4 This is more or less complete list of tasks that has to be done before
5 SILC 1.0 could ever be released. It is clear that the list does not
6 include all the bugs that exists. At the end of list are tasks that
7 needs to be done but are probably post 1.0.
9 Feel free to contribute if you have the ability and free time - all the
10 help is really appreciated - and needed.
18 o We should replace all short, int, long, unsigned short, unsigned int,
19 unsigned long with some pre-defined datatypes that really are what
20 we want on all platforms. int16, uint16, int32, uint32 etc. are
21 what we could use or maybe SilcInt16, SilcUInt16 etc. Also, boolean
22 datatype should be defined.
25 TODO In SILC Client Library
26 ===========================
28 o Client library crashes if for example server timeouts protocol
29 execution and disconnects the client. The client, on the other hand
30 may still assume that the connection is active, even after receiving
31 the EOF. Reason for this is that the clien library does not handle
32 the SilcSocketConnection reference counter at all. This must be
35 o Logic for handling multiple same nicknames for example in private
36 message sending. I guess the logic is done in server side but is
39 o I guess, public key authentication (when connecting to a server)
40 is not working currently. It is just matter of loading the keys
41 from file and using them (see corresponding code in server, it should
42 support public key authentication already).
44 o Non-blocking connection on the background must be stopped if some
45 other connection on same window has established. Now it is possible
46 that some non-blocking connection timeouts on the background when
47 we already have a working connection to some other place; things
50 o Add client library parameters or options that handle what kind of
51 messages the library should print out (using `say' client operation,
52 for example) and what is left for the application to print. The
53 appliation could for example set that it handles all command printing
54 but all error printing should be handled by the library, etc...
55 This is not a showstopper.
57 o Input line on UI is buggy. Cursor movement etc bugs. Too lazy to
64 o TODO in commands (command.c and command_reply.c):
66 o RESTART is not implemented
67 o In servers all command reply funtions should still call the
68 pending command reply even if the reply was error. In client
69 it is not called but in server, I think, it must be called.
70 When implementing this check that all commands handle the
71 situation correctly when it is called as pending command
72 (it should most likely check that cmd->pending == TRUE/FALSE).
73 o WHOIS and IDENTIFY command reply sending is buggy because
74 it may put status to be SILC_STATUS_LIST_START even though
75 there is only one valid entry (thus should be SILC_STATUS_OK).
76 This happens because it does not check the data.registered == FALSE
77 clients before setting the status. Good fix for this would be
78 to make sure that the clients table does not include any
79 unregistered clients in the first place.
80 o Cleanup LEAVE command's channel key sending, it can be done
81 simpler by one function: silc_server_send_channel_key.
83 o TODO in notify types (packet_receive.c):
85 o SERVER_SIGNOFF notify type is not implemented
87 o Packet processing can be made faster. All packet function in the
88 packet_receive.c has same prototypes. Instead of calling those from
89 huge switch() make a table of callback functions that can be called
90 directly by the packet type.
92 o DNS/IP lookup blocks the server. This must be fixed. Check the
93 resolver stuff (resolver(3), resolver(5)). Either we have to do the
94 own resolver stuff (through scheduler, if possible without writing
95 too much own stuff) or use threads.
97 o Acceptance of incoming connections (client and server connections)
98 should be checked before key exchange protocol. Currently it is
99 checked at the authentication phase after KE, that is ok, but it should
100 be checked before starting KE, as well.
102 o Server says that it is able to listen on multiple ports but currently
103 that is bogus. It can, but internals are for single server.
105 o Protocol execution timeouts are hard coded, should be configurable.
107 o IP address fields in configuration file should accept mask format
108 as well, IP/MASK, and not just plain IP.
110 o Connection classes should be actually implemented in serverconfig.c.
111 They can be defined but they are totally ignored currently.
114 TODO In SILC Libraries
115 ======================
117 o Implement PFS (Perfect Forward Secrecy) flag in SKE (and in client and
118 server, actually). If PFS is set, re-key must cause new key exchange.
119 This is required by the SILC protocol.
121 o Re-key in general is actually missing (from everywhere) and must be done.
123 o Compression routines are missing. The protocol supports packet
124 compression thus it must be implemented. SILC Comp API must be
125 defined. zlib package is already included into the lib dir (in CVS,
126 not in distribution), but it is not used yet, and it requires some
127 tweaking on the Makefiles (we want static lib not shared).
129 o Rewrite the task system. I made it too complex and too "neat" and
130 it really should be rewritten. We don't need priorities really, one
131 priority is enough. This will simplify a lot the task system.
133 o SIM support for SILC PKCS API needs to made so that they could be
134 used as SIM's. At the same time some work is required on prime
135 generation as the way it is done now sucks. Read from code for
138 o Random Number Generator needs some tweaking. Reading /dev/random may
139 block resulting slow initialization of RNG. Some other things in the
140 RNG may block as well. Also, I have some pending changes to the RNG
141 that needs to be commited (from Schneier's Yarrow-160 paper). They
142 should make the RNG even better.
144 o Scheduler needs to be analyzed on high load as it might be unfair
145 towards select() because it may run timeout tasks before select() and
146 after select(). If it is found to be unfair the timeout task running
147 before select() should probably be removed.
150 TODO in the protocol before SILC 0.x
151 ====================================
153 o New commands and features in the commands
154 (draft-riikonen-silc-spec-xx.txt):
156 o Define GETKEY command to fetch the public key of a server
157 and/or a client in the SILC Network.
158 o Define SENDKEY command to send your public key to a client
159 in the network. Sending to the server must not be done due
160 to various security reasons (the server must not trust the
161 public keys blindly without third party verification; that's
162 why SENDKEY is not for servers).
163 o Define AWAY command to set the indication flag whether the
164 client is present or not. Do not save the away message to the
166 o Define the channel founder property to be permanent locally in
167 the server so that channel founder can regain its rights even
168 if it disconnects from the server. Thus, define a new command
169 or channel user mode that can be used to set the channel founder
170 passphrase or public key that can be used in the authentication
171 when regaining the founder rights.
173 o New features in the KE/auth protocol
174 (draft-riikonen-silc-ke-auth-xx.txt):
176 o Define group exchange support for the SKE so that the SKE
177 could be performed among more than two entities. This is not
178 a showstopper and may be defined later.
184 o Pthreads support. A lot of problems are solved with server (and with
185 client as well) if we add pthread support. We can forget things such
186 as non-blocking connecting etc, and we can do things such as DNS/IP
187 lookups async. The server itself also benefits great deal from
188 threads, especially from performance point of view.
190 But, this is not a small task and almost entire SILC Library has to
191 be made re-entrant. Own API is probably added for the threads support
192 to make changes in the future as painless as possible. So the API
193 would have things like silc_mutex_lock, silc_mutex_unlock and
196 o X.509 certificate support. SILC protocol supports certificates and
197 it would be great to have support for them. This is a big task as
198 support has to be made for ASN.1 as well. I've looked into OpenSSL
199 package as it has X.509 certificate support (and ASN.1 as well).
200 The code does not look very good to my eye but it has some potentials.
201 This should be looked at more closely.
203 Naturally own SILC Certificate API has to be defined regardles what
204 the actual X.509 library is (OpenSSL X.509 or something else). Other
205 choice is to write own X.509 library but I'm not going to do it -
206 I can help to migrate the OpenSSL X.509 into SILC and I can help if
207 someone would like to write the X.509 library - but I'm not going
208 to start writing one myself. Anyhow, the OpenSSL X.509 lib should
211 o SSH2 public keys support. Maybe - not really needed but could be
212 nice as SSH is widely used all over the place. SILC Protocol
213 supports SSH2 public keys.
215 o Cipher optimizations (asm, that this) at least for i386 would be nice.