5
- overlapping helper feature implementation (plain or stateful or both).
6
- Someone other than Robert to review the draft squid 3 release notes (in CVS already).
10
- Windows port (Merging in process, won't delay 3.0).
11
- Generalize socket binding to allow for multipe ICP/HTCP/SNMP sockets
12
(get rid of udp_incoming_address) (???)
17
- reply_body_max_size syntax cleanups. (done)
18
- Other sponsored work that needs to get into mainline sooner rather
19
than later (None known of - RC).
20
- Update automake and autoconf on squid-cache.org for the squid-3 CVS snapshot generation. (done)
21
- ESI client side streams support framework (done).
23
- Authentication tidyups in proxy_auth acl managagement (Done).
24
- Appropriate parts of rproxy (Done).
25
- Some kind of Range support to at least the level of 2.5 and earlier (Done).
30
- always/never direct syntax cleanups.
31
- refactoring of acl driven types to reduce amount of duplicated
32
code (acl_check, acl_tos, acl_address, acl_size_t, ...)
33
- Go over the old TODO (below) and sort out what has
35
o) Needs to be done, and when
38
- Customized logging capabilities
41
- Other devel.squid-cache.org projects
42
- header filtering improvements (bug #601)
46
- Rework the store API like planned
47
- Improved event driven comm code
48
- Buffer management (required by store API and comm code improvements)
49
- Improved reply processing to not reparse the reply headers several
51
- Implement HTCP authentication
52
- Log HTCP queries somewhere, like htcp_query.log
53
- There are various places where ICP means ICP/HTCP, and
54
other places where it does not. For example, cachemgr 'counters'
55
stats like icp.pkts_sent count only ICP. However, in 'peer_select'
56
stats, ICP does include HTCP as well.
61
******************************************************************************
62
Our OLD Todo List, still containing some interesting items.
63
******************************************************************************
65
() dont re-sibling requests which came from a sibling
68
(1) is high priority, (5) is low priority
69
(**) Means we need this for the real 1.2 release
70
==============================================================================
72
(1) A customizable cache replacement policy. Ugh, this could be
73
interesting since we just optimized the LRU replacement with a
76
(2) be more persistent in forwarding requests; return fewer error pages
77
(2) Fix some problems with peer-selection.
78
- Allow for 'backup' parents
79
- An orderd list of siblings to try.
80
- Hash function load balancing
81
(2) Calculate Content-Length for multipart range replies (AR)
84
(3) ETag support (makes metadata bigger--ack!)
85
(3) Efficient public peer access control *without ACLs*
86
Bill Wichers <billw@unix0.waveform.net>
87
Dave Zarzycki <zarzycki@ricochet.net>
88
# distant_peer_deny (icmp rtt) (hops)
89
# If the ICP client is farther than "z" hops away,
90
# or if more than y% of pings to the client exceed
91
# x milliseconds, then deny.
92
distant_peer_deny 200 75 10
93
(3) deny/allow tags for HTTP header fields (AR)
94
(3) LDAP support. See http://miso.wwa.com/~donley/ and
95
Clayton Donley <qa1049@ilms04.cig.mot.com>
96
(3) What to do about ACL's and URL escaping?
97
(3) Review RFC 2068 and RFC 2109 for header handing, especially
100
(4) ** Everywhere that we use 'pattern' or such, use ACL elements instead.
101
stoplist_pattern, refresh_pattern, ... (DW)
102
(4) Refresh based on content types. This means we'll need an enum of
103
known content types added to StoreEntry. Unknown types will lose.
104
(4) Write binary headers as metadata?
105
(4) HTML-escape special characters in errorConvert().
107
(5) X-Proxy-hops header?
108
(5) internalize redirectors?
109
(5) For _popular_ objects, return hits if known to be stale, but then
110
initiate an IMS/refresh in the background.
111
(5) #ifdefs to disable IP caching
112
(5) REST for failed ftp transfers.
113
(5) Customizable access logging.
114
(5) Inline DNS queries with native resolver calls.
116
(5) return "ERR_TOO_LARGE" messages for replies with
117
content-lengths larger than some limit.
118
(5) Bind HTTP/ICP sockets to multiple interfaces
119
(5) Custom X-Forwarded-For
122
+ s/sprintf/snprintf/g
123
+ Write the text for all the error message files
124
+ fix all the assert(!ERR_FOO) placeholders
125
+ Finish hacking at storeCopy() to achieve in-memory sliding window
126
for objects being retrieved.
127
+ make USE_PROXY_AUTH standard
128
+ single linked list for disk LRU replacement
129
+ Incorporate Pelletier's latest retry patch for 1.1
131
(1) reconfigure (SIGHUP) causes problems with Config.cacheSwap.swapDirs[].
132
(1) I think we're assuming persistent connections and not looking at
134
(2) Write src/store_key_md5.c and lib/md5.c
136
(5) convert ipcache/fqdncache to double-linked list
137
and lose qsort() calls.
139
- whois -h whois.ra.net '\!gAS194'
141
(3) SIGHUP should reread error text files
142
(4) support ftp://host/pub;type=d
143
(5) should we write clean store logs from the double-linked list so
144
that they are sorted in order of last reference? Then upon
145
restart we won't have to worry about the list being unsorted
146
and causing problems with the LRU magic value.
147
(3) ** Make checkFailureRatio() work again. (DW)
148
(2) ** copy over -z semantics from 1.1; don't automatically wipeout
150
(3) ** Make swap.state binary with fixed-size cache keys (AR)
151
(3) ** Make sure every squid.conf option has documentation (??)
152
(3) ** Store cache key and other metadata in swap files (KA)
153
(3) ** Finish implementing the *_dump() functions in cache_cf.c (DW)
154
(2) ** Fix cachemgr.cgi. It doesn't understand some of the new
156
(2) ** Add "Squid" trailer to all error messages. (AR)
157
(2) ** copy over 'test_reachability' from 1.1 (DW)
158
(2) ** we don't properly handle the HTTP/1.1 'Connection' header (KA)
159
(2) ** Code cleanup: (KA)
160
- Make sure the lib/*.c files do not use printf/fprintf.
161
- Replace bcopy with xmemcpy
163
(1) SSL doesn't work with Proxy Authentication
164
(3) HTTP Header todo list (AR)
165
(5) replace "for" loops in md5.c with memcpy and memset
166
(3) ** shrink-able memory pools (AR)
167
(3) ** request-range (AR)
168
(4) ** Double check that MemBuf.size is used correctly everywhere (AR)
169
(3) 'no_cache' access list for specifing objects which should not
171
(3) Full request headers for CONNECT requests.
172
(1) ** Send full HTTP requests in ICP messages (DW)
173
(2) Track cache keys (MD5) in memory utilization
175
(3) Arjan's external proxy auth patch is at
176
http://www.iae.nl/users/devet/squid/proxy_auth/
177
(3) In some situations, a private-key reply needs to eject a
178
public-key cached object. For example when an object is
179
first cachable, but later becomes uncachable.
180
(4) Configurable proxy-auth realm
181
(4) Configurable FQDN cache size