Introducing the Swiss Subnet: A Secure and Compliant Blockchain Infrastructure Leveraging the Internet Computer Protocol

General
,
73

Waterneuron rejected this proposal - they are clear against good for ICP

74

Adam rejected it lol

75

I already said this.

I know WTN is controlled so I just vote no and sometimes yes depending on how i feel. If the vote mattered Iā€™d give it more than 1ms of thought.

Me voting on NNS : 8%
Me voting on WaterNeuron : 7% of 1% where the 7% means nothing.

1 Like
76

In this case your vp is a major part of the reason it got rejected lol.

77

Yup sure, just like when you voted with 67.007%

1 Like
78

I wonder if you know this but, between you Dfinity and the ica neurons you three own ~ 39% of vp on the nns.

Why donā€™t we ever discuss this threat vector?

But in the end you canā€™t sit here and spin narratives after you voted to reject the proposal and weā€™re a key vote to push the dao to vote no. Next time vote responsibly then you can sit there and ask questions to the other members who voted no on it.

79

This is so irresponsible. You can do better. Just follow DFINITY if you arenā€™t going to take your governance responsibility seriously.

Here is their WTN neuron ID: 0c9288198b6dc21da28113a842f1c6cb4248eb54273f1b2702c12afbeb5b416f.

Here is proof: OpenChat

1 Like
81

Sounds like a projection after being exposed for the way you voted in wtn for this.

2 Likes
82

So you guys @Swiss_Subnet are working with DFINITY Foundation and their UTOPIA project? They are okay with you guys creating a brand around ICP subnets hosted in switzerland?

I still do not understand. Your responses lack clarity

83

In principle I think rented subnets are a good idea. We could do with a more diverse set of subnets, where each one meets nuanced criteria (such as regulatory compliance, but other things come to mind such as building subnets where all the node providers have a certain amount of staked tokens, or much larger or smaller subnets, subnets specialised for running certain workloads like LLMs etc.).

Itā€™s already the case that new subnets can be created by the community. I think the trouble is that thereā€™s typicall not a good way for the voting community to gauge whether or not thereā€™s enough demand for a certain type of subnet (will demand be enough to pay for running the nodes).

I gather that the point of subnet rental is that the proposer for the new subnet fronts the ICP for rewarding the nodes for roughly 6 months of compute, and keeps topping it up. This way it somewhat removes the concern about there not being enough demand, as the proposer has put their money where their mouth is.

I think itā€™ll be interesting to see how this goes. If there are pitfalls they can always be improved. If it works well I have a feeling this could be a great feature for the IC.

2 Likes
84

@volibear , when we developed the concept of a private, geographically defined subnet, we consulted closely with DFINITY, who supported us in preparing the core protocol infrastructure and the subnet rental canister.

While it may appear similar on the surface, the Swiss Subnet is entirely separate from and in no way connected to the UTOPIA project.

3 Likes
85

Hi @Swiss_Subnet, could you elaborate a little regarding why youā€™re choosing to rent a subnet rather than making the case for a permanent Swiss subnet (same sort of approach as the European subnet and proposed US subnet)?

86

Thanks for the question @Lorimer I checked with the team and have this message to relay:

The European subnet canā€™t deliver true GDPR compliance ā€” no real support for Article 17 (right to be forgotten), Article 20 (data portability), or Swiss FADP standards.
Itā€™s just a basic 13-node parking lot, with no advanced compliance or innovation.
Meanwhile, the Swiss Subnet is a true data fortress: private, permissioned, with 100% encryption and nodes spread across 10 Swiss cantons and Liechtenstein, including mountain bunkers for unmatched security.
The EU subnet is only a temporary facade. The real challenge is actual compliance ā€” and good luck achieving that there.
Itā€™s only a matter of time before serious projects move to the Swiss Subnet to avoid legal and compliance risks.

87

Thanks for the additional information.

Could you elaborate? The European subnet ensures data is processed within the EU. Other than that, dapps can prove compliance by open sourcing reproducible builds.

Iā€™m guessing what youā€™re saying is that not all dapps on the European subnet will necessarily be fully compliant, and compliance cannot be forced on any of the dapps. Is that your meaning?

Are you able to share more about how compliance will be enforced on the Swiss subnet?

88

apologies for the delayed response. As the CTO I need to go back to the team to get answers about non-tech questions. Weā€™re looking into speeding things up. This is what I learned:

One of the key areas weā€™re exploring with this subject is Compliance as a Service (CaaS). Through this approach, we would continuously update privacy policies and proactively monitor new legislation or regulations, so that small and medium-sized enterprises (SMEs) would not need to maintain their own dedicated privacy compliance office. We will ensure full compliance with both FADP and GDPR, on both sides of the Rhine.

89

No worries. Could you share more about compliance with respect to the canister-in-canister approach?

Iā€™m only asking lots of questions because Iā€™m fascinated and would like to know more, given that this is the first example of a rented subnet.

As a ā€˜permissionedā€™ subnet, can I ask who has the say about who deploys to it, and who has control over the canisters?

90

Access to deploy canisters into rented subnets is fully controlled by the renting principal, this is how Dfinity designed it.

We intend to enable a self service capability which might work like this:

  1. develop locally using dfx
  2. commit working code to Github
  3. create principal using dfx (if not already done)
  4. use Swiss Subnetā€™s website to configure your project, providing
    • Github link
    • principal
    • domain (optional)
  5. Swiss Subnet will then
    • compile the code from Github into a wasm
    • deploy a new canister with the wasm
    • add the provided principal
    • remove its own principal (optional, tbd)
  6. tadah - the app is live, and can be maintained by the developer going forward

Let me know if you see anything that would be particularly valuable in this context, thanks!

1 Like
91

Thanks @Swiss_Subnet. What sort of trust assumptions need to be made by those who deploy to the subnet? I see that lots of effort is being put into ensuring physical decentralisation, in terms of the nodes running the canisters. Theoretically speaking, how many people would need to come to consensus when/if it comes to forcefully uninstalling a canister from the subnet (such as due to non-compliance)?

92

We will likely launch early access with a simple, centralized control model, and then add more options over time.

Using Orbit or similar tools we can configure almost every possible setup.

But itā€™s clear that we need to retain the possibility to stop or uninstall a canister, one way or other, to prevent a negative impact on our other clients, in case of a buggy or hacked canister consuming too many resources.

Developers who need more control, and isolation from others, should talk to us about getting their own instance of Utopia.

1 Like