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
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
|
# -*- text -*-
AM-UTILS-6.0 TASKS TODO
Please volunteer to do any of the following:
- complete testing of untested platforms in INSTALL file
ncr2
sunos 3.5
- autofs support: see README.autofs for details.
needs to be re-ported to solaris 2.6 (headers changed)
- deal with everything that has XXX on it in the sources
- documentation update ("XXX: FILL IN" sections).
- hlfsd should be able to not use /var/alt_mail, but send code 75 back to
sendmail (telling it to re-queue the mail and retry later).
- compatibility with Sun's automount maps?
should be easier via the amd.conf file to specify type of map
perhaps done at the same time autofs support is done.
- a selector ala if_exists() for networkmask(1.2.3.4/5.6.7.8), will match
against all known IP addresses of this host.
- convert to using my own rpcgen .x files for amq/amq/nfs (v2 and v3)
- support multiple "fail-over" NFS mounts in Solaris 2.6.
- use packaging info for various OSs (such as RPM, Redhat Package Format)
- $mindelay and $maxdelay in milliseconds
- multiple nfsl should be matched if one matched and nfs mount failed. fall
through.
- fall through syntax? opts:=nofail,fallthrough?
- random nfs rhost:={srv1, srv2, srv3}
- nfslx, same as nfsl, but using linkx
- find out why this sometimes works and sometimes not:
mcl -rhost:=minetta host!=${rhost};type:=nfs host==${rhost}
- loadable info_*, amfs_*, and ops_* modules (lazy evaluation).
- hlfsd for ~root/.mailspool (getpwnam("root"))
- hlfsd should daemonize even if -DDEBUG, then use -D nofork
- fixmount should use generic code from transp/transp_{tli,sockets}.c
- report netbsd/openbsd bugs in BUGS file
- reverse notion of -F and other cmd-line options, so they override the
amd.conf file (right now amd.conf overrides cmd-line options).
- y2k compliance.
- an un/install script for binaries
- update README
- add am-utils URL and COPYRIGHT to all distributed scripts and sources and
programs etc. also to amq -v output.
- ion: browsable_dirs only works for nis if in [global] broken if it is file
maps.
- all if statements should have an else clause because of Ultrix's /bin/sh.
- look for obsolete AM_* macros and use newer AC_* ones
- use 3rd arg to AC_CHECK_TYPE (headers)
- FreeBSD 3.1 doesn't have SYMTTL flag, so the mtime of the hlfsd symlink is
always new. This breaks hlfsd on FreeBSD 3.1, b/c the acregmin/max code
depends on the time that the symlink was created. Solution: best is to
support NOAC in the FreeBSD kernel, otherwise hlfsd will have to know not
to turn on this code in hlfsd/stubs.c:
if (++slinkfattr.na_mtime.nt_useconds == 0)
++slinkfattr.na_mtime.nt_seconds;
- support nqnfs as per Jeffrey C Honig <jch@BSDI.COM>: "Besides having
NFSMNT_NQNFS set when you try to do the mount, NQNFS requires that you do
the nfssvc() call to create an nqnfs client daemon. I can't see any
obvious way to check if the server supports NQNFS."
- browsable 'auto' style maps.
- amq man page: amq -f also reloads maps that are not updated
- mention signals in amd man page
- ALLOWED_MOUNT_TIME of 40 seconds is way too long!
- need option to turn off pings of portmapper before trying a mount
- type:=program should not require umount command. can use default
umount(). it works for smbmount/umount.
- document var[0-7] variables/selectors
- need a way to export a nfs-mount point after it is being mounted (jukebox
mounts). can do this with type:=program.
- support port=n, for machines that use NFS on a different port
- support WebNFS (cf. Solaris mount_nfs(1M) -o public)
- nfsl may be buggy when fs:= is specified explicitly (lockup?)
- detecting down'ed hosts faster: use default portmap TTL variables?
- entries w/ $key in /defaults, see "defaults" instead of the actual entry.
- y2k: amq /home (mounted time uses YY instead of YYYY)
- new amd.conf option to disable amq listener altogether (security)
|
