Software Aggregator

• db-centric webapp (maybe Catalyst)
• non-blocking multi-tasking backend (maybe POE)
• multiple datasources:
  • software repositories like sourceforge, berlios, savannah, cpan, gna, etc.
  • software directories like ports, pkgsrc, portage, etc.
  • software catalogues like freshmeat, tucows, etc.
  • separate sites like winamp, mozilla, etc.
  • crawlable mirrors like ibiblio, citkit, osuosl, etc.
  • parseable rss feeds
• single database
• multiple front-ends:
  • web-interface
  • web-services
  • e-mail notifications
  • rss, etc.
• essential data:
  • names
  • websites
  • distfiles, where to fetch
  • packages, where to fetch
  • branches
  • current version on each branch
  • descriptions
  • sourceforge-like summary
  • alternating dependency trees