# Signaling session outline

### A. General information
| | |
| -------- | -------- |
| Session title | Signaling ring |
| Facilitator | James Hancock |
| Scribe | *Text* |
| Time and date | 13:30 - 15:00, 29.10.2018. |
| Material requirements | *Text* |
| Space requirements | *Text* |
| Number of participants | *Text* |

### B. Expected outcomes
1. Action plan of needed tools based on review of the following events.
2. Block Reward Reduction
3. Ropsten Bork
4. Delaying Constantinople

### C. Expected outputs
1.

### D. Timeline

| Duration | Segment | Description |
| -------- | -------- | -------- |
| 20 mins | Product status + Feedback | |
| 15 mins | Context in which the magicians may want to know stances? | When, how often and for which cases Discussion of latest issues: mining reward, ropsten and pushback of the hardfork |
| 15 mins | Backlog and feature priorities | Are there additional signals we can use? |
| 40 mins | Fishbowl discussion | |
| 10 mins | Wrap up | |

James:
We came up with having some kind of website.
V1.0 includes coin and gas signalling + stances on twitter. Thanks, Giveth and Aragon for the support.
Since the last meeting we had 3 community events:
* reducing the miners' fee
* ropsten fork
* Constantinople push back

We need to discuss tools that could help with managing that better.

Other agenda items:
* Multisigs and DAOs endorsement. How do you Signal / DAOs best practices?
* List of the key sources of the information we need. Like hash power or specific influencers like 0x or maker. Criteria definition
* How do we create incentives for people to signal?
* Translate data from signalling / ethereum community to the investors (family offices), hedge funds, miners and educate them on what is good info and data. There's a lot of info – telegram, magicians, calls – too much for a usual investor. Also convenient distribution channels
* Do we care about the standardized format for those signals? (there was a fork choice coordination proposal, e.g. specific block to transit by Dan, Strange loop, Metamask)

# Events and signal sources

On fee reduction: The idea to reduce the reward was thrown out. There was some debate, but how do people signal for support. Core devs took some miners representatives on the dev call. And some compromise happened regarding the fee and ASICS. The guy with big powers said – if I can't make money I can use my powers elsewhere. No idea who runs carbon vote.

The biggest challenge with coinvotes – legitimacy is not very high. Devs won't take in consideration at all. So I'm not putting my funds out of ledger if they won't be checking anyway. If you win – you'll point to that, but not otherwise. How do you legitimize them?

Griff: coinvotes are not that valuable without other signals. Miners will, of course, be against. Most impact comes from influencers. If we can grab them from sources and track their decisions in something like a wiki format.

There's no common belief on who should influence the decisions. People don't necessarily want miners or devs to decide. The challenge we're gonna face is to have a better-formed ideology on who should decide.

Afri: some issues are not contagious, but people can make them such.

Maybe we can assign someone, do delegative voting for tactical issues.

As long as you have a grouping – it's ok. Pure coninvote can be one-sided (e.g. let have the minimum miner fee).

## Timing of voting/signalling

EIP has a maintainer and he makes the last call. So last call timing makes the most sense.

Also, it's important – who asks the questions. What kinds and how they're asked?
Another thing is EIP can be wrongly represented.

## Types of signals

Do we have any other data points besides coins? We have gas and miners voting. You can even allocate power to different types of users (e.g. miners and app devs).

Maybe we need to start with identifying the stakeholders and ways to reach them. Spoke word is one, written work, in-person meeting and etc. You can also check for non-verbal communications. One way to signal emotions – is signaled by emojis. There could be a lot of "confused" reaction. Better then yes or no.

Signals are non-binding. But somehow it can come expensive to go against the majority decisions.

That's how it works now – you can fork but that's impractical. Devs push the client release and they can choose. If they disagree also by selling the coins. The signals important for the core team would be more about the tech questions rather than economic. And investors do vice versa.

You can find a lot of legitimacy from the influencers stances (Vitalik, Vlad etc.).

In non-profit world, consensus is a qualitative thing, not quantitative. Not ideal to quantify that.

You can signal with some percentage of uncertainty. But we can use some metrics and criteria. The uncertainty can be solved with the liquid democracy/delegation.

You can end up with no decision (bitcoin situation), but the ability to fork is valuable. It's a very difficult problem, but moving to better things seems like a better solution.

If we can make exits frictionless, then deciding would be way easier. With blockchain has an opportunity to choose. And in the end, the dominant strategy is to get there.

If the signal is clear we can avoid contagious issues and splits (e.g. 60% are for this).

On ropsten fork, would there be a signal that helped you – no.

Why isn't it easier to help stakeholders to reach core devs. Even for core developer, it's hard to understand some questions (like the roadmap). So can be for many others.

That's where research teams come in and can structure that.

There's no concept & constitution to guide that.

It's not that we don't have enough data, it's just is not disclosed.

There's a couple of datapoints restricted by the tech e.g. full node voting. Nodes stuff is possible – Infura knows IP and software running. It's not sybil resistance, but incentivizes node operators. What if Catalonia becomes independent and decides to use Ethereum for governance and block all other IPs. This is out there in Bitcoin (ethereum as well at ethernodes.com). You can put a flag in the client.

Coin.dance could show how many nodes use the right software. On ethereum, 27k nodes are available with OS, version, IP etc. Something like this should be available and open sourced.

# Prediction markets

Markets are the best way to aggregate the information.

You can leverage the prediction market in the situation of high asymmetry. With 2-dimensional prediction market, if you can combine those results it's more informative.

The best signals are giving you the info you don't know. In prediction markets, stances are known. It's more suitable for avoiding things like hard forks.

In EIP0 we had an extensive list of influencers.

Weighting the signals can feel wrong. The best course is to provide the data without the interpretation.

Influencers list is the most controversial topic. But it's the potential way for delegative representation (e.g. delegate all my votes for Vitalik).

The action item can be to raise that issue (tasks) further in magicians. Like Quora, HiveOne.

Discussing it here is better than on Reddit. It's a delicate balance. Last time discussing 999 we've found several people using multiple accounts to signal.

Action item:
* List of stakeholders
* 2 new EIPs (standardized format of signals, miners )
* Ethernode for tennagrahp
* Join Quora and Riot/Telegram chat, or Tweet to James and Griff