CIP 1 Governance Proposal: Tech Agnostic

We should always use our own products when and if we can, however, not at the cost of not being able to function at 100% capacity as an organisation and not at the expense of the DAOs members. I believe this charter edit is necessary. That being said, I wholeheartedly believe we should dogfood when possible, test through the DAO when applicable, and build the best DAO whilst building the best tools. This edit allows for the flexibility we need to succeed, especially at the speed and pressure we are under.

1 Like

i’m going to be @mheuer’s understudy on this process, and would like to encourage anyone interested in seeing some new changes to please ping me on discord or in here and i’ll try to follow up with some tests on my forked copy

rock on!

2 Likes

I agree with Anthony and Ivan a, use 6th grade-level English so people will READ and secondly refer and understand our charter b, Know where to find clause in case of conflict or support and suggest improvement if there is a gap.

Charter is the skeleton of governance the “meat” is in policies and ethics. If the skeleton is sturdy which ANY member of Aragon can refer then the charter has done its job and we can fill in the gaps.

I would prefer to keep Charter tech agnostic but include it in our policies as we gotta use and promote our own products.

Updating via Google Docs or Github can be best option. I think Charter needs to seperate a few things and add some more sections. Will expand in other forum

2 Likes

I reviewed the proposed charter changes listed here:

Here my main criticism of the tech agnostic changes

In my opinion, words that specify a concrete tool such as ‘Aragon Voice’ or ‘Aragon Court’ can’t simply be replaced by the phrase ‘tech agnostic mechanism’.
‘Tech agnostic mechanism’ specifies nothing (except that there is some form of mechanism) so that people can theoretically flip a coin instead of using Aragon Voice or Aragon Court.
More generally, I also don’t understand how a DAO living on a blockchain and using web3 technology can have a tech agnostic charter. Do we mean vendor-agnostic?
Instead of naming a specific tool/vendor such as Aragon Voice, we can generalize this by naming the characteristics and requirements (e.g., ‘a percentage-based decision-making tool with properties X, Y, and Z’).
The proposers / people motivating this charter change should be precise what the characteristics and requirements are.

Example: ‘a trustless, sybil-resistant, percentage-based decision-making tool for signalling purposes with the possibility for on-chain execution’.

The same line of argument holds for Aragon Court, which cannot simply be replaced by ‘tech agnostic mechanism’.
Aragon Court as well as Kleros Court and Celeste are decentralized arbitration services. If we don’t want to have arbitration processes in the charter, we have to specify concrete alternative processes that replace them.

For the full review, see CIP-1 - Tech agnostic mechanism language by alibama · Pull Request #9 · aragon/network-dao-charter · GitHub .

1 Like

Does this not suffice?. If the DAO majority vote coin toss, then thats on us not the Charter…

How prescriptive do we want to be while aspiring also to a user friendly doc?

2 Likes

In a governance system, the checks and balances can only be meaningful if there is certainty about the jurisdictions and mechanisms of the institutions involved. Having placeholder language in constitutional texts can create problems down the lane, because each of these tools/institutions scope and limitations affects the other in ways that are difficult to predict in advance.

Experience in real life shows us that constitutional documents benefit from clarity/sharpness/focus, rather than inclusion/breadth/optionality, to the extend possible and as needed by the specific document.

In my humble opinion, especially in a situation where we do not have clear-cut, well defined alternatives to the tools in place right now, it is better to leave the Charter as is, despite the shortcomings of the current toolset.

As stated in the discussions, if there is a better tool out there, industry tested and with clearly better use-case than the current tools, then the community can discuss the pros/cons, analyze its potential effects across governance and checks and balances, and if the community signals the benefits to be greater, then these can simply be added to the text with a vote. Otherwise, I see boilerplate texts or agnostic-definitions as inviting more trouble than it is worth.

2 Likes

Ok I’m no dev but this seems a bit brilliant.

&

So building then on what @mheuer and @sertac advise would it be enough to propose that:

As Aragon transitions to a DAO we propose that:

  1. Governing technology must be
  • sybil-resistant
  • trustless
  • approved by vote
  1. Update the Charter to remove references that limit the DAO’s choice of technology.

Grade 6

EDITS: the opening proposal has been edited to simplify language and to introduce a definition of “tech agnostic”

We seek your feedback on a definition of “tech agnostic” that will best serve the Aragon Network long-term

Thanks for the change here Renee, this is certainly a difficult one to get right but as soon as we do hopefully everything else will get easier.

On the definition of Tech Agnostic, I think this is perfect with one further point:

  • Open source tools - we have seen the risks of closed source tech recently
1 Like

The issue I’m not seeing addressed is what makes the use of tool X legitimate instead of tool Y. Could someone decide to use a mechanism that fit the criteria but no one else in Aragon has used before? And what if said mechanism gives them advantages?

The tools/mechanisms need to be specified somewhere. Where will that be and how will that be governed?

Removing it from the charter is just outsourcing the problem but where to?

2 Likes

The way I am envisioning it would be these tools would be separate votes, eg the charter states that we should use a decentralised voting tool.

This would be voted on eg Snapshot vs Voice and then the tool that has been agreed will be used.

2 Likes

Updated to include open source. @daniel-ospina is this better?

approved via governance proposal & vote

As a process outside of the charter we can maintain a whitelist of approved tools here on the forum, as the tech committee does for security audits.

I don’t think that solves it.

The charter serves as “single source of truth”. And if this proposal passes, we now need a separate document that defines what’s the actual tool(s) used. That creates a usability challenge of how to make the new document visible/accesible (adding an extra entry in every web property to then add this extra document? that adds noise) and als that leaves the question open of how does one change this new document.

If the answer to the question is the MetaGov process, then why not leave the tool specification in the charter as it’s the same process and hence same requirements (quorum, etc) to change it.
If the answer is a new process, then that needs designing, adding complexity.

So instead of making a change in the charter to reduce specification, this proposal could become a change in the charter towards the actual tool that we want to use: just propose the tool and then replace (e.g. a proposal to replace all mentions of Aragon Voice with Spanshot)

The official whitelist can be linked in the Charter. And shared in places such as the Governance transparency efforts as a standard/process

I imagine the document is simply updated each time a tech tool is approved by the proposal + vote mechanism. Otherwise we end up editing the Charter, for every new tool. We still have to propose and vote on tool AND then propose and vote on charter edits

Apologies for cross-posting, but before writing a new Charter/Constitution, I recommend reading this: https://constitutions.metagov.org/

3 Likes

Thank you for the link @mlphresearch. For the optimistic governance setup that the AN DAO currently uses, the charter needs to be more specific and detailed because it constitutes the rules (“the law”) under which the DAO should operate. If it is too unspecific, it would be difficult for jurors to decide if a scheduled execution violates the charter or not.
When we switch from optimistic to direct voting mechanisms, the governance setup is clearly specified and enforced through the smart contracts and their configuration (“code is law”).
I am only facilitating and versioning the charter changes, but I imagine that the reworked charter will be much simpler and similar to the ones discussed by MetaGov.

While on the topic, let me say that I think that the sweet spot of DAO governance is having the possibility to do both, direct and optimistic execution, but in different scenarios, which will be easily possible with Aragon App.
Small proposals (e.g. people requesting a small bounty <$500) are not worth having a vote by the entire DAO (voter fatigue) and can be scheduled optimistically. A single watcher can then raise a dispute on a violating scheduled execution. An alternative to optimistic execution are committees with executive powers, to which such requests can be made.

4 Likes

I think we may be getting into semantics now in the sense of naming (and thus structuring) different parts of DAO documentation. Obviously, any level of detail can be included in the project documentation as a whole. But I think it’s better to move from core mission/values/principles to increasing levels of operational/procedural details as (1) the former informs the latter (i.e. it makes it easier to tie things together) and (2) it’s a great way to sequence introducing and onboarding new participants to the DAO when it comes to presenting and consuming information (i.e. you start with a high level understanding of the structure + core mission/values alignment, and then gradually move deeper into specifics as engagement and contributor roles become more specialized/sophisticated). Moreover, the problem with being overly focused on the details of written procedures is that these are much more likely to become inadequate over time (and thus require iteration), whereas the core principles/values tend to be more resistant to change (which is related to the fact that the spirit of the law can override the actual text when it comes to deciding whether something violates “the rules” or not). So my intuition is that commingling the two makes future iteration more resource-intensive. EDIT: Just to be clear, that’s not to say that various procedural details don’t need to be clearly established and version-controlled.

To draw a parallel with traditional governance systems: there’s a difference between establishing whether something violates a fundamental principle/value (ultimately decided in courts), or whether it violates a low-level procedural rule (operational checks and fail safes built into, for example, parliamentary or other organizational proceedings; of course, the latter question can also end up in court but the relevant rules to be checked are often not in the Constitution but in institution/field-specific legislation or even bylaws/guidelines internal to the organizations involved). The analogy may not be perfect but hopefully illustrates the point.

Of course, if existing structures/procedures absolutely require that the preamble to other AN DAO documentation (which is partly how I like to think of the Charter/Constitution) be more specific and detailed, then that’s that. But if the long-term objective is to move to a different structure (from core mission/values/principles to increasing levels of detail), then the link shared is probably as good a resource as it gets at this point in time.

1 Like

Agree in principle. Although such bounties may even be better administered not at the level of the whole DAO but in/by subDAOs that focus on the area the bounty is related to (e.g. some areas may require an ongoing bounty program).

2 Likes

<3

+1

LIVE VOTING Aragon Voice