What a vote transaction actually contains
A close reading of a single vote transaction, line by line, with the on-chain signature and the slot it landed in.
Read the entry →Solana reference, edited by hand
A reading desk for one proof-of-stake network, with margin notes in fountain pen.
Mentisarcway is a small library of articles, glossary entries, and ecosystem deep dives. Each entry defines its terms, traces design decisions back to on-chain proposals, and pairs the concept with a worked example you can verify against a block explorer.
Reader questions
The things a first-time reader asks before trusting this shelf as a reference rather than a blog.
Where do the entries get their information?
Each article cites primary sources: Solana Improvement Documents (SIMDs), client release notes, validator calls, and on-chain account data. When we paraphrase secondary commentary, we name the author and date it. A footnote block at the foot of every entry lists what we read.
How often is an entry re-checked?
Every entry carries a changelog noting which network version and client release it was last verified against. We re-read entries when a mainnet upgrade lands or when a referenced account changes state, and we record the date of the re-check in the changelog.
Do the editors hold SOL or run validators?
Contributors disclose any related token position or validator interest on their profile and on each entry they touch. The library does not take sponsorship from token issuers, foundations, or RPC providers. Operating costs are met by the editors.
Can I cite an entry in my own writing?
Yes. Each article has a stable URL and a "last verified" line. We recommend citing the entry by title, the version date in the changelog, and the URL, the same way you would cite a printed reference work.
What if I find an error?
Write to [email protected] or use the contact form. Tell us the entry, the line, and the source that disagrees. We log the correction in the changelog with your name unless you ask to stay anonymous.
Is there a paid tier or a token gate?
No. The library is free to read. There is no Discord behind a token, no bootcamp funnel, and no premium tier that unlocks the harder entries. The only thing we accept is corrections and topic suggestions.
From the journal shelf
Teasers for further reading that sits one step harder than the home page. Each one names a primary source you can open in a second tab.
Why stake accounts separate authority from balance
The design rationale for custodial versus withdraw authority, traced to SIMD-93 and the original staking contract discussion.
Open the article →From a wallet seed phrase to a first program
A curated track that moves a reader from keypairs through transactions to deploying a small program, with the order explained.
Follow the path →How we version an entry against a network upgrade
The checklist an editor walks when a new client release ships, from re-reading the entry to updating the changelog line.
See the method →The shelf, by count
Numbers a reference library can defend. No satisfaction scores, no growth percentages.
142
Entries maintained this quarter
318
Terms defined in the glossary
9
Reading paths across the ecosystem
2023
Oldest entry still on the maintenance rota
Flagship reading path
From a wallet to your first program
A structured track for a reader who can sign a transaction but has not yet deployed code. Seven ordered entries, each with the rationale for why it sits where it does, and a worked example that compiles or signs against a public network.
- Keypairs, seeds, and what a derivation path actually fixes
- Transaction structure: instructions, accounts, and the order of execution
- Reading a program on-chain: account layout and the cost of each byte
- Stake accounts as a first non-trivial program to read end to end
- A worked deploy against devnet, with the signature to verify
Who maintains the shelf
A small editorial team. Each contributor names the area they maintain and discloses any related token or validator interest, so a reader can weigh the voice behind an entry.
Dohyeon Ahn — founding editor
Maintains: consensus entries, vote credits, leader schedule
Dohyeon started the library after three years of answering the same questions in validator chats. He reads SIMDs in full before summarising them and keeps a personal rule of never writing an entry the same week he changed his own validator config, to avoid bias. Disclosure: runs a small, non-commissioned validator on testnet only.
Minjae Park — glossary editor
Maintains: glossary terms, cross-references between entries
Minjae treats the glossary as the spine of the shelf: a term used in any article must already have a definition, or get one before the article ships. She trained as a translator and writes definitions that a reader can quote without paraphrase. Disclosure: holds a small SOL position acquired before joining the project; no validator interest.
Soyeon Han — programs editor
Maintains: program deep dives, account layouts, client release notes
Soyeon writes the program entries. She decompiles the on-chain bytecode when the source is unavailable and quotes both. Before Mentisarcway she wrote internal documentation for an exchange and left because the work was not allowed to be public. Disclosure: no token position, no validator interest, no paid consultancy.
Two honest ways into the subject
We are not pretending one path is morally superior. Both readers below are real; both arrived at the same working understanding.
The reader who learned from scattered sources first
"I spent eighteen months piecing this together from Discord threads, validator call recordings, and a folder of bookmarks that broke every few months when docs moved. I got there, but I cannot tell you how many hours were spent re-checking things that turned out to be out of date. What I want from a reference now is a date stamp and a footnote, so I can trust the page in front of me for one evening of reading without opening six tabs to verify it."
The reader who started here
"I came in cold from a non-engineering background. The glossary let me read the first article without stopping every line, and the reading path gave me an order instead of a menu. The trade-off is that I have less of the 'I survived the messy sources' instinct — when an entry is wrong, I am slower to notice it. That is why the changelog matters to me; it is the only way I know an entry has been re-read recently."
Send a topic, a path suggestion, or a correction
A request starts a conversation, not a sales call. We reply with a citation, a draft paragraph, or a question — usually within five working days.
Write to the editors