• Forum has been upgraded, all links, images, etc are as they were. Please see Official Announcements for more information

V12.1 Testnet Launch Thread

Status
Not open for further replies.
Yes, testnet was reset, but we didn't create a new genesis block - it is still one mined at 2014-01-25 --> https://test.explorer.dash.org/bloc...75e2ec92894837288a481e5c005f6563d91623bf8bc2c

Normally the client should find peers by itself after some minutes, if you still have problems you can add the masternodes available: https://test.dashninja.pl/masternodes.html#mnlistdetail

Thanks for the info!

Using those masternodes as "addnode" fixed my problem!

I'm now up and running as a liquidity provider!
 
I'm doing some mixing on several wallets (Windows 10, 64bit, v0.12.1.0-16671cd) and i still got some unconfirmed transactions, stagnating the whole mixing process

InRQXLK.jpg


I noticed this behaviour on Mainnet too as i was mixing there a few weeks ago too .. it just took a lot of zipwallettxes to get through those stuck transactions.
Only the Privatesend Denominate transactions are sometimes just not confirming, the other kind of transactions (colleteral / selfpayment) are confirming just fine.
I think it has to do with Privatesend Denominate transactions not having transaction fees and are therefore sometimes overlooked by miners ? (eventhough they have a prioritize condition)

Status: 0/unconfirmed
Date: 24-5-2016 18:29
Debit: -0.10000100 tDASH
Debit: -0.10000100 tDASH
Debit: -0.10000100 tDASH
Debit: -0.10000100 tDASH
Debit: -0.10000100 tDASH
Debit: -0.10000100 tDASH
Credit: 0.10000100 tDASH
Credit: 0.10000100 tDASH
Credit: 0.10000100 tDASH
Credit: 0.10000100 tDASH
Credit: 0.10000100 tDASH
Credit: 0.10000100 tDASH
Net amount: 0.00000000 tDASH
Transaction ID: e19938e5a81ad80d27b82fae39a553a09fae2aa6ee6909315c7a26f71793ae05-000

Perhaps we should consider implementing small transaction fees to Privatesend Denominate transactions....
 
@qwizzie - I'm not entire sure, but unconfirmed-tx by miners has to do with block getting full. So increasing fees will prioritised them into the blocks when the mempool is full? These are unconfirmed propagated tx, and why -zapwallettxes gets rid of them. If this is right I deserve gold star.

If this is a propagation issue... Is it possible to create a PrivateSend style system, where the mix txs are broadcast to a few MN's for better eficiency? Does this make sense at all ?
 
@qwizzie - I'm not entire sure, but unconfirmed-tx by miners has to do with block getting full. So increasing fees will prioritised them into the blocks when the mempool is full? These are unconfirmed propagated tx, and why -zapwallettxes gets rid of them. If this is right I deserve gold star.

If this is a propagation issue... Is it possible to create a PrivateSend style system, where the mix txs are broadcast to a few MN's for better eficiency? Does this make sense at all ?


Full blocks at dash testnet? WTF?
 
Full blocks at dash testnet? WTF?

No dude, I'm not saying that, I'm answering to qwizzie's suggestion, implying that it's not the case, and that fee's won't fix this :p (but wtf do I really know about code?)
 
No dude, I'm not saying that, I'm answering to qwizzie's suggestion, implying that it's not the case, and that fee's won't fix this :p (but wtf do I really know about code?)

Okay, I am testing privatesend and I am getting conflicted transactions. What is their purpose actualy?
 
Okay, I am testing privatesend and I am getting conflicted transactions. What is their purpose actualy?

I think at this point we are still at phase 1 and are working on Message Propagation

Testnet Agenda

We’re opening testing to the public and are going to be testing everything in the following order:

  • Phase 1: Configuration / Syncing / Message Propagation
  • Phase 2: Normal Network Functionality (DS/IX)
  • Phase 3: Masternode Trezor Support
  • Phase 4: Enhanced Budget System | Sentinel
  • Phase 5: Pre-launch
  • Phase 6: Launch
 
@TanteStefana I doubt daemon can have or needs some console interaction - it's daemon after all. More than that, I'd like to note that it doesn't require rpcuser and rpcpassword anymore so non-mn users can start using daemon without any setup and for mn users it's as simple as adding 2 lines to a text file - "masternodeprivkey=blablabla" and "masternode=1". Not such a difficult task imo. Next, I think hot masternode setup should go away completely one day and the earlier the better, there should be no setup like that where nodes with $$ are online 24/7 imo. So I'd actually cut that out from our code rather than start wrapping things to make it nice and easy. As for gui part for masternode.conf and hotcold setup - I agree that's a field for improvement :)

Cool! Thanks so much for looking! I had no idea that the dash.conf was simplified so much!
 
Mix result - 3 wallets, 2k tDash @ 8 rounds, dodgy internet, laptop from work to home.
8 hours total mix time
Restarted QT zapping tx anyway

Wallet 1 - 19%
Wallet 2 - 23%
Wallet 3 - 42%

wallet 3 seems legit

eteOv1A.jpg
 
Last edited:
Mix result - 3 wallets, 2k tDash @ 8 rounds, dodgy internet, laptop from work to home.
8 hours total mix time

Tried with one wallet, 1000 / 2 rounds, after 120 mins it was 19%, stopped there.
 
Weird thing happened.

So I suffer from the known lagging-after-sleep-OS bug with VMware Ubuntu 15. So I quit the QTs and rebooted. Never had a problem, reboot just makes it snappy again.
When trying to open the QT with -datadir= ... I got Permission Denied from terminal ... on all 3 mix wallets. I had to sudo the first one, and then the other one's were ok. Nevertheless...

3pYAq8E.png


Feature request> Is it possible to create a sub-tab, or check-mark to auto-hide mix txs? It would be great so it doesn't clutter the transaction tab
 

I cannot reproduce this one, neither on Linux nor on Windows 8.1 / 10.

Maybe you have corrupt settings. Please go to Settings -> Options and click on "Reset Options".

After that restart your wallet, switch back to the traditional theme and report back.
 
I was under the impression that 12.1 would feature some new hyper-mixing breakthroughs making it much faster. Maybe not enough other mixers?
 
I was under the impression that 12.1 would feature some new hyper-mixing breakthroughs making it much faster. Maybe not enough other mixers?

I kindda feel like privatesend doesn't work and is not production ready due to the high mixing times (many hours to a few days), this makes it impractical for the mainstream user (who's attention span is measured in minutes). I remember Evan said he was gonna work on a new solution that was automatic several months back, but I think that got dropped. Still, it's our weakest link and perhaps its time to move on from the anonymity concept and just focus on being digital cash. At least until we can implement a better solution than coinjoin. Coinjoin is only useful for trolls to attack us and I have to say I agree with them, it's not practical as it stands today.

Pablo.
 
I kindda feel like privatesend doesn't work and is not production ready due to the high mixing times (many hours to a few days), this makes it impractical for the mainstream user (who's attention span is measured in minutes). I remember Evan said he was gonna work on a new solution that was automatic several months back, but I think that got dropped. Still, it's our weakest link and perhaps its time to move on from the anonymity concept and just focus on being digital cash. At least until we can implement a better solution than coinjoin. Coinjoin is only useful for trolls to attack us and I have to say I agree with them, it's not practical as it stands today.

Pablo.
I think the slow speed on testnet is due to the low number of concurent users mixing. The real test will be on mainnet where there will be orders of magnitude more people mixing. "move on from anonymity"? No chance of that. Even as is, it's practical. Digital cash = privatesend.
 
I think the slow speed on testnet is due to the low number of concurent users mixing. The real test will be on mainnet where there will be orders of magnitude more people mixing. "move on from anonymity"? No chance of that. Even as is, it's practical. Digital cash = privatesend.

I'm sure there is a logical reason for why it's so slow, but the fact remains it's impractical. We should remove it, or at least stop advertising it until it works at a level something like Instantsend, which is a charm. I look at privatesent like a 20 kilo weight on an otherwise streamlined machine, its just not there yet.

Again, that's just my opinion :D. I do wish we could come up with a more efficient solution but I'm not the one to do it unfortunately :).

Pablo.
 
Quick note for those who is trying to mix and measure results: please wait for most of the masternodes to update to 70200 first. Few nasty bugs were squashed and hopefully mixing speed should improve a bit. Also try -darksendmultisession flag (or -privatesendmultisession after this https://github.com/dashpay/dash/pull/804 gets merged) and see how it goes.
 
Status
Not open for further replies.
Back
Top