Skip to content

devlog.md

Latest commit

 

History

History
209 lines (171 loc) · 7.46 KB

devlog.md

File metadata and controls

209 lines (171 loc) · 7.46 KB

Dev Log

June 25, 2025

Problems

  • Current self.chain stores all blocks in memory → leads to scaling issues on long run.
  • Old test case failed after changing gas rules → now transactions with gasLimit = 0 are considered valid.
  • The name MMB (old brand) feels unprofessional and lacks clarity.
  • Current self.chain are used everywhere because in many cases I need to create a dummy chain, dummy node to fast compare, or deploy simplify abstract.
  • Current chain deserialization assumes loading the entire chain at once, which is inefficient and unrealistic for syncing nodes in a real network.
  • Block synchronization logic should not rely on serializing/deserializing the full chain, but rather support range-based fetching from peers.

Solutions

  • Added FileStorage as new persistent chain backend (off-memory).
  • Created initial devlog to track major decisions.
  • Renamed project from MMB → VSTD (Vietnam Stable Digital) to improve clarity and positioning.
  • Updated test logic to reflect gas-free transaction support.
  • Refactor all self.chain references to use new storage wrapper.
    • Replace all self.chain to the new storage wrapper
    • REWRITE all the old dummy chain logic for like, serialize, deserialize.
    • Rewrite chain sync, chain deserialization

June 26, 2025

Problems

  • Continue solving the chain sync problems and chain deserialization problems.
  • Continue solving Current chain deserialization assumes loading the entire chain at once, which is inefficient and unrealistic for syncing nodes in a real network.
  • Rename mmb_layer0 to layer0 for better naming
  • Write some more test just in case for furthur testing environment
  • Rewrite chain sync
  • When reorg -> State change, need to save state diff to block for reverse.
  • Create Fast API (temporary) for the blockscan -> Need to create block scan website
  • Future task: Implement dynamic minting transaction!!!
  • Save and request transaction
  • Execution fail doesn't mean block is invalid

Solutions

  • Depricated any chain sync and remove any other chain dummy objects.
  • Finish replacing all imports from mmb_layer0 to layer0
  • Refactor all self.chain references to use new storage wrapper.
    • Replace all self.chain to the new storage wrapper
    • Make normal block sync work
    • REWRITE Depricated all the old dummy chain logic for like, serialize, deserialize.
    • Rewrite chain sync, chain deserialization
  • Rewrite chain sync logic
  • Implement state diff logic
    • Capture state diff during tx execution
    • Store diff mapped to block height
    • Revert logic on reorg using diff
    • Optional: persist diff to disk (future-proof)
  • Implement blockscan REST API
  • Save and search transaction
  • Fix unexpected bug of nonce where nonce updated for 0x0 not for the sender
  • Refactor gas for transaction

June 27, 2025

Problems

  • The state diff -> Not planed
  • Now implement REST API (or RPC) for blockscan and a smallest wallet
  • Make a chrome extension for a wallet
  • Need to fix pytest error
  • Finish blockscan
  • Refactor gas for transaction
  • Need some addition metadata for transaction tho :D
  • Very wierd worldstate mismatch for syncher ??
  • I think transaction need more data like timestamp and stuff lol

Solutions

  • Implement blockscan REST API (Ok it fine)
  • Save and search transaction
  • Implement wallet
  • Implement web interface for blockscan (Yo it so niceeeee)
  • Yo we need to fix the test!!! (EZ:D)

June 28, 2025

Problems

  • Need some addition metadata for transaction tho :D
  • Very wierd worldstate mismatch for syncher ??
  • I think transaction need more data like timestamp and stuff lol
  • Make a chrome extension for a wallet (maybe :D)
  • Need to fix network test, either implement multiserver testing or find a better way to test network

Solutions

  • Implement wallet
  • Check the worldstate mismatch
  • Implement transaction metadata
  • Fix integration test for node (because of peer limit)
  • Fix unit test when append metadata for transaction
  • Actually use those metadata for transaction
  • Fix why the node can't connect to the network (Or luck?)

June 29, 2025

Problems

  • Make a chrome extension for a wallet (maybe :D)
  • Nah Just make a react app for wallet
# Sign block
consensus.sign_block(block)
block.miner = consensus.get_validators() # Hardcoded
  • This shit took me many hours to figure out
  • Not a error, Just signing (seal) block before setting the miner and it make me struggle :DDDD

Solutions

  • Implement wallet UI
  • Fix some wallet bugs (Because the UI is so cool (AI generated))
  • Fix python typing because I switch to vscode :DD
  • Rewrite the address logic
  • Implement wallet connetivity to the network
  • Somewhere they need to implement miner that cause signature error

June 30, 2025

Problems

  • The address logic is so bad (ok my life is suck it I don't care anymore)
  • Continue interact react app for wallet
  • Very weird transaction mismatch for client to server (Client side signing fail!!!)
  • WHYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYY
  • ok my life is suck
  • Javascript ECDSA SUCK
  • Why there is space between the string because of jsonlight dumps??????
  • {a: b} is not {a:b} ok my life is suckK
  • Nah this bug make me sick
  • Still not fixed it yet. The message is sync but still bad signature error
  • Again, ok my life is suck today

Solutions

  • Implement wallet connetivity to the network
  • Somewhere they need to implement miner that cause signature error
  • First step at smart contract
  • Dynamic config (mint, validator, ...)

July 1, 2025

Problems

  • Swtich to daemon wallet, CLI wallet.
  • CLI wallet but UI is Webase (Nah you know what)
  • Cry, rage, scream into the logs — but don’t stop.
  • Because that bug?
  • It didn't beat you.
  • It forged you. 🔥
  • Now I fix the bug
  • No one can beat me.

The bug is relatively simple

tx.signature = signature.slice(2); // Assumes signature starts with '0x'

There is no 0x so cutting the first 2 characters mean cut the whole signature. lol

“Sometimes, it’s just a damn 1-line bug that makes you beg your entire ecosystem of JavaScript.” — “Quanvndzai 2025”

  • Enough talking, let me make the sign page work

Solutions

- [ ] Implement CLI daemon wallet that provide API for the wallet web app - [ ] Rewrite the wallet web app.

  • [Status: FIXED]
  • Finishing web base wallet + signing + handle transaction
  • First step at smart contract
  • Dynamic config (mint, validator, ...)

July 2, 2025

Problems

  • Smart contract first implementation and syntax

Solutions

  • Syntax of smart contract
  • Storage and state of smart contract
  • First smart contract implementation
  • Dynamic config (mint, validator, ...)

July 3, 2025

Problems

  • Reconfig consensus to use BFT
  • BFT confirmation
  • Because the BFT, also because the transaction process after the block is sent, so we need to send the receipts root to all nodes if we are validator leader are validator any node that receive > 2/3 of the receipts root can mark a block finalized

Solutions

  • Implement BFT confirmation

July 4, 2025

Problems

  • BFT confirmation
  • Continue implementing smart contract

Solutions

  • Implement smart contract
  • Implement BFT confirmation

October 15 2025

Problems

  • IDK just refactoring and cleaning up

Solutions

  • Try to fix the tests