mdw@git.distorted.org.uk Git - sgt/agedu/blob - TODO

   1 TODO list for agedu
   2 ===================
   3
   4 Before it's non-embarrassingly releasable:
   5
   6  - work out what to do about atimes on directories in the absence of
   7    the Linux syscall magic
   8     * one option is to read them during the scan and reinstate them
   9       after each recursion pop. Race-condition prone.
  10     * marking them in a distinctive colour in the reports is another
  11       option.
  12     * a third option is simply to ignore space taken up by
  13       directories in the first place; inaccurate but terribly simple.
  14     * incidentally, sometimes open(...,O_NOATIME) will fail, and
  15       then we have to fall back to ordinary open. Be prepared to do
  16       this, which probably means getting rid of the icky macro
  17       hackery in du.c and turning it into a more sensible run-time
  18       abstraction layer.
  19
  20  - cross-Unix portability:
  21     + use autoconf
  22        * configure use of stat64
  23        * configure use of /proc/net/tcp
  24        * configure use of /dev/random
  25        * configure use of Linux syscall magic replacing readdir
  26           + later glibcs have fdopendir, hooray! So we can use that
  27             too, if it's available and O_NOATIME is too.
  28        * what do we do elsewhere about _GNU_SOURCE?
  29
  30  - man page, --version.
  31
  32 Future possibilities:
  33
  34  - IPv6 support in the HTTP server
  35     * of course, Linux magic auth can still work in this context; we
  36       merely have to be prepared to open one of /proc/net/tcp or
  37       /proc/net/tcp6 as appropriate.
  38
  39  - run-time configuration in the HTTP server
  40     * I think this probably works by having a configuration form, or
  41       a link pointing to one, somewhere on the report page. If you
  42       want to reconfigure anything, you fill in and submit the form;
  43       the web server receives HTTP GET with parameters and a
  44       referer, adjusts its internal configuration, and returns an
  45       HTTP redirect back to the referring page - which it then
  46       re-renders in accordance with the change.
  47     * All the same options should have their starting states
  48       configurable on the command line too.
  49
  50  - curses-ish equivalent of the web output
  51     + try using xterm 256-colour mode. Can (n)curses handle that? If
  52       not, try doing it manually.
  53     + I think my current best idea is to bypass ncurses and go
  54       straight to terminfo: generate lines of attribute-interleaved
  55       text and display them, so we only really need the sequences
  56       "go here and display stuff", "scroll up", "scroll down".
  57     + I think the attribute-interleaved text might be possible to do
  58       cunningly, as well: we autodetect a basically VT-style
  59       terminal, and add 256-colour sequences on the end. So, for
  60       instance, we might set ANSI-yellow foreground, set ANSI-red
  61       background, _then_ set both foreground and background to the
  62       appropriate xterm 256-colour, and then display some
  63       appropriate character which would have given the right blend
  64       of the ANSI-16 fore and background colours. Then the same
  65       display code should gracefully degrade in the face of a
  66       terminal which doesn't support xterm-256.
  67        * current best plan is to simulate the xterm-256 shading from
  68          0/5 to 5/5 by doing space, colon and hash in colour A on
  69          colour B background, then hash, colon and space in B on A
  70          background.
  71     + Infrastructure work before doing any of this would be to split
  72       html.c into two: one part to prepare an abstract data
  73       structure describing an HTML-like report (in particular, all
  74       the index lookups, percentage calculation, vector arithmetic
  75       and line sorting), and another part to generate the literal
  76       HTML. Then the former can be reused to produce very similar
  77       reports in coloured plain text.
  78
  79  - http://msdn.microsoft.com/en-us/library/ms724290.aspx suggest
  80    modern Windowses support atime-equivalents, so a Windows port is
  81    possible in principle.
  82     + For a full Windows port, would need to modify the current
  83       structure a lot, to abstract away (at least) memory-mapping of
  84       files, details of disk scan procedure, networking for httpd.
  85       Unclear what the right UI would be on Windows, too;
  86       command-line exactly as now might be considered just a
  87       _little_ unfriendly. Or perhaps not.
  88     + Alternatively, a much easier approach would be to write a
  89       Windows version of just the --scan-dump mode, which does a
  90       filesystem scan via the Windows API and generates a valid
  91       agedu dump file on standard output. Then one would simply feed
  92       that over the network connection of one's choice to the rest
  93       of agedu running on Unix as usual.