and exploring ideas like displaying Cred scores on the forum. Stacks Advocates DAO is still running ~$2,000 worth of tokens (via CredEquate). The Smart Contract Research Forum (SCRF) is distributing 5,000 DAI/mo, and I’ve been working with them on their instance. Here’s an article I wrote for them that addresses some of the technical and operational challenges of maintaining and building on SC instances.

Flying the SourceCred Plane

There are also a number of other instances out there, which admittedly we haven’t been tracking programmatically for a while. But if you search @sourcecred on Twitter, you’ll see a stead drumbeat of projects using or recommending SC. E.g. last week Geo Web gave us a shoutout

SC has also, despite its lack of shipping, joined many people’s lists of recommended DAO tools. E.g. another tweet from last week:

God’s work!

@Misanth

Integrating Coordinape into it means spending more time on something that doesn’t work & isn’t likely to survive

Whether or not will it survive depends on the fact whether we abandon it or not.

Unclear IMO. It would be a blow for sure. But SC has broken through enough, and it’s solving a problem that I don’t see other tools solving.

@Misanth

Cred(XP in MG) can in no shape or form be replicated with any other known tool. If it can, someone please let me know asap

I would also like to know! I think SC has reached PMF, despite its lack of shipping. At a minimum, it’s proved valuable as an engagement measurement and incentivizaiton tool, which has proved particularly useful for recruitment into more formal roles. It can also be viewed as a general value exchange in-community, which is revolutionary if not yet profitable. Lots of other exciting use cases have been talked about, if not actively explored. I also sense some projects out there building interesting stuff in private, but who knows .

@peth

We’re the last carriers of SourceCred, if we had the resources to maintain it & take it to the next level, it could be a part of our product offering

@polimyl

Example Possibility: We could pitch SourceCred to some people for a grant to officialy become the maintainers of it. Then we could make it a part of MetaOS, integrate QuestChains etc.

FYI, as some here may know, I was exploring the last few months a reboot of SC, finding funding. But ran into headwinds. Notably, a death in my family, a huge mess surrounding that, mental health issues; depression, ADHD diagnosis, some heavy grief and trauma work
). As difficult as it is to say, I feel I should put it out there. tldr; doesn’t look like I have the spoons for cofounding/leading anything, at least for a bit. I apologize for being unresponsive to people reaching out, including some MG’ers.

While I’ve not been as active in MG recently, I have been tracking the possibility of handing MG maintenance (and development if it chooses) to MG (i.e. transfer GH repos to MG, officially announce SC’s support). I floated the idea on the SC community call, as well as to some key contributors still around, and am seeing broad support for the idea. MG is one of (if not the only) community with enough value-alignment and history to trust with that. Especially considering myself and others are still planning to build on SC.

@peth

We can keep on talking about fine-tuning the weights but after 3 years still dont have anybody committed to doing it & ultimately:

A Heavy Lift

If MG were to commit to developing on SC, I would love to jump in as a proper player! I could bring some missing pieces, such as knowledge of the algo/product, connections (I can think of dozens of deep-pocketed people/projects interested in SC), governance of algorithms and implementing configurations, decent meme game.

However, I would not recommend MG pursue taking over SC unless it has the resources and alignment to 1) actually execute new features (a heavy lift) and/or 2) have a viable business/cryptoeconomic model that can fund builders. The demand is there, no doubt (we’re still getting serious interest here and there but can’t respond to it). But there appear to be some big challenges to meet it.

@peth

There’s lots of costs related to maintaining it & keeping it running, then there’s a literal shitton more required to build on top of it & take it to the next level.

So I don’t think simply maintaining and running it would take that much tbh. As I outlined in Flying the SourceCred Plane, the core algorithm and plugins are solid. SC may not have been able to ship new features (that got traction anyway). But it has spent ~3 yrs fixing bugs, building some new features in response to real users, and generally hardening the codebase. There don’t appear to be any critical vulnerabilities in the algorithm when using the main plugins (at the level of the algorithm). I’m determined to work on it, even if that means just paying somone off UpWork a few hundred dollars a year keeping the APIs pulling data (commodity work). But I would be surprised if other projects didn’t pick it up.

However, SC predictably runs into two issues when a community starts using it:

• It’s labor intensive. If your instance is small, and you’re not running large amounts of money, it can basically be set and forget. But eventually most communities require some skilled labor to just maintain it.
• Desire for new configurations/features. If significant rewards are on the table, communities inevitably want to tweak the parameters. Which raises governance issues/labor. People also invariably get ‘source pilled’, and want to create new features, plugins, etc. It’s too damn compelling!

Building new features and plugins does require significantly more resources and coordination. For instance, @dysbulic’s Coordinape plugin idea.

@dysbulic

I propose that we build an adapted version of Coordinape that natively supports multitiered distribution and builds a graph to be a part of the SourceCred graph.

This idea is soooo good!! It’s totally doable and I want to support exploring this. It’s also really hard for me to gauge how much work it would be, what unseen challenges it could run into. I don’t know if it’s tens of millions of dollars like @peth is suggesting stuff like this could cost, but it’s not a side project either.

I do think it’s possible for MG to leverage SC to raise a lot of money, should it decide to recommit. Especially if it can coordinate well enough to deliver and support new plug-ins/features (so much damand for that…much strategic partnerships). The main thing I worry about though, whether its SC or another MetaOS product, is decentralized development. Please forgive my tangent here, but it is a potential blocker for me contributing.

@Misanth

It’s something different - My understanding was that metagame is trying to change how we collaborate and manage interactions by flipping things upside down, turning the powers at play on their heads (even our logo breathes that connotation). The appeal with sourcecred was that it didn’t care who you are, as long as you participated. I really do not know of any other system that said fuck you to the corporate way of managing staff like sc did.

Feel this! This is certainly the vibe SC was founded with, what still motivates me personally, and what gives MG the credibility with SC to even have this convo. AND, after spending ~3yrs in SC struggling to ship, watching MG and countless other DAOs do the same, suffering through a lot of community conflict (which I’m still healing from ), it appears that decentralized development is still a largely unsolved problem.

Reading the latest governance proposals, I’m really hopeful. It seems MG is right on the cutting edge of what is possible in terms of decentralized coordination. I can feel the hard won experience in every detail. I think using Coordinape is a great way to achieve the minimum viable centralization to achieve its goals. I’m here for it. I also think for these mechanisms to work, you’d need a solid, aligned team willing to slog through the frustration and inevitable conflict/failure modes you encounter along the way. I speak from wounds that are still healing, so please take what I’m about to suggest with a grain of salt. But I suspect to pull this off, MG needs to rally around leaders that are strongly aligned and have the authority to make hard decisions, including removing permissions and payouts if need be. The visible conflict I see in the discord makes me twitch and hesitate to contribute, and I’m probably not alone. The leaders also need material support to step back as these mechanisms decentralize that authority (yet another basically unsolved problem on top of everything else).

Perhaps now I’m speculating beyond my depth, but I also wonder how MG will differentiate itself without SC? Many DAOs are using Coordinape now, and the DAO tooling landscape has expanded significantly–mostly by VC-backed startups with fast-moving hierarchies. In a bear market, it’s hard for a traditional centralized startup to raise funds. Communities of rebels with experimental decentralized governance structures I imagine are an even harder sell. Having watched MG since the beginning, it’s been amazing to watch its beautiful vision slowly materialize (and it is materializing!!). But to make it not painful and frustrating for the core contributors to keep re-committing, it needs capital. Which probably means a solid business model and/or proven cryptoeconomic model.

Just my 2 satoshis. I apologize for the delayed reply, possibly ignorant takes and wall of text. Also apologize if any additional replies to this are delayed, as I’m leaving tomorrow morning for Costa Rica from Portland and facing a full travel day. Will be there for the next ~5 weeks and working, but bandwidth will be less.

Regardless of how this goes, <3 you MG!

[Misanth]

Just put up an official vote and lets get done with it, there’s no need to further clash individual community members philosophies as that’s what i see is going on here. Every single reply on this thread was against the idea of abandoning SC so it’s not a good enough conclusion to go forward with your proposal without a vote.

Also, I tried to resist replying to the following but I can’t :

@peth

The whole point of this thread is for those who are against to voice their side of the argument.

@Misanth

I don’t see how one is different than the other in that regard. Anxiety comes from within, not from outside->me, an extremely anxious person that took 30 mins for every report regardless of its length.

Have you tried it?
Its very much different because its not a chat & its not written where 500 people can see it but limited to 8.

You casually forgot it was @tenfinney, @dysbulic and me who initiated coordinape here way back and also suggested the grouping by guilds (before we consolidated our internal guilds). I also believe @FearlessThompson started the idea even before that. So, yes. I have tried it.

[peth]

@Misanth

Just put up an official vote and lets get done with it, there’s no need to further clash individual community members philosophies as that’s what i see is going on here.

It’s much healthier to come to conclusions through discussion than resorting to voting without necessarily actually addressing the concerns or going deeper than scratching the surface.

Also, Dysbulic is the one actually building stuff on SourceCred. Luxumbra is the one using it to distribute rewards, Alec is the one bugfixing & maintaining it and I’m the one fundraising for MetaGame, so us 4 agreeing matters a lot more than random people voting “no” from the sidelines.

And seriously? You see this as a difference in member philosophies?

@Misanth

You casually forgot it was @tenfinney, @dysbulic and me who initiated coordinape here way back and also suggested the grouping by guilds (before we consolidated our internal guilds). I also believe @FearlessThompson started the idea even before that. So, yes. I have tried it.

I casually didnt forget none of you actually started formal proposals & actually got us all on Coordinape.
Everybody likes “suggesting” things without seeing them through as official executed proposals.

You casually forgot to address the actual point & questions I posed. But ok.

@Misanth

Every single reply on this thread was against the idea of abandoning SC

Factually incorrect. Vid agreed with abandoning and neither Alec nor Seth (whom I actually expected to be biased because he’s from SourceCred & still suffering from SC dying) actually picked sides but talked about both the good & bad.

That makes:

• 2 for abandoning
• 3 against
• 2 inconclusive

Also, not a single reply actually addressed the concerns that started this thread

As a personal side note, I’ve noticed this pattern with you before & honestly, I really don’t like the way you debate things. You cling on to the weakest points made & not the actual concerns. You spin personal accusations & try to make me look dishonest while saying things that are factually incorrect, yourself. Its like politics 101 instead of trying to get to the bottom of things.

[Misanth]

Project as much as you wish, no one is stopping you ser. You’ll do as you please in any case. Best of luck regardless of the outcome. I’ve nothing more to contribute in this discussion

[peth]

@s_ben

Stacks Advocates DAO is still running ~$2,000 worth of tokens (via CredEquate). The Smart Contract Research Forum (SCRF) is distributing 5,000 DAI/mo, and I’ve been working with them on their instance. Here’s an article I wrote for them that addresses some of the technical and operational challenges of maintaining and building on SC instances.

Love hearing there are others still using it - is there a list available somewhere? Would be great to reach out & see if they’d find the MyMeta profiles/leaderboard or the Coordinape plugin that Dysbulic just built useful.

@s_ben

SC has also, despite its lack of shipping, joined many people’s lists of recommended DAO tools. E.g. another tweet from last week:

Also great to hear but it does make me wonder how many of these people simply know what SourceCred is/that its an awesome tool vs have actually used it extensively or at least tried deploying…
I love it but personally stopped recommending it unfortunately - I still believe its incredibly cool, just not worth the hassle for most projects (esp. those just getting started)…

@s_ben

I think SC has reached PMF, despite its lack of shipping.

I don’t think it has but honestly think the only thing keeping it back is UX. If only the devs focused on building the UI & some kind of an installer/deployer, it probably would have went through the roof. As it is, Coordinape kind of just came along & ate its lunch

@s_ben

As difficult as it is to say, I feel I should put it out there. tldr; doesn’t look like I have the spoons for cofounding/leading anything, at least for a bit. I apologize for being unresponsive to people reaching out, including some MG’ers.

As one of the people who reached out to you & offered help in rallying support & getting funding (feel the need to say this only because people in this thread seem to think I have something against SourceCred itself ) - still down to do it.

@s_ben

While I’ve not been as active in MG recently, I have been tracking the possibility of handing MG maintenance (and development if it chooses) to MG (i.e. transfer GH repos to MG, officially announce SC’s support). I floated the idea on the SC community call, as well as to some key contributors still around, and am seeing broad support for the idea. MG is one of (if not the only) community with enough value-alignment and history to trust with that. Especially considering myself and others are still planning to build on SC.

Feeling honored good ser! I offered the same above mentioned support to Dysbulic in case he wants to continue with improving the integration in the coming months, building SC into Neo4j etc. - just still not sure it should be done by MetaGame itself or if it should be another project, maybe by SC & MG (+ other interested projects) together?

@s_ben

However, I would not recommend MG pursue taking over SC unless it has the resources and alignment to 1) actually execute new features (a heavy lift) and/or 2) have a viable business/cryptoeconomic model that can fund builders. The demand is there, no doubt (we’re still getting serious interest here and there but can’t respond to it). But there appear to be some big challenges to meet it.

Right, thank you - that’s pretty much my whole point here. I think its very much a worthwhile effort, just don’t think it’s reasonable for us to do it at this point in time when we’re struggling even with metagame.wtf itself…

@s_ben

However, SC predictably runs into two issues when a community starts using it:

• It’s labor intensive. If your instance is small, and you’re not running large amounts of money, it can basically be set and forget. But eventually most communities require some skilled labor to just maintain it.
• Desire for new configurations/features. If significant rewards are on the table, communities inevitably want to tweak the parameters. Which raises governance issues/labor. People also invariably get ‘source pilled’, and want to create new features, plugins, etc. It’s too damn compelling!

These points are what make me push against it. Besides the lack of UI & other UX problems, main problem with SC is that it requires constant tweaking which adds overheads not the least of which is governance. Can’t even imagine what it must be like with big projects using it. The need for governance should be minimized & having 1000 variables doesn’t help

@s_ben

I don’t know if it’s tens of millions of dollars like @peth is suggesting stuff like this could cost, but it’s not a side project either.

Oh, sorry for the confusion, I didn’t mean to say this specifically would require $10m - I meant to say that if MetaGame as a whole had $10m & more than 2 devs, then it wouldn’t be a problem to dedicate a few hundred k & a dev or two to work on this full time…

@s_ben

Feel this! This is certainly the vibe SC was founded with, what still motivates me personally, and what gives MG the credibility with SC to even have this convo. AND, after spending ~3yrs in SC struggling to ship, watching MG and countless other DAOs do the same, suffering through a lot of community conflict (which I’m still healing from ), it appears that decentralized development is still a largely unsolved problem.

Yessir! Don’t think anyone has nailed it yet & tbh starting to think projects shouldn’t decentralize until they already have an actual product market fit & are at least somewhat sustainable. Decentralized from day 1 only means people pulling in different directions from day 1 when alignment is most important.
Once there’s a powerful project, it makes sense to decentralize it so it doesn’t become corrupt. But starting decentralized… Its like, what exactly are we decentralizing?

@s_ben

I speak from wounds that are still healing, so please take what I’m about to suggest with a grain of salt. But I suspect to pull this off, MG needs to rally around leaders that are strongly aligned and have the authority to make hard decisions, including removing permissions and payouts if need be. The visible conflict I see in the discord makes me twitch and hesitate to contribute, and I’m probably not alone. The leaders also need material support to step back as these mechanisms decentralize that authority (yet another basically unsolved problem on top of everything else).

Not necessarily around leaders but we should definitely align around north stars & treat every side quest as irrelevant until we reach a point of sustainability.
Right now each of us spent hours on this discussion when the main problems of MetaGame are 1) that there are 30 people using the platform & 2) that we have almost 0 funding & we won’t be able to get more until we at least partially solve 1).
Arguing over which contribution tracking system we’re going to use feels completely irrelevant to the pressing issues of actually building a massive online coordination game.

@s_ben

Perhaps now I’m speculating beyond my depth, but I also wonder how MG will differentiate itself without SC? Many DAOs are using Coordinape now, and the DAO tooling landscape has expanded significantly–mostly by VC-backed startups with fast-moving hierarchies. In a bear market, it’s hard for a traditional centralized startup to raise funds.

I definitely don’t think using Coordinape would make us a traditional centralized startup, the main difference I see is that it would allow us to have less overheads & more control over the budget.
Without SourceCred, MetaGame would hopefully differentiate itself by building MetaGame itself, the platform, the massive online coordination game (SourceCred is used for building MetaGame, not using MetaGame). There still isn’t anything quite like MetaGame out there

I really appreciate you weighing in here with sober thoughts. Seeing you post this huge message, at first I thought it would be 100% pro-SourceCred but I was glad to see that you put in proper effort into explaining the downsides as well

[markop]

Hi all, I’d like to take this opportunity to introduce Praise as a potential solution to aforementioned problems outlined by Peth.
Disclaimer: I’m affiliated with Praise.

Praise is a project backed by the dev team at General Magic, it’s being continuously developed and maintained, with a defined roadmap ahead. It is already being used by communities and DAOs such as Gnosis, DAppNode, BrightID, ReFi DAO, Commons Stack, TEC, Giveth and more being onboarded ever day!

Currently integrates with Discord and Telegram, with more integrations to come (Slack and other communication tools).

Praise is super easy to get started with, both for the community and for the builders:

• It doesn’t require everyone to be added to a new product
• It doesn’t require everyone to learn a new tool

All you have to do is start praising others for their contribution and accumulating the data which you can use and turn into whatever you want.

Praise provides a transparent and verifiable record of community contributions and they are rewarded based on actual contributions and impact, and not “just talking”.

Praise can produce a powerful contribution graph (currently in development) and with the power of AI it generates user profiles, bios, skills and contributor scores and more.

To address the “team-bound” evaluation, Praise has a built-in Quantification process where you decide who the quantifiers will be (team mates) and how to assess the evaluation.

The Praise system is user-generated data, trustworthy, and un-gameable.

For a limited time and selected number of communities the Praise team is offering to help with:

• Configuring Praise for the first time (discord roles, allowed channels etc.)
• Helping you get started praising
• Help you get started with quantifications

Learn more about Praise.

If you’re interested, we can do a community demo to show you how it works.
Alternatively, if you’re a member of any communities already using Praise, head over to their Discord and try /praise

[Dave]

One more vote for Coordinape. It’s been so refreshing to use it, as opposed to SC.

I probably don’t know much about the history and the broader context. What I do know is that as a contributor, I experience SC as unclear, unfair and overly complex to use.

If not Coordinape, then something else, but please no more SC and -did-a-thing

[peth]

So it also turns out that, contrary to what we thought for the past 2 years - github PRs are in fact all worth the same amount of XP regardless of the number of commits inside.
Commits temporarily gain XP but then get merged into 1 when the PR is merged to main.

Idk what’s more problematic:

• The fact we still don’t know how SourceCred rewards things, after 2+ years of using it, or
• The fact its rewarding all PRs the same, regardless of whether its a 30 second styling fix or a full codebase refactoring