Meta: Convention Creation over Standardization

This post is intended to begin a meta conversation about the role of standards and conventions in the ecosystem, as well as the roles and responsibilities of ecosystem actors in creating such standards and conventions. Here I make the case for focusing on creating community conventions over standards based on an analysis of optimization processes and social capital.

In technology development or any other optimization process, all strategies service two competing goals: exploration of new territory, and exploitation of existing territory. Strategies which are over-exploratory tend not to find optima at all, whereas strategies which are over-exploitative tend to converge to local optima, rather than global. In the early stages of optimization, exploration tends to be more useful than exploitation, and this trend reverses later on. We, the community, are running an optimization process to enhance the ecosystem, and must as well balance these two competing concerns.

The term “standardization” implies some level of finality, i.e. standards are meant to develop very slowly, with broad consensus, and to service a very well-understood problem domain. Standards have both a higher bar for completeness as well as legitimacy (see the classic XKCD comic).

Conventions, while similar to standards in the sense that they both specify properties and behaviors, carry a more organic and exploratory connotation. The key difference between standards and conventions is the level of legitimacy required for their success. Conventions and guidelines tend to vary across industries and problem domains, and tend to develop more implicitly than explicitly as compared to standards. Standards require broad adoption and conformity in order to be successful, whereas conventions only require adoption and conformity within a given social niche.

The main questions when it comes to developing standards in the Polkadot ecosystem are these:

  1. Which groups have the legitimacy and social authority to create standards at all?
  2. Which problems are important and well-understood enough to standardize solutions for?
  3. What overlaps exist between (1) and (2)?

The answers I currently see to (1) are the Web3 Foundation and Parity Technologies. The Fellowship is emerging as another live player in the technical expertise domain, and may be the most apt as a future standardization authority, given its ability to signal on-chain.

The answers to (2) are very few. There is agreement over problems which are faced by the community: cross-chain accounts, XCM transacting, commonly used pallets in Substrate Runtimes, etc, but it is not clear that any of these problem domains are sufficiently well-understood enough to merit standardization.

The disconnect arises when looking at question (3): the type of expertise housed within W3F, Parity, and the Fellowship is primarily core expertise whereas the problem domains seeking standards tend to be ecosystem issues.

Therefore, I would argue that the ecosystem is not currently in a suitable place to focus on standard creation, and instead should focus on convention creation as there is a lack of highly legitimate authorities on technical ecosystem development, as well as a lack of well-understood and studied problem domains.

Instead I propose that these issues be addressed via the creation of conventions within the ecosystem. In fact, the ORML is already an example of a conventional set of pallets to be used by Substrate projects.

Focusing on conventions serves these purposes:

  1. Initial exploitation of promising solutions, while leaving the door wide open to further exploration. Conventions may be later refined into standards.
  2. Building deeper understanding of shared problems
  3. Faster iteration and non-finality
  4. Legitimacy-building within the ecosystem, for developers seeking to provide solutions to ecosystem problems. i.e. creators of useful conventions today will gather legitimacy to be the standard-setters of tomorrow.

Let’s discuss here.

11 Likes

I would like to echo this.

Create standards can be hard and actually not necessary the right thing to do for some cases. A standard that’s only used by a single party is just an implementation spec.

That’s why I stopped pushing for token standards but instead working on ORML. A lot of time, a de facto standard / convention is what we need. It is always a solid foundation to create a proper standard from it.

1 Like

This is definitely an important discussion to have. My opinion on the subject of standards in general is that in order for the ecosystem of interoperability to succeed, there have to be some ecosystem-wide standards for tokens, contracts, interfaces etc. I do not think that the onus is on w3f/parity to enforce these standards, it should be community driven and involve a wide range of participants from the ecosystem

Interesting topic rob - good to see a discussion digging deeper into potential strategic frameworks and insights. I realise this thread is primarily focused on technical conventions, but we should also actively discuss the emergence of conventions and standards that exist in a kind of meta-space within the ecology of the polkadot ecosystem, since future technical conventions may well be inspired by initiatives, rather than being a necessary precursor to them.

Almost all social networks were designed for a purpose that did not come to pass, instead it was the early users of Myspace, YouTube and Facebook who abused the platforms for their own initiatives, hacking, tweaking and reframing the systems to serve their needs, often to the bemusement of the founders.

In this respect important that we set the context for what we consider the Polkadot ecosystem - if we see the relay, parachain and on-chain orgs/collectives as a more cohesive ecology - economically, socially, and culturally interdependent, rather than as separate entities, then we are progressing to a point where there are areas where Parity/W3F/Fellowship are not the sole or primary drivers of emerging conventions (and the standards that follow).

Taking this a step further, it is existential to the future success of the polkadot protocol - and Parity/W3F/Fellowship for this to happen - indeed, the emergence of conventions within areas that these three orgs are not experts in, or even understand, is a necessary foundation for the ecosystem to flourish.

In this respect, grass-roots initiatives focused on leveraging substrate’s basic super-powers can be better placed to deliver insights, strategies and to uncover novel applications - that in turn may address fundamental challenges better than those who ‘built the system’, simply through naivete, risk taking and scratching their own itch - rather than using the systems, concepts and frameworks established by the experts.

In this regard, we return to the original purpose of Kusama - and the expect chaos motto. When I saw that presentation, the exciting thing was that it set out an intention from the creators that here is a new technology, we can’t possibly know what it will be useful for, or what eventual value proposition will be, but lets really beat it up, abuse the system and think naively about how it might be reoriented - especially in relation to the parachains / onchain orgs it enables.

It feels that Kusama is seen as less relevant by many, but if we take the counter position, that Parity/W3F/Fellowship cannot possibly see the future, are not omnipotent gods, and have not designed the perfect system we can open up the conversation around conventions into more interesting spaces.

So in summary, agree on your perspective on conventions ahead of standards - I also think this insight has the potential to reveal a lot more than maybe you originally intended if we pull the thread a little more, since before we can even move towards technical conventions (and related domains such as governance and value accrual), we should really dig into the foundations where we may have conventions that are simply there by default by virtue of being the system that bootstrapped the protocol in its earliest form.

With relays, parachains and collectives - alongside a bunch of working tech, talented people and a lack of exploration, it feels like a good time to ask what sacred cows exist, what orthodox thought may turn out to be of fleeting and consider just how this ecosystem may be recomposed to further the ideals that attracted us all - and indeed hold the keys to delivering real world impact and long term viability.