Skip to content

Latest commit

 

History

History
49 lines (46 loc) · 2.14 KB

TODO.md

File metadata and controls

49 lines (46 loc) · 2.14 KB
Raw

Current priorities

  • Write torrent server
  • Complete code to retrieve page .torrent from dht
  • Write frontend
  • Add progress events throughout
  • Implement GridFSStorage for webtorrent
  • Implement writeAssetsToGridFS transform for assetgraph

Configuration

  • Web port
  • DHT port
  • Torrent port
  • Proxy settings

Rework

  • Figure out a cleaner way for assorted modules to access lookupService and torrentService without passing it around everywhere

Misc tasks

  • Change debug messages to be of format EverArchive.MODULE
  • Use torrent-stream, when retrieving page archived elsewhere... FOR GLORY
  • Proxy support?
  • Some external resources not being archived (e.g: typekit js when testing against http://nodejs.org)
  • Need to give some consideration to upgrade path - we don't want to leave orphaned torrents
  • Hook outgoing links, notify user they're leaving archive
  • Package as node-webkit app :3
  • Mobile (cordova?) app? :O
  • Fix errors in assetgraph transforms being swallowed.
  • Render pages in web component for style encapsulation, or render huge and have a hidable, pos-absolute overlay
  • Provide ability to download zip, magnet, torrent file, also cache these
  • Provide image mode as well as html mode
  • Allow client to seed too! :D
  • Handle archiving multiple revisions of page - WILL BREAK DATA COMPAT
  • Audit gridfs-stream memory usage (Is gridfs-stream issue #58 fixed?)
  • robots.txt
  • Document that model methods return promises
  • move let declarations to closer to first use
  • Investigate if we should use trackers or not
  • Change most model methods & statics to return promises, rather than directly being generators
  • Remove most asserts (or only run in dev mode), and write tests
  • Follow on from above, null & type checking, error handling
  • Spdy push of assets
  • load config from nconf
  • Support precompiled client resources AND on-the-fly compile
  • Support CORS
  • Validate input url is text/html - we don't want people directly archiving huge binary files... Or do we?
  • if in production mode (NODE_ENV), disable mongoose autoIndex
  • Consider queing crawler tasks?
  • redis caching (LRU?) of data from gridfs - would this even be necessary? probably not