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.
14 [Latest Note: The protocol has changed a bit in some parts which
15 causes that the current implementation violates some requirements.
16 These are not listed here, currently.]
22 o Extended SIM (SILC Module) support. Currently only SILC Cipher API
23 and SILC Hash API may be used as SIM's. What I have in mind is to
24 have extended support for SIM's so that basically any SILC API could
25 be used as SIM's. This would open tremendous possiblities but
26 opens also issues on security that needs to be dealt with.
28 Some sort of SIM compilation environment should be defined so that
29 the SIM's could use SILC specific symbols from the modules (which they
30 cannot do currently). In the future modules could add new features
31 to SILC easily with this support. I'm more thinking this from client's
32 perspective to add new features to client (such as IRC support as SIM)
33 but server should have the support as well. Anyhow, this is an
34 interesting feature...
36 This maybe post 1.0 task - dunno.
38 o SIM support for other platforms than just for Linux. Apache has
39 example code (code that we could use directly pretty easily) for
42 o We should replace all short, int, long, unsigned short, unsigned int,
43 unsigned long with some pre-defined datatypes that really are what
44 we want on all platforms. int16, uint16, int32, uint32 etc. are
45 what we could use or maybe SilcInt16, SilcUInt16 etc. Also, boolean
46 datatype should be defined.
48 o More platform supports should be added. The code is pretty much
49 generic but there are some parts that require porting (SIM). Also,
50 some support for different platforms is needed into configure.in.
52 o SILC requires currently GCC to work because we use GCC specific
53 compilation options. Generally any compiler that supports inline
54 functions and can build shared libraries (for SIMs) should work.
55 These cases should be included into configure.in.
61 o Implement all commands. A lot of commands are still yet to be
62 implemented. Most of them are trivial but some will require some
63 planning. Go see the command.c for unimplemented commands.
65 o Non-blocking connection on the background must be stopped if some
66 other connection on same window has established. Now it is possible
67 that some non-blocking connection timeouts on the background when
68 we already have a working connection to some other place; things
71 o Finish WHOIS, finish JOIN and other commands that are partly
74 o Input line on UI is buggy. Cursor movement etc bugs. Too lazy to
77 o Logic for handling multiple same nicknames for example in private
78 message sending. I guess the logic is done in server side but is
81 o Private message key setting is missing and must be implemented.
82 Currently private messages are encrypted with session keys. This
83 is required by the protocol.
85 o Channel private key setting is missing and must be implemented.
86 Currently there cannot be private keys for channels. Normal channel
87 keys (generated by server) are used. This is required by the protocol.
89 o Public and private key generation is now done everytime the program
90 is run. Now, this is only for testing period as I've been lazy to
91 do it any better for now. This must be fixed.
93 o I guess, public key authentication (when connecting to a server)
94 is not working currently. It is just matter of loading the keys
95 from file and using them (see corresponding code in server, it should
96 support public key authentication already).
98 o Multiple windows support. Basic support for multiple windows already
99 exists but a lot is still missing to get it working. Also, some
100 of the existing stuff probably needs to be tweaked a bit before the
101 multiple windows support could be done. And of course the actual
102 commands that control the windows needs to be written (/WINDDOW).
104 o Implement /KEYMAP (or similiar) command to remap control and function
107 o Implement /ALIAS command to make command aliases.
109 o Implement /set/if/do/while etc as in IRC2. Maybe post 1.0 task.
110 Some scripting would be good.
112 o Connection Authentication request resolving is missing and must be
113 done. This is required by the protocol.
115 o Key Exchange protocol's responder side is missing from client.
116 Generally it is possible for the client to be responder so it should
117 be implemented (See corresponding code from server). Error handling
118 in the KE protocol is also in pretty bad shape in client.
120 o Configuration file loading from global and from local dirs. This
121 is currently missing and I guess the global is only used. Old SILC
122 version (in 1997) had ~./silc directory that I guess should be done
123 now as well. The code for handling those exists but not in current
126 o Configuration file format - could be better.
128 o Write help files for commands. Nice format for the help files should
129 be selected. I'm open for ideas.
131 o All allocations and freeing needs to be checked for memory leaks.
132 Also, return values from various allocations and functions needs to
139 o Implement all commands on server side. A lot of commands are still yet
140 to be implemented. Most of them are trivial but some will require some
141 planning. Go see the command.c for unimplemented commands.
143 o DNS/IP lookup blocks the server. This must be fixed. Check the
144 resolver stuff (resolver(3), resolver(5)). Either we have to do the
145 own resolver stuff (through scheduler, if possible without writing
146 too much own stuff) or use threads.
148 o Lenght of the packet processing timeouts needs to be checked whether
149 they are too short or too long. I haven't really tested whether they
150 are suitable. They should be tested on high load which I haven't done
153 o Public and private key generation is now done everytime the program
154 is run. Now, this is only for testing period as I've been lazy to
155 do it any better for now. This must be fixed.
157 o Server says that it is able to listen on multiple ports but currently
158 that is bogus. It can, but internals are for single server.
160 o Command lagging must implemented. Those commands (all currently) that
161 has the LAG flag set they must not be allowed to be executed more than
162 once, say, in 2 seconds.
164 o Command flag usage in general is not implemented yet.
166 o Client history must be implemented. Protocol says that server must
167 keep history information about clients for some period of time.
169 o Channel flags and user modes on channels are not implemented yet as
170 /MODE command is not implemented yet in client and server.
172 o Protocol execution timeouts are hard coded, should be configurable.
174 o Channel message sending routines uses a lot of common code. Should
175 create a common function for those instead of writing the same code
176 again everytime, as done now.
178 o serverutil.c I guess should be created for util-like functions that
179 now resides in server.c, which is getting too big.
181 o serverconfig.c and the function naming in it is inconsistent. It is
182 not silc_config_server* it should be silc_server_config*. As should
183 all the SilcConfigServer* types be SilcServerConfig*.
185 o Implement DENY_CONNECTION section in serverconfig.c and in server.
187 o Implement REDIRECT_CLIENT section in serverconfig.c and in server.
189 o Configuration file format - could be better.
191 o IP address fields in configuration file should accept mask format
192 as well, IP/MASK, and not just plain IP.
194 o Connection classes should be actually implemented in serverconfig.c.
195 They can be defined but they are totally ignored currently.
197 o Acceptance of incoming connections (client and server connections)
198 should be checked before key exchange protocol. Currently it is
199 checked at the authentication phase after KE, that is ok, but it should
200 be checked before starting KE, as well.
202 o Statistics are totally missing from the server. It would be nice
203 to gather some statistics.
205 o All allocations and freeing needs to be checked for memory leaks.
206 Also, return values from various allocations and functions needs to
210 TODO In SILC Libraries
211 ======================
213 o Public key verification in SKE (SILC Key Exchange) protocol is missing,
214 thus currently we trust on all public keys. This probably doesn't cause
215 bad problems but the mechanism of verifying it from local database
216 (from files) needs to be done (it can open man-in-the-middle-attacks).
218 o Implement PFS (Perfect Forward Secrecy) flag in SKE (and in client and
219 server, actually). If PFS is set, re-key must cause new key exchange.
220 This is required by the SILC protocol.
222 o Re-key in general is actually missing (from everywhere) and must be done.
224 o SKE does not send correct status types. Types are defined but not
227 o Connection authentication protocol does not send correct status types.
228 These types are not defined currently at all.
230 o PKCS#1 style RSA public key encryption/decryption/sign/verify is
231 missing, and should be added for interoperability reasons. The thing
232 I've done now is bad and should be removed as soon as possible (or
233 the protocol should then state the method of how they should be done).
235 o SILC public key file type is bad. I'd like to see PEM encoded files.
236 I have public domain code for base64 encoding but it needs to be
239 o Slow ciphers should be removed. I think we don't need more than
240 the AES finalists plus blowfish and RC5.
242 o These slow ciphers actually don't work currently as I've tested
243 only the ones that are worth testing. The CBC mode on these slow
244 ciphers probably don't work. No need to worry, these ciphers should
247 o Scheduler needs to be analyzed on high load as it might be unfair
248 towards select() because it may run timeout tasks before select() and
249 after select(). If it is found to be unfair the timeout task running
250 before select() should probably be removed.
252 o On select() issue; maybe we should use poll() instead if it is
253 available? poll() doesn't have max fd limit...
255 o SIM support for SILC PKCS API needs to made so that they could be
256 used as SIM's. At the same time some work is required on prime
257 generation as the way it is done now sucks. Read from code for
260 o Compression routines are missing. The protocol supports packet
261 compression thus it must be implemented. SILC Comp API must be
262 defined. zlib package is already included into the lib dir (in CVS,
263 not in distribution), but it is not used yet, and it requires some
264 tweaking on the Makefiles (we want static lib not shared).
266 o Cipher API needs to be made more consistent. Some parts of the
267 code generated with current Cipher API looks really bad. Same
268 is with PKCS API, even worse actually. They need to be made
269 cleaner. Introducing silc_cipher_encrypt/decrypt/set_key etc.
270 functions (I actually don't understand why have I left these un-done).
272 o Scheduler should automatically allocate task queues if NULL pointers
273 are passed to the silc_schedule_init. Would make initialization
276 o Packet processing routines in client and server are actually pretty
277 much generic and should be moved from the client/server to the library
278 as generic routines (silc_<client/server>_packet_decrypt_rest* etc).
279 This requires heavy changes to the client and server.
281 o Random Number Generator needs some tweaking. Reading /dev/random may
282 block resulting slow initialization of RNG. Some other things in the
283 RNG may block as well. Also, I have some pending changes to the RNG
284 that needs to be commited (from Schneier's Yarrow-160 paper). They
285 should make the RNG even better.
287 o Logging should be made more generic in a way that application can
288 set to where the logging is destined to. Now, it is always destined
289 to stdout (or stderr) which is a bad thing for client. Ie. some
290 sort of logging registration functions or similiar should be done
291 (silclog.[ch] in core). The actual output of logs should be done
292 by callback function in the application not in lib.
294 o I don't like the ID cache system currenly implemented. Ugly and
295 not so good. Must be rewritten very soon.
297 o All allocations and freeing needs to be checked for memory leaks.
299 o There are also checks missing from allocations whether the allocation
300 returns valid memory or NULL. These are missing in library as well
301 in client and server. Either all checks has to be added or we will
302 have to make sure that silc_*alloc()s always return valid memory
303 and assert()s if the system's memory allocator (*alloc()) fails.
305 o silc_buffer_[un]format() needs to be made more stable as it may
306 crash the SILC if malformed data is sent as argument. There are a
307 lot of places in client and server where we trust directly data coming
308 from network and try to unformat it. The unformatting routine needs
309 to be able handle situations where data sent is malformed, by mistake
310 or intentionally. This is important as it is easy to crash the SILC
311 now by just sending malformed data. Also, in client and server we
312 must start checking the return value from silc_buffer_[un]format.
318 o Write manuals for server.
320 o Write manuals for client.
322 o Write SILC Library Reference manual. This would include all the SILC
323 API's with simple examples how the functions are to be used. This is
324 pretty easy to create by taking all the functions plus their comments
325 from source/header files. However, same sort of reference manual
326 should be written for client and server as well.
332 o Pthreads support. A lot of problems are solved with server (and with
333 client as well) if we add pthread support. We can forget things such
334 as non-blocking connecting etc, and we can do things such as DNS/IP
335 lookups async. The server itself also benefits great deal from
336 threads, especially from performance point of view.
338 But, this is not a small task and almost entire SILC Library has to
339 be made re-entrant. Own API is probably added for the threads support
340 to make changes in the future as painless as possible. So the API
341 would have things like silc_mutex_lock, silc_mutex_unlock and
344 o X.509 certificate support. SILC protocol supports certificates and
345 it would be great to have support for them. This is a big task as
346 support has to be made for ASN.1 as well. I've looked into OpenSSL
347 package as it has X.509 certificate support (and ASN.1 as well).
348 The code does not look very good to my eye but it has some potentials.
349 This should be looked at more closely.
351 Naturally own SILC Certificate API has to be defined regardles what
352 the actual X.509 library is (OpenSSL X.509 or something else). Other
353 choice is to write own X.509 library but I'm not going to do it -
354 I can help to migrate the OpenSSL X.509 into SILC and I can help if
355 someone would like to write the X.509 library - but I'm not going
356 to start writing one myself. Anyhow, the OpenSSL X.509 lib should
359 o SSH2 public keys support. Maybe - not really needed but could be
360 nice as SSH is widely used all over the place. SILC Protocol
361 supports SSH2 public keys.
363 o IRC support for SILC client. This would be nice to have on client
364 as it could be used to connect to SILC and IRC. People wouldn't
365 have to have two different clients when same would work on both.
366 I'd like to see this done as SIM, after the extended SIM support
367 has been added to SILC.
369 o Cipher optimizations (asm, that this) at least for i386 would be nice.