I can confirm that starting from block 396000 (round 99) onward, there are only 2500 new coins being generated, and they are being paid enterily to miners. Also, none of my wallets are downloading the receiver_99.csv file, yet they are still downloading blocks.
My guess is that miners are generating blocks while ignoring the latest receiver file, which should never happen. This might be due to a bug or a network attack. We should first discard the bug. Which is not an easy task because there are different versions of devcoin running the network, we should analyze each of them individually.
- develCuy
Now we have the issue reported in the issue queue
https://github.com/devcoin/core/issues/76So far we know that round 99 started at block 396000 on 2019-10-26 11:07:49 UTC, and that at least 4 of the 7 file custodian mirrors where well setup by 2019-10-23 17:04:45 UTC or 2 days and 18 hours in advance to round 99. Yet, we should investigate further to determine if that was enough time in advance and if not having the next receiver file prior to generating block 396000 makes the nodes to halve miners rewards and not paying shares, as proposed by Markm.
- develCuy
After studying the source code a bit more, I'm more convinced that we neither have a bug or a network attack, instead we have the miner nodes doing what they are programmed to do as Mark already said. Why are they generating only 2500 DVC per block? Because miner nodes are unable to validate that the 50% + 1 of file custodians have their mirror of receiver_99.csv in sync. For that reason they aren't downloading that file, rather they apply a programmed "fallback reduction" that lowers the rewards to 2500 DVC.
There is an assumption that we had to have the file custodians in sync 2 weeks in advance to the new round, but I can't confirm it yet. So far I read in the source code that miner nodes keep trying to download the next receiver file before, during and after discovering a new block and regular nodes do it when syncing the blockchain with the network.
Long history short, we currently have a contradiction, because the file custodians doesn't seem to be hosting the exactly same copy (bit by bit) of receiver_99.csv. At least, that is what the nodes "can say". But when I compare the files, they are exactly the same. Another option is that somehow the nodes aren't able to download all copies of the receiver file, they might be getting a truncated file or even empty. This idea is reinforced by the fact we had 3 file custodians outdated or with their server offline, the remaining 4 should be enough but it doesn't seem to be the case.
I keep debugging the source code for finding a solution. All help is welcome! We have an open issue here
https://github.com/devcoin/core/issues/76- develCuy
I have finally figured out what is happening and I'm working on a fix. We had a mixture of situations:
1. Two custodians have their servers not working properly, they tend to go offline for different issues with their infrastructure
2. Ctya's server was shutdown in the middle of round 98, rather than at the end, not allowing Traxo to take over smoothly
3. Another custodian has enforced HTTPS connections but most of the miners nodes don't support it
4. The custodians not were in sync early enough for round 99
For the reasons above, we had 4 of the 7 mirrors "out of sync" for the standard of Devcoin blockchain, that leaves only 3 working mirrors when we need at least 4, the nodes are then rejecting the receiver files. The consequence of is that miners apply a penalty know as "fallback reduction", as an incentive for solving the issue with file custodians.
Hopefully, we can finally fix this situation, but it requires consensus from all the file custodians and perhaps adding a couple mirrors as I anticipated since previous rounds. My proposal is to implement one or two automated receiver files mirrors, so that we can lower the chance for this situation to happen again. Also, we have to be more strict with file custodians, if they are unreliable, we can't hold them for a single round more, and that is what I'm going to do from now on. Along releasing the receiver file earlier, we should define a safe deadline soon.
People interested on the technical details of what originated this situation please check the issue queue:
https://github.com/devcoin/core/issues/76Last but not least, the latest version of Devcoin supports HTTPS, but it looks like the majority of miners never managed to update their nodes. I don't blame them because we never released any official stable version for a long time. So we should do it now!
Let's clean up all the mess!
- develCuy
We are very close to have round 99 fixed. If some miner is interested on collaborate, please try the new prerelease version of Devcoin 0.8.6.0 from the official repository:
https://github.com/devcoin/coreI recommend you to follow Qtera's instructions:
https://steemit.com/busy/@cpol/how-to-compile-your-very-own-devcoin-wallet-in-ubuntu-18-04AND use the following command to download the source code:
$ git clone --branch 0.8.6.0
https://github.com/devcoin/coreFINALLY, don't use it with your current wallet, please create a new separate wallet for your trial.
WARNING: Please mine ONE and only ONE block, then shutdown your node and report your findings here.
Go go Devcoin!
- develCuy
Did everything as you said here in Windows sub-system with "Ubuntu 18.04 LTS" -- no luck. Was missing files even though the instructions I followed to a T a few times in a try-fail-erase-repeat experience that took me about 3 hours last night.
So obviously somethings broken. Great version control. I don't mind waiting. I read your Steemit articles -- the Lean release process is fine by me. As are Mark's recently discussed long term viewpoint when it comes to "arbitagers" and "market makers"
Thanks for all the help I know you all are working hard on this. That's why I keep coming back here to report my trials and errors. Like you said on Steemit, develCuy, you want the public involved in the process. I am aware of the role playing or MMORPG/whatever aspects to what is said around here. Even that helps me realize computing problems in a theatrical sense. Like say, MarkM not getting his Axiom coins from Yobit but admitting a third chain may have existed unknown to him when he made AXON after AXIOM stranded some of the Milieu on Yobit's chain fork. Or like Vlad being bipolar crazy in his antics, severe DSM-IV like paranoia over Yobit and his IXC taking 6 month's to withdraw/yet MarkM having market maker posts about IXcoin recently on Yobit. I myself cannot even deposit IXC directly to Yobit, but I can do whatever I want with my Axiom. What does all of this mean to me? You guys already know - I'm just letting everyone here know I give feedback as my part of the "real human element" in the role playing game/fake-chain/game currency/multi world/process that is making one hell of a technological and humanitarian financial-like/gaming product that I absolutely value for more than its "value based fiat" (fiat being a government order not necessarily tied to money - thus it defies fiat, even if not money, and is liberation for the masses that works better than other institutional transfer of information that is valued. If also, as nova said, it is like money in some aspects - yes markm was right even if he couldn't buy a pizza with BTC as your story from the past reveals. Lastly, develCuy said on Steemit (I believe) that the intrinsic value is not the president on the bill, but the clerk who accepts your fiat money as payment. Maybe that's needed for something like STEEM and SBD, but all DevSTEEM needs to accomplish is the automatic and reliably quick use and conversion ability of DVC via an app or something in the future to join the rest of the cryptocurrency world in these type of traditional intrinsic value giving transactions. Because, this is still a simulated test, or a lean process, version control - whatever you all are doing - I see it, and I don't see it out of arrogance or for any other reason than I care. Besides Vlad crossing my personal lines on Telegram by talking about my personal information he somehow harvested both on and off social media - I can handle the monkeying around because it reveals a simulated personal experience that helps me understand the concept of a network or a system that I may not have the education to understand in technology. Keep up the good work - there is light and so much more than a fiat based currency value of these coins. In fact, I'm glad for the opportunity to snap up coins cheap because they are being manually controlled instead of part of a fully automated type book/machinated market making like environment on exchanges. Markm's posts this week helped me understand how the books are played for the ancient coins, and although he does build both sides of the book admittedly, there's arbitage in the fact that a coin right now that I can afford is not tied to $ in any type of fiat that would in all likelihood be on an exchange where I could not realistically "beat the machine" in trading for profit. In the future it'll pay off to have that buffered space, delayed fork space, fake-chain space/lean process for long term value in ways I appreciate, somewhat grasp, and probably don't even realize.
Please keep the good info flowing. Someone's listening from the "real world" of the mass public. I appreciate the game because of the time involved in the effort on the developer side to pull it off so exquisitely. A secure chain is my only interest besides dev time spent on the project and seemingly they go hand in hand. It's a long term view - just like the original AXIOM "AxiomCryptocurrency" account described before it vanished in 2015 - they still have a say in full control over many coins involved in the Galactic Milieu, STELLAR/HORIZON listings, etc. Maybe one day the action potentials sent out via the neurological fibers, axom's if you will, that began at the source, collective minded people who founded this project will eventually "make you realize" you're part of the founding project, one of 12 "shares" of DeVCoin even that came from one place and found its way back to the other end.
Go team DevSTEEM.
