by

Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Activity
  • Sep 20 14:56
    sudachen synchronize #2143
  • Sep 20 14:56

    sudachen on elk-fix

    storageclass fix (compare)

  • Sep 20 01:01
    lrettig synchronize #2060
  • Sep 20 01:01

    lrettig on github-actions

    Update testnet join instruction… Merge branch 'develop' into git… (compare)

  • Sep 19 21:35
  • Sep 19 08:01
    avive edited #2145
  • Sep 19 07:55
    avive edited #2145
  • Sep 19 07:54
    avive edited #2145
  • Sep 19 07:43
    avive labeled #2145
  • Sep 19 07:43
    avive opened #2145
  • Sep 18 21:51

    bors[bot] on lrettig-patch-1

    (compare)

  • Sep 18 21:51
    bors[bot] commented #2144
  • Sep 18 21:51
    bors[bot] closed #2144
  • Sep 18 21:51
    bors[bot] edited #2144
  • Sep 18 21:50

    bors[bot] on develop

    Update testnet join instruction… (compare)

  • Sep 18 20:52

    bors[bot] on staging.tmp

    (compare)

  • Sep 18 20:52

    bors[bot] on staging

    Update testnet join instruction… (compare)

  • Sep 18 20:52

    bors[bot] on staging-squash-merge.tmp

    (compare)

  • Sep 18 20:52

    bors[bot] on staging-squash-merge.tmp

    Update testnet join instructions Use raw link for config file [ci skip][skip ci][skip netlify… (compare)

  • Sep 18 20:52

    bors[bot] on staging-squash-merge.tmp

    (compare)

cryptohuang
@cryptohuang
if a prover has X data of out Y, then they will win X/Y times, if only one challenge
in your case, if prover has X data out of Y, then they will win (X/Y)^c, where c is number of challenges
but (X/Y)^c is overkill since X/Y already fits a security of a system like yours
probability of winning is proportional to space proven (X/Y)
what do you think @y0sher
rehs0y
@y0sher
@cryptohuang Definition 4.1 in the Spacemesh protocol paper defines our soundness. An adversary that does not store the entire data should win with negligible probability. if the probability of winning remains high (though less than 1), then the adversary can change a single label to have a totally new challenge, thus being able to "amplify" his probability of success.
yaelmhoffman
@yaelmhoffman
Hey everyone! We just added a bunch of protocol docs to the protocol repo. Come check them out here: https://github.com/spacemeshos/protocol
cryptohuang
@cryptohuang
AH I understand now - this is a fiat-shamir issue.
Thank you it makes a lot of sense
Every new amplification reduces the soundness
really good!
what is number of labels to challenge today? In code I see 100, where is it from?
rehs0y
@y0sher
@cryptohuang you are right, the current code uses 100, pending mainnet and more changes in security proofs it might change.
cryptohuang
@cryptohuang
how do you to derive that?
Noam Nelke
@noamnelke
@cryptohuang derive what? You mean how the Fiat-Shamir is implemented? Or where we get the config that says that 100 Merkle-leaves are required?
cryptohuang
@cryptohuang
sorry, how do you derive the number of challenges?
new question and thank you for answering
cryptohuang
@cryptohuang
why does miner need to mature power?
Noam Nelke
@noamnelke

how do you derive the number of challenges

The number of Fiat-Shamir challenges is not derived, it's pre-configured. The same configuration is used for generating proofs and validating them. In other words, we validate that the proof includes the number of challenges that is configured.

why does miner need to mature power?

I'm not sure I understand this question. What do you mean by "mature"? The way the protocol works is that miners have to create an activation transaction that proves that they have exerted effort (committed space over time / spent the cost of regenerating their space). The published activation transaction makes them eligible to participate in the protocol (produce blocks and participate in the hare protocol).

Vaibhav Agarwal
@vaibhavbarmy
Smesher app is not detecting external seagate hardisk? Any thoughts on this
Aviv Eyal
@avive
@vaibhavbarmy The testnet is not out yet - we expect to make it available on March 1st. The app is basic first release of the platform and doesn't have all features yet such as variable space commitment sizes or external drives. These features will be added quickly after the first release - they are on our roadmap
Vaibhav Agarwal
@vaibhavbarmy
Awesome I love spacemesh, any idea as to when smesh (SMH) will available to buy on exchanges?
Noam Nelke
@noamnelke
@vaibhavbarmy as @avive said, the testnet is not out yet, not to mention mainnet, so SMH doesn't exist yet. Once it exists, we still won't be able to predict when exchanges decide to list it, but we hope that enthusiastic users like you will mine it well before it's tradeable.
Vaibhav Agarwal
@vaibhavbarmy
@noamnelke thanks for your detailed reply :)
cryptohuang
@cryptohuang
@noamnelke the word mature is used in your protocol pdf document

from document: "
The first k ATXs are considered immature; every succeeding ATX for the same identity is mature.

A valid, mature ATX activates the associated identity for a single epoch. A node ID is considered
“active” starting at the first epoch boundary following the publication of the ATX (e.g., if an epoch is
∆epoch = 1024 layers, and a mature ATX was published in layer 1700, the ID becomes active at layer
2048, and remains active until layer 3072)."

@noamnelke: how do you choose how many challenges to do? sorry
Noam Nelke
@noamnelke
@cryptohuang ATX maturity was a concept explored by our research team and eventually deemed unneeded (so essentially k=0). The original idea was to make it more expensive to create an active identity. This is ultimately achieved by making it more expensive to generate the init data, instead of requiring users to wait for several weeks before they start mining. It also simplifies the protocol, since all ATXs are considered mature.

I already answered your question about the challenges:

The number of Fiat-Shamir challenges is not derived, it's pre-configured. The same configuration is used for generating proofs and validating them. In other words, we validate that the proof includes the number of challenges that is configured.

If this is not what you meant, please explain what you did mean

If you mean how we choose what to configure, this is a security parameter that is still not final and will be finalized closer to the release of our mainnet based on the protocol at that time (it's still being tweaked) and security requirements that will be met with minimal bandwidth and storage space cost.
cryptohuang
@cryptohuang
thank you
this answers
mathoticus
@mathoticus

Are there succint ways to check post pow? Like low resource or logarithmic work verification?

Flyclient paper says it would be applicable but curious how you all see it in practice

Noam Nelke
@noamnelke
Hey @mathoticus , that’s a great question and we have some innovative ideas that may reduce the proof size and validation effort considerably, but they’re still a little raw and not ready to share yet, so.... please stay tuned 😉
mathoticus
@mathoticus
ok, thank you @noamnelke !
Aviv Eyal
@avive
Hi all - we released Tweedledee - an open Spacemesh testnet! Please join the testnet discord channel here: https://discordapp.com/channels/623195163510046732/623195511331225620
There are basically 2 ways to join - using the Spacemesh desktop App for linux / OS X / Windows or by building go-spacemesh and a cli wallet from source code. See: https://github.com/spacemeshos/go-spacemesh#joining-a-testnet-with-mining.
There are also pre-compiled go-spacemesh and CLIWallet binaries you can grab from these github releases: https://github.com/spacemeshos/CLIWallet/releases/tag/v0.0.1 and https://github.com/spacemeshos/go-spacemesh/releases/tag/v0.1.3
To join using the desktop App head over here: https://spacemesh.io/testnet/
See you on the testnet!
Vaibhav Agarwal
@vaibhavbarmy
When is the mainnet planned?
yaelmhoffman
@yaelmhoffman
@vaibhavbarmy We have several releases planned over the next 12 months, with mainnet coming towards the end of that period.
Alon Askal
@AlonAskal_twitter

Hi everyone!

👾 We're moving our community discussions to Discord! 👾

For those wishing to be involved in the discussion on Discord ->
🖖🏾 https://bit.ly/dSpacemesh 🖖🏾

Looking forward to seeing you on Discord!

Sincerely,
Team Spacemesh
🤖

Alon Askal
@AlonAskal_twitter

👾REMINDER👾

For those wishing to be involved in the discussions, joins us on Discord ->
🖖🏾 https://bit.ly/dSpacemesh 🖖🏾

Alon Askal
@AlonAskal_twitter
        🖖🏾Dear Smeshers! 

█ ✪ █▓▓▓▓▓▓▓▓▓▓▓█ ✪ █

Make sure to follow @TeamSpacemesh on Twitter to stay updated:
https://bit.ly/2Jpr2hq

Chris Spannos
@cspannos
Hi Spacemesh team! I'm Chris. I'm super excited about Spacemesh. I'm currently a fellow at Insight Data Science in their decentralized consensus program. I'm working on a project that assesses differences between API protocols applied across various DLT protocols. I'm curious, why does the Spacemesh API use protobuf and gRPC for both internal and external APIs? I understand that protobuf would offer performance gains internally, so wouldn't suggest any change there. However, as an opensource project Spacemesh wants to attract as many developers and applications as possible. So why not for your public facing API use OpenAPI instead since it's among the most widely used specifications? Wouldn't this help onboard developers to your protocol? For these reasons, I'm interested, and investigating, converting your protobuf / gRPC API to an OpenAPI description model, which I think would be useful for tooling, documentation and automation. Would you find this useful? If so, I'll probably have some questions for the team, as this is new to me. I'd love to know the trade-offs you've considered for your choice of API protocols. You're guidance would be much appreciated!
rehs0y
@y0sher
Hi @cspannos ! And thank you for your interest in our project. We choose gRPC and protobuf for the external API because it has very good support for go and other languages and very nice tooling. This tooling includes a middleware that exposes a swagger (OpenAPI) mirror of the gRPC API implemented on a different port, Which we use. I hope this answers your questions feel free to ask more 😉
Chris Spannos
@cspannos
Hi @y0sher , Many thanks for providing that context! That makes sense. Is the middleware you mention grpc-gateway?
rehs0y
@y0sher
Yes, exactly.
Chris Spannos
@cspannos
Cool. Thx!