Skip to content

Latest commit

 

History

History
58 lines (58 loc) · 2.15 KB

Services.md

File metadata and controls

58 lines (58 loc) · 2.15 KB
  • Repository services v1.0
    • AI
      • NER
        • Title and Description to Std Name
      • SeqToSeq
        • Temporary solution: Regex script
    • Url of a file to CID on IPFS
    • Cache service (database interface)
      • spigot / bukkit etc. id / name to repository name
    • Scraper
      • Integrate Sentry
      • Spigot scraper
        • New plugins
          • Check if plugin is already in the repository, after pushing, add it to the repository
        • Update
          • Plugins data
          • Versions
      • Bukkit scraper
        • New plugins
        • Update
          • Plugins data
          • Versions
    • spigot.json / bukkit.json etc. to data.json
      • Make versions object the file (in gameVersions)
  • Repository services v1.1
    • UNIT TESTS!!!
      • Project is getting larger, and larger, and developing any feature at this point is very hard. After the initial release (v1.0) i plan on adding unit tests to it.
    • GitHub fetcher
      • Update
        • Plugins data
        • Versions
    • GitLab fetcher
      • Update
        • Plugins data
        • Versions
    • BitBucket fetcher
    • Scraper
      • ? Automatically check "Are you a human" cloudflare captcha
    • Supported game versions and api versions for certain version
      • "-mc." in plugin name detection
      • analize of plugin file
  • Repository services v1.2
    • Automatic dependency detection
      • By using GPT-3 trained NER model on README, Title, and Description
    • Background image generator
      • For newly created plugins
      • Using midjourney/playgroundai.com / locally run Stable-Diffusion
  • Repository services v1.3
    • Get Bukkit usernames from web.archive.org
    • Version to Semantic Versioning
      • !BLOCKER! Only works with OpenAI Davinci model, which is too expensive!
      • Try gpt3.5 chat model
      • Maybe use trained NER model
  • Repository services v1.4
    • ? Calculate difference between packages to make updates smaller in size
      • Similarly to Steam update process / Delta RPM
      • Is it a good idea? I mean it seems cool, but at the same time, it would be hard to implement, and to maintain.
    • Rewrite parts of services, with performance in mind.