File: TODO

package info (click to toggle)
rlinetd 0.9-3
  • links: PTS, VCS
  • area: main
  • in suites: jessie, jessie-kfreebsd
  • size: 2,928 kB
  • ctags: 1,291
  • sloc: sh: 12,370; ansic: 3,212; yacc: 1,585; makefile: 174; lex: 131; sed: 16; perl: 4
file content (35 lines) | stat: -rw-r--r-- 1,701 bytes parent folder | download | duplicates (12)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
-*- text -*-
o pre-exec logging as well as post-exit
o further error and sanity checks in initialization process. improve
reporting of errors back to user.
o redirection service (a la rinetd) sendfile doesn't seem to work.
o better backend data structures than unsorted lists
o make the tcp_wrappers stuff a module
o rfc1918 module - kinda done with the tcp_wrappers stuff
o environment % substitutions
o timers for maximum process life and periodic events (stat generation)
o groups other than 'default' in config (macros, effectively)
o is there any way of deducing the protocol in use given the socket
type? ie, having called socket(x, y, 0), how can you find out the
protocol number to pass to pmap_set()? this will be useful in allowing
people to only specify a sockettype of "stream" without needing to
specify "tcp" as well...
o signal-triggered actions (logging, killing services,...?)
o get the log string fragments into the stringtab
o do some profiling :)
o banner arg fed through argvtab?
o work out how the error recovery stuff in bison works, and use it
o cache tcp_wrappers' request structure, and re-use it in multiple
requests.
o fenceposts in *get functions?
o abstract out basic string storage from logtab, argvtab, and
hypothetical dynstringtab
o get the hisock stuff happening again
o make instances nest - will require an unroll sort of thing, where
you register cleanups to be taken when a service exits.
o i18n/l10n? what's the difference, anyway?
o mod_perl
o use libltdl for the dynamic fruit
o provide a replacement for getopt if it's not available
o freebsd seems to have an rlim_t - really should use it
o sanitize function and file layout. everything's all over the place