TODO

   1 Things to do for opkg, in a partial order:
   2
   3 1) Bug fixes
   4
   5 1.a) fix all code marked "XXX: BUG"
   6
   7 1.b) fix all outstanding opkg bugzilla bugs
   8
   9 1.c) Make sure Suggests, Recommends, (Enhances?), all work, (I know
  10      some of these are not yet implemented)
  11
  12 1.d) Audit all dpkg package fields for both parsing and printing support.
  13
  14 1.e) "opkg status" and "opkg info" need to complain about unknown
  15   packages. For that matter, these two commands should probably just
  16   become aliases to the same functionality.
  17
  18
  19 2) Implement new features
  20
  21 2.a) start with all "XXX: FEATURE" comments
  22
  23 2.b) implement all accepted opkg bugzilla feature requests, (also
  24      check old post "Subject: Reinventing opkg" to see if we forgot
  25      any of the good ideas we had once upon a time)
  26
  27 2.c) Add support for:
  28         opkg install foo from some-feed
  29         opkg install foo-0.4.3
  30         opkg install foo-latest # Only needed if we add a feed-order-priority option
  31      or something like that.
  32
  33 2.d) Experiment with making a small statically linked binary. Can we
  34      eliminate some library calls, (glob, regcomp)? Can we get
  35      something like uclibc to work well?
  36
  37 2.e) Add support for a deb-src in /etc/opkg.conf. (ARGH! Name clash
  38      alert! In apt-land, deb-src means something very different than
  39      what I would mean here. Hrm... Maybe src-deb would be cleaner? )
  40
  41      (What is this?  Jamey 7/23/2002)
  42
  43 2.f) Figure out a clever chroot mechanism for running maintainer scripts in
  44      offline_root mode.
  45
  46 2.g) Implement opkg history mechanism, with undo and redo?  Keep track of
  47      packages removed/installed by each call to opkg.
  48
  49
  50 3) Cleanup the code
  51
  52 3.a) Start with all comments marked "XXX: CLEANUP"
  53
  54 3.b) Clean up out of date comments. That really confusing
  55
  56 4) refactorying opkg_install_pkg into more precise functions
  57 4.1) refactory upgrade list first
  58 4.2) Finding canditate is linear search O(P*PN) and is very slow (been called very frequently)
  59       P provider
  60       PN pkgs in a provider
  61      It's can be O(P) if there we use hash table.
  62      It should be refacotry to a faster one.