Topic: [ANN] Triangles [TRI] Now Stable & Working Anonymous Cloak over TOR - ON BITTREX - page 18. (Read 90226 times)

We might actually be okay now.  Chain movement is much better.

Anyone who wants to test TRI with us but has no TRI, sync your chain to the current block, and I'll send you 5 TRI.  PM me your TRI address.

Everyone with TRI, update to latest client.  I think we finally got it.

Wallets:

First make a backup of your wallet. Everyone should do that from time to time. Copy the "%appdata%\roaming\triangles" folder (Windows) to a safe place. Done.

WINDOWS wallet, v 3.4.3.1 "Cheops":

https://mega.co.nz/#!dUkW1TBY!tt7a9rVi4qhdkVi-LiT2MXdBTnV_XtLGmq3KGLlptZA


Included is a folder with a bootstrap.dat file with 10881 blocks (if you wallet isn't synced to 10881 blocks before you update OR IF YOU PREVIOUSLY USED YESTERDAYS RELEASE - see the included readme!). Also a folder with a peers.dat file (if you have trouble finding peers - see the readme!). This bootstrap.dat file is also useful if u just downloaded the wallet: create a folder "triangles" in %appdata%\roaming and place the bootstrap.dat into that folder - then start the wallet.

If u have coins in your wallet, please start your wallet  with the "-connect=lkta5xj5nlnxxmrt.onion:24112" switch via commandline the first time after you updated your wallet. Its a non staking node and syncing to that node first prevents initial forking. Wouldn't  be much of an issue if it happens though. This also helps if u have trouble getting a connection despite using the peers.dat (see above).

\o/

i  just made a bootstrap @ 11222 - so we can use that tomorrow in the worst case.

So much for today.
Network is alive, forks need to and will be adressed. Also 0 reward blocks and fast "Minting-Chains".

Hey thats my chain too!

no:




getblockhash 11072

0ac5afcdbf8a47eb6ddd6d0d31e8f01a093f79dd4de6e8edc1250f35322926f8


getblockhash 11200

ece781fdff375c95a692d8551d809fe20dc15ac72478a161d1f2f4e5d8a591fd

thats my chain.

you still have the same result for getblockhash?

What a forky mess. I will definitely address that with a min minting interval tomorrow.


U should keep a wallet dat from block 10881 backed up. if u resync use that wallet.dat with the bootstrap.

if u dont have such wallet.dat: create and empty wallet with only bootstrap.dat in %appdata%\roaming\triangles, when its started import your wallets privkeys (read last pages). close wallet and safe your wallet.dat - its your clean 10881 blocks wallet from now on. then sync to the network.

if u used "-connect" switch, restart ur wallet without it once you are sync - otherwise your wallet wont connect to other nodes!

I would say we let the network run until tomorrow so everyone can play with it etc pp. Dont do any reasonable transaction or trades now!


I will fix that tomorrow and also select a suitable blockchain then - one which syncs from 10881 to the end without errors. Shouldnt really matter which one we/I select.
Expect the fix in approx. 24 hours, lets have fun with this playground until then.



edit: currently  both my wallets are synced to 11209. they did that on their own with bootstrap.dat.

Same here. tried to get back onto right fork and now stuck on 11051
edit: stcuk at same point on both wallets. got my win8 up and running without crash. just catching up with you guys... long day at work

let me check, maybe i forked. give me a few mins.

I'm sticking on 11051 too.
I've done everything including starting from scratch.  

4 connections 80 blocks remaining

Were you able to connect to wurst's fork or were you getting hung up on 11051 too?

I started with no wallet and only bootstrap.dat in the triangles Appdata folder and connected to your node, but there's still something about block 11051 that will not let me go forward.  I tried restarting it a few times, but it still wouldn't sync.  I guess we'll just have a couple of chains this weekend and see how many more we have on Monday.

Edit: Seems like there are already a few chains. 

Sure.  Should everyone try to get on the 608ab1d91366798abe291d7c103bb6d354c7cd593000aad5e1e57accdf9dff1e fork?

edit: Nevermind, read your post.  The answer is yes.

anyways....lets see how it evolves the next 12h.

please report how often u forked afterwards - maybe with some details like approx coins. if that isnt to personal.

remember:

Recovering from a fork:

Delete both blockchain files, drop bootstrap.dat into that folder, start wallet with "-connect lkta5xj5nlnxxmrt.onion".

As long as u have the same blockheight as that client youre fine (you can check via "getpeerinfo" - or if not there - search for the address in your debug.log or use the connect switch, see above).

We will decide tomorrow how and if this needs to be addressed.

Awesome. See above! Same thought. although a fixed minimum blocktime would be dangerous: different clients would try to mint a block the second those 2 mins are over - bi forking risk.  better do it with an interval per client, maybe even with some random part. 2.5 +/- 0.5 minutes or something.

I think the problem might be that we don't have high enough network difficulty so we're pumping out way too many blocks per minute when it's a person's turn to stake.  If we don't stop forking over the weekend, would making some simple change like going to 2 minute blocks reduce that?

a-ha i was using the old -connect

It takes two to fork. I currently have 6 nodes "on my my side", so you better join me.

Delete blockchain files, put bootstrap into folder, start wallet with "-connect lkta5xj5nlnxxmrt.onion" - thats my sync node".

Reason for the fork happening: Large part of the weight distributed over few nodes. Will stabilize once more nodes come online.

For now, the right fork is where I am on (so you can easily sync to it with "-connect lkta5xj5nlnxxmrt.onion" and bootstrap in case u forked). If bittrex is online, bittrex is the right fork. Currently I'm thinking - if those forks continue to happen -  it might be wise to reduce the amount of blocks generated over time per wallet:

1st thought: 3mins min between blocks per wallet/node/address.

Effects/Facts:

- Easy fix
- Blockchain is working now, so we can do updates with proper forking at given block height without even interrupting the network - with enough time for everybody to update the client. As it is meant to be done.
- Forking risk would be greatly reduced
- The network would be a little slower (wouldnt matter with only 4 confirms needed - if u ask me).
- PoS rewards per block mintet gets bigger
- 0.00 reward blocks wouldl be greatly reduced up to almost no more 0.00 blocks.
- Chain-mining (where you mine a sequence of blocks) wont occur any more.

Any thoughts on this?

I did too.  I'm bootstrapping now to see if it will sync right.

edit: It didn't.  I just went back to the 05adfeb1a920fb5e1a8dba6b6f52aad6e690e5e35cfd3ee28d1e8c889ddfeb8c fork for now.

Wurst, what's wrong with our fork?  Can you get on that one?

I had that too lol.

So I should bootstrap?

getblockhash 11072
05adfeb1a920fb5e1a8dba6b6f52aad6e690e5e35cfd3ee28d1e8c889ddfeb8c
thats what im getting.