brk/docs/TODO.md

TODO

• CRATES
  • BUNDLER
  • CLI
    • launch
      • if first, test read/write speed, add warning if too low (<2gb/s)
      • check available disk space
      • pull latest version and notify if out of date
    • add custom path support for config.toml
  • COMPUTER
    • add rollback of states (in stateful)
    • add support for per index computation
    • fix min fee_rate which is always ZERO due to coinbase transaction
    • before computing multiple sources check their length, panic if not equal
    • create usd versions of vecs structs instead of having options everywhere
    • datasets
      • sats version of all price datasets (average and co)
      • pools
        • highest dominance
        • consecutive blocks
        • max consecutive blocks
        • add indexes back to the mempool.space version of pools as we need the correct one for the API
      • price
        • oracle (https://utxo.live/oracle/UTXOracle.py)
      • cohorts
        • costs basis by percentile
        • address counts relative to all datasets
        • revived/sent supply datasets
        • add p2pk group (sum of p2pk33 and p2pk65)
        • add utxo count, address count, supply data for by reused addresses in groups by address type
        • add more date ranges (3-6 months, 5-7 years and more)
        • % of volume
      • indicators
        • pi cycle
      • https://mainnet.observer
      • https://glassnode.com
      • https://checkonchain.com
      • https://researchbitcoin.net/exciting-update-coming-to-the-bitcoin-lab/
      • https://mempool.space/research
  • ERROR
  • FETCHER
  • INDEXER
    • parse only the needed block number instead the last 100 blocks
      • maybe using https://developer.bitcoin.org/reference/rpc/getblockhash.html
  • INTERFACE
    • Maybe change json to:

      {
        "price_close": {
          "data": [115976.6, 115943.4],
          "unit": "USD",
          "range": {
            "from": 19627,
            "to": 19628
          }
        }
      }
      

    • create pagination enum
      • from to
      • from option
      • to option
      • page + option default 1000 max 1000
    • from/to/count params don’t cap all combinations
      • example: from -10,000 count 10, won’t work if underlying vec isn’t 10k or more long
  • LOGGER
    • remove colors from file
  • MCP
  • PARSER
  • SERVER
    • api
      • copy mempool's rest api
        • https://mempool.space/docs/api/rest
      • add extensions support (.json .csv …) instead of only format
      • if format instead of extension then don't download file
      • ddos protection
        • against API params varying in range
        • search
        • fuzzy on typo
          • https://github.com/rapidfuzz/strsim-rs or stick with current impl
          • create map of all single words
          • do some kind of score with that ?
      • discoverability
        • catalog (tree/groups)
        • search
      • failover to /api
      • no HTML / redirects ?
      • change /api/vecs/{index}-to-{metric} to /api/{metric}/index
      • change /api/vecs/query to /api/bulk
      • support keyed version when fetching dataset: {date: value} / {date: [value]}
    • add support for https (rustls)
  • STORE
    • save height and version in one file
  • STRUCTS
  • GLOBAL
    • https://davidlattimore.github.io/posts/2025/09/02/rustforge-wild-performance-tricks.html
• DOCS
  • README
    • add a comparison table with alternatives
    • add contribution section where help is needed
      • documentation/mcp/datasets/different front ends
    • add faq
• WEBSITES
  • PACKAGES
    • move the fetching logic from bitview website to an independent brk package which could be published to npm
      • https://www.npmjs.com/package/@mempool/mempool.js
      • auto publish with github actions
  • BITVIEW
    • explorer
      • blocks (interval as length between)
      • transactions
      • addresses
      • miners
      • maybe xpubs
    • charts
      • selected unit sometimes changes when going back end forth
      • add support for custom charts
      • price scale format depends on unit, hide digits for sats for example (if/when possible)
      • shows certain series as scatter plots with a solid sma/ema
        • mainly datasets with a big variance like raw hash_rate
      • hide pane if no series on it
      • fix (and reset) pane size (50/50) when changing charts
      • units: add short name / long name / title
      • verify that "compare" folders aren't missing charts/datasets
      • legend
        • add link to explanation for each name (to glassnode ?)
    • table
      • pagination
      • exports (.json, .csv,…)
      • improve dataset selection
      • display 1k values (instead of 10k) but to avoid caching multiple times the same values apply everywhere
    • search
      • improve
      • datasets add legend, and keywords ?
      • support height/address/txid
    • api
      • add api page with interactivity
    • glossary ?
    • nav
      • move share button to footer ?
      • when clicking on already selected option, pushes to history, bad !
    • global
      • improve behavior when local storage is unavailable
        • by having a global state
      • font:
        • https://fonts.google.com/specimen/Space+Mono
    • keep as many files as possible under 14kb
    • No classes
    • Organic animations
• GLOBAL
  • check TODOs in codebase
  • rename output to txout or vout, input to txin or vin
  • https://マリウス.com/thoughts-on-cloudflare/