What does and does not count as an Initiative?

The CredSperiment
1

This topic is not an Initiative. As mentioned in the comments below, this thread is an experiment to explore what does and does not qualify as an Initiative. This topic does not qualify as it’s really a contribution to a larger initiative.

Description:

A place for ideas that could be turned into initiatives.

Status: Proposal

Completed

Champion:

@burrrata

Benefits:

Then we can keep track of useful ideas rather than losing them in chats. Then people can browse the Initiatives Wish List to find a list of open problems they can contribute to.

Implementation plan:

Create an Initiatives Wish List category. Then make each new topic in that category a wiki. It is expected that ideas are rough, so making them editable by the community will help to refine them over time. People can then champion ideas by creating an Initiative and linking to it in the wish list topic.

Deliverables:

Initiatives Wish List category

Initiatives Wish about page

Dependencies:

References

Contributions:

@burrrata brought the idea up

@decentralion created the category

@burrrata then organized the category

2

So… I’m starting to dive into the Retroactive Cred Activation project and trying to wrap my head around what does and does not deserve it’s own Initiative. Wasn’t sure about this one so I thought I’d write it out as an Initiative to then see how I felt about it. In this case, I would lean towards saying that this Initiative could be rolled into the Organizing SourceCred's Discourse Initiative. Curious to hear everyone else’s thoughts on what does and does not deserve to be an initiative :slight_smile:

@Beanow @decentralion @s_ben what do you guys think?

3

This does feel very small for an initiative, and like it would be better seen as a single contribution to a larger initiative.

It’s a fuzzy dividing line / grey area. But it seems like the total amount of work here was maybe an hour or less? That feels like “single contribution” zone to me.

4

That’s what I was thinking too. Do you think the best way to accomplish this would be to just link to the Initiatives Wist List category in the Please help us organize this Discourse Initiative? and if so, what would that Cred flow look like on the back-end?

Agreed. Maybe we should add a section to the SourceCred contributor guides that if something takes an hour or two first check to see if it could be added to a current Initiative and/or a current Artifact. Then if it doesn’t fit anywhere else, consider creating a new Initiative or Artifact.

5

If you just link to the category, then it won’t really flow cred to you for the contribution.

I think you can add a contribution line like:

# Contributions
- Setting up the [initiatives wish list](...): by @burrrata

If you want to give me some cred for helping out, maybe we could use the following syntax (not yet support, but I think it will be important to add eventually)

# Contributions
- Setting up the [initiatives wish list](...): by @burrrata (4x) and @decentralion (1x)

Which would send you 4x as much cred, which makes sense IMO, since you did a lot more work for it than I did.

2 Likes
6

Good idea. I like it :slight_smile:

That being said, we might not want to keep this process open ended. Who gets X% of each contribution could create a lot of cognitive friction and/or conflict. On the other hand, if we put constraints on the process then Cred flow to each contribution might be less efficient. Overall, however, I think a simple standard for contributions might increase the user experience and lead to more people creating better initiatives.

What if we said that each contribution has 5 points, and you can allocate those points however you want among contributors. That way we can still give people credit for helping out, but we don’t have to go through an entire evaluation process every time - instead we can juts give them one of the points.

7

Among different solutions, I think the system I demo’d (choose a multiple for each contribution) is the easiest to work with. Saying that there are 5 points is basically the same as saying there are 100 points, i.e. assign percentages upfront. This becomes a problem whenever we want to add more contributions to something that exists.

Here’s an example. Let’s say we have an initiative with two contributions, one of which is much more important than the other, so say we do:

  • Contribution1: 4 points
  • Contribution2: 1 point

Now say that someone adds another contribution which is in between the other two. The simplest would be to say:

  • Contribution1: 4 points
  • Contribution2: 1 point
  • Contribution3: 2 points

However, now the points no longer sum to 5. So we’d need to instead do some math for:

  • Contribution1: 2.85 points
  • Contribution2: .7 points
  • Contribution3: 1.4 points

It’s much simpler if we simply use the multiplier approach:

  • Contribution1: 4x
  • Contribution2: 1x
  • Contribution3: 2x

(You’ll note this is a slightly different example than what we’ve been talking about already – it’s an initiative where we are adding another contribution, rather than a contribution where we are adding another author. But we should be able to use the same system in both cases. And we can just as easily imagine this being a single contribution, where the third person came along and added some more work to it, or we realized after initially writing it that we forgot to credit them.)

8

Agreed. I was thinking about it as splitting up a pie vs growing a pie. This approach seems better.

To make sure I understand, is the multiplier then acting as a boost to the amount of Cred? Ex: Cred flows through an Initiative and then flows to the contributions to that Initiative. If there are multiple contributions to a larger Contribution, then multipliers determine how much flows to whom.

If this is the case, then would all contributions have a default multiplier so that some Contributions don’t get more Cred simply because lots of people made lots of small contributions?

Also, if I’m not understanding how the multiplier system works then can you please explain how it would work in more detail?

9

The multiplier determines how much cred flows out from the node to various connections.

If we think of every cred as water, every node as a “pool” that can hold water, and the edges as “pipes” that connect the pools, then every contribution edge is a pipe that connects an initiative to various contributions. Setting a higher multiplier is like increasing the width of the pipe.

For a concrete example, suppose that we have an initiative I and contributions C1 and C2. Let’s suppose I has 100 Cred and initially neither C1 nor C2 are connected to it.

If we connect I->C1 @ 1x (a 1x weight edge from I to C1) then C1 will be worth around 100 Cred, which flows from I.

If we connect I->C1 @ 1x, I->C2 @ 1x, then C1 and C2 will both wind up with around 50 Cred, since the Cred from I is split.

If we wanted to express that C1 is more important than C2, but we still want C2 to get some cred, then we could do I->C1 @ 4x, I->C2 @ 1x. In that case C1 winds up with around 80 Cred and C2 winds up with about 20 cred.

The default multiplier is 1x, so 2x means “twice as important as a standard contribution”, 0.5x means half as important, and so forth. So the following would be a valid example:

Contributions

  • Contribution 1
  • Contribution 2
  • Contribution 3 (5x)

Which corresponds in the formal syntax I invented to:

I->C1 @ 1x, I->C2 @ 1x, I -> C3 @ 5x

Does this clarify things?

10

Yeah I think so. So really, the multiplier is multiplying by a percentage because there’s only so much Cred that can flow downstream. Total Cred flowing to a contribution is then split between the contributions. If there’s a single contribution then all the Cred goes there, but if there’s multiple contributions then Cred gets divided between those contributions based on the percentage multiplier on each contribution.

100 Cred => 1 Contribution (1X) = 100 (100 * 1/1)

100 Cred => 3 Contributions (2X, 5X, and 1X) = 25 (100 * 2/8), 62.5 (100 * 5/8), and 12.5 (100 * 1/8)

Like that?

11

Yep!