Parano1d ① Proof-native Layer 1 ordered by PoW

[Hashrateoptions]

! I found the exact cause. This was my deployment mistake.
I deployed the seed nodes using a local build whose embedded HistoryStep matrix whitelist differed from the public v0.1.0 binaries. Public clients correctly rejected the proofs produced by those nodes.
I have stopped the network. I am replacing every seed with the exact published v0.1.0 binary and restarting the chain from genesis.
Please close the client for now. Your downloaded v0.1.0 package is correct, so you will not need to download it again. I will post here when the network is ready; then simply reopen the client.

I mined 1 block. do I need to del data dir?

Yes. Since your client actually mined and accepted a block on the stopped chain, close the application and remove the chain data before restarting.
I will post the exact reset instructions for each OS together with the network-ready notice.

Do you plan to do it today ?
As really I dont get it setting two seed nodes taking so long ?

[tomassamot]

There is a bug regarding the maximum number of supported threads: with my 7742 CPU (64 cores, 128 threads), I can only select a maximum of 64 threads; similarly, with the 9654 CPU (96 cores, 192 threads), the maximum selectable limit is also just 64 threads.

[Hashrateoptions]

There is a bug regarding the maximum number of supported threads: with my 7742 CPU (64 cores, 128 threads), I can only select a maximum of 64 threads; similarly, with the 9654 CPU (96 cores, 192 threads), the maximum selectable limit is also just 64 threads.

It is normal on my side same propably leaving some resources for safety

[IgnotusNemo]

Update: I found the full cause.

The public v0.1.0 binaries were built by CI with the pre-mainnet test HistoryStep pack, while my locally tested build contained the final mainnet pack. The source code, genesis and mainnet parameters are correct, but the published v0.1.0 binaries are not.

I have hidden v0.1.0 and stopped all seed nodes. Do not run or mine with v0.1.0.

v0.1.1 is now being built for all supported platforms. The complete five-platform rebuild may take up to two hours.

When I confirm that v0.1.1 is ready:

1. Close v0.1.0 completely.

2. Delete the entire Parano1d data directory. Uninstalling v0.1.0 is not required and would not remove this directory.

Linux:
~/.parano1d/data

macOS:
/Users/YourName/.parano1d/data

Windows:
%USERPROFILE%\.parano1d\data

For example:
C:\Users\YourName\.parano1d\data

3. Install v0.1.1 over v0.1.0 and start it normally. If macOS asks whether to replace the existing application, choose Replace. The data directory will be recreated automatically.

Please wait for my confirmation before restarting. I will publish the links only after all binaries are available and the seed network has been restarted and verified.

This was my mistake. Sorry for the failed start.

[Arttemkaaa]

Update: I found the full cause.

The public v0.1.0 binaries were built by CI with the pre-mainnet test HistoryStep pack, while my locally tested build contained the final mainnet pack. The source code, genesis and mainnet parameters are correct, but the published v0.1.0 binaries are not.

I have hidden v0.1.0 and stopped all seed nodes. Do not run or mine with v0.1.0.

v0.1.1 is now being built for all supported platforms. The complete five-platform rebuild may take up to two hours.

When I confirm that v0.1.1 is ready:

1. Close v0.1.0.

2. Delete the entire Parano1d data directory. Nothing from the stopped chain needs to be retained.

Linux:
~/.parano1d/data

macOS:
 /Users/YourName/.parano1d/data

Windows:
%USERPROFILE%\.parano1d\data

For example:
C:\Users\YourName\.parano1d\data

3. Download, install v0.1.1 and start it normally. The directory will be recreated automatically.

Please wait for my confirmation before restarting. I will publish the links only after all binaries are available and the seed network has been restarted and verified.

This was my mistake. Sorry for the failed start.

How long?

[alt_x]

This was my mistake. Sorry for the failed start.

How long?

By amount of files to build, my guess is tomorrow

[Hashrateoptions]

There was 2nd release now it is hidden bootstrap nodes was working now again down it is really a bad start and acting
Gui wallet works 2 peers but just when you run mining one peer is missing.
In my opinion network is isolated

[IgnotusNemo]

Please do not run either withdrawn release until I explicitly confirm the replacement. Running them only adds noise to the live diagnostics.
I will post again only after the replacement has passed the full local and multi-platform release checks.

[tomassamot]

Frankly speaking, a 15-second block time makes the network highly susceptible to forking. If a surge of forks occurs, chain reorganization requires extremely high processing speeds; otherwise, if the reorganization takes longer than 15 seconds, the network enters a state where consolidation becomes persistently difficult, making it impossible to reach consensus on a long chain.

[Hashrateoptions]

Please do not run either withdrawn release until I explicitly confirm the replacement. Running them only adds noise to the live diagnostics.
I will post again only after the replacement has passed the full local and multi-platform release checks.

Ok we will wait till you test and make premine time waste here it is 3 am on my side wallets have 3 peers but mining do not work also you run bootstrapnodes and then switch them down becasue of that network have a lot of forks as people use isolated nodes and that is main issue. Going sleep wasted half of my night Good Night.

I was thinking it is mainnet and have all checks but seems we are on testnet

[IgnotusNemo]

Frankly speaking, a 15-second block time makes the network highly susceptible to forking. If a surge of forks occurs, chain reorganization requires extremely high processing speeds; otherwise, if the reorganization takes longer than 15 seconds, the network enters a state where consolidation becomes persistently difficult, making it impossible to reach consensus on a long chain.

A reorganization does not have to finish within one 15-second block interval. What matters is whether nodes can verify and apply blocks faster than they arrive over time.
Parano1d does not replay transactions or rebuild proofs during a reorg. It verifies each {block, HistoryStep} bundle and only rolls back or reapplies the bounded 18-block suffix. This work does not grow with chain height.
A 15-second target still makes propagation and stale-block rates important, but crossing one block interval does not by itself prevent consensus.

[IgnotusNemo]

I was thinking it is mainnet and have all checks but seems we are on testnet

the protocols that changed crypto were not born battle-tested. mainnet is what made them battle-tested.

[alt_x]

It is late night here.

what about if make it fair, relaunch tomorrow?