1 TODO for 1.1 And Beyond
2 =======================
4 NOTE: Any item that doesn't have (***DONE) in it, isn't done yet. The
5 (***TESTING NEEDED) means that the item has been done but not yet properly
9 lib/silcclient, The Client Library
10 ==================================
12 o silcclient.h clean up and API rewrites.
14 o silcclient_entry.h finishing, all entry relates APIs to this header.
16 o SilcChannelEntry, SilcServerEntry, SilcChannelUser, allocating,
17 freeing, finding, etc. rewrite. Also making them reference counted for
20 o Finish all the missing SILC packet processings, rewrites.
22 o The client_notify.c rewrite.
24 o Resuming to client_register.c (remove client_resume.c)
28 o Remove protocol.[ch].
30 o File transfer rewrite.
32 o Starting key exchange directly, rewrite.
34 o Channel messages, channel private keys, channel entires, channel
37 o For many APIs leave the hash context allocations to the caller instead
38 of using client->sha1hash and client->md5hash, or some kind of thread
39 safe (no locking) concept.
41 o Password auth test, public key auth test.
43 o Key agreement rewrite.
45 o Connecting to remote client, peer-to-peer private messages
47 o Private message waiting API (in threads)
49 o client_attrs.c, attributes rewrite.
51 o No SilcBuffer lists back to application in command_reply operations.
52 Convert them all to real lists and/or structures for easier use.
54 o Nickname formatting rewrite.
58 o Remove silc_client_run and silc_client_run_one from calling SilcSchedule.
59 Leave the scheduling entirely to application.
61 o All packet waiting timeout tests and error condition tests.
64 lib/silccore/silcpacket.[ch] ****DONE****
65 ============================
67 o SilcPacketEngine. (***DONE)
69 o New SILC Packet API. (***DONE)
72 lib/silccore/silcpacket.[ch] ****PARTLY DONE****
73 ============================
75 o IV Included flag support, UDP transport support (***TESTING NEEDED)
78 lib/silccore/silcid.[ch] ****DONE****
79 ========================
81 o Add silc_id_str2id to accept the destination buffer as argument
82 and thus not require any memory allocation. Same will happen
83 with silc_id_payload_* functions. (***DONE)
85 o silc_id_str2id, silc_id2str to non-allocating routines. (***DONE)
91 o All payload encoding routines should take SilcStack as argument.
93 o All payload test routines into lib/silccore/tests/.
99 o Removing key from the repository is not possible currently. It should
103 lib/silcske/silcske.[ch]
104 ========================
106 o IV Included flag support in SKE (***TESTING NEEDED)
108 o UDP transport changes; retransmission support by using exponential
111 o SilcConnAuth header file documentation.
114 lib/silccrypt ****PARTLY DONE****
117 o Implement the defined SilcDH API. The definition is in
118 lib/silccrypt/silcdh.h.
120 o SSH2 public keys support, allowing the use of SSH2 public keys in
123 o OpenPGP certificate support, allowing the use of PGP public keys
126 o SILC PKCS (silcpkcs.h) reorganizing when other PK supports added.
127 Move the SILC Public Key routines away from the crypto library into
128 the core library (silccore). silc_pkcs_public/private_key_* routines
129 to silc_public/private_key_* routines. The silc_public_key_* routines
130 should also automatically handle SILC Public Keys, and other keys
131 and certificates as well. Add fe. silcpk.h into silccore. It should
132 also include the Public Key Payload encoding and decoding routines.
137 o Cipher optimizations (asm, that this) at least for i386 would be nice.
142 lib/silcutil ****PARTLY DONE****
145 o silc_stringprep to non-allocating version.
147 o Compression routines are missing. The protocol supports packet
148 compression thus it must be implemented. SILC Zip API must be
151 o Add builtin SOCKS and HTTP Proxy support, well the SOCKS at least.
152 SILC currently supports SOCKS4 and SOCKS5 but it needs to be compiled
155 o bool -> SilcBool (***DONE)
157 o Silc FD Stream to WIN32 (lib/silcutil/silcfdstream.h)
160 lib/silcutil/silcbuffer.h ****DONE****
161 =========================
163 o Remove the `truelen' field from SilcBuffer as it is entirely
164 redundant since we can get the true length of the buffer by
165 doing buffer->end - buffer->header. Add silc_buffer_truelen
166 macro instead. Consider also removing `len' field too since
167 it effectively is buffer->tail - buffer->data, and adding
168 silc_buffer_len macro can do the same. These would save
169 totally 8 bytes of memory per buffer. (***DONE)
172 lib/silcutil/silcbuffmt.[ch] ****PARTLY DONE****
173 ============================
175 o SilcStack aware silc_buffer_unformat (***DONE)
177 o SilcStack aware silc_buffer_format (***DONE)
179 o silc_buffer_format reallocates automatically (***DONE)
181 o SILC_STR_OFFSET (***DONE)
184 lib/silcutil/silcstack.[ch] ****DONE****
185 ===========================
187 o Data stack implementation (***DONE)
190 lib/silcutil/silcstream.[ch] ****DONE****
191 ============================
193 o Add abstract SilcStream. (***DONE)
196 lib/silcutil/silcsocketstream.[ch] ****PARTY DONE****
197 ==================================
199 o Add SilcSocketStream (***DONE)
201 o Add SilcSocketStream for WIN32
203 o Handle EOS sending to upper layer properly
205 o Test QoS after the changes made to socket stream
208 lib/silcutil/silcschedule*.[ch] ****DONE****
209 ===============================
211 o Scheduler can be optimized for FD tasks by changing the fd_queue
212 to SilcHashTable instead of using linked list. We need to do
213 one-to-one mapping of FD to task and hash table is more efficient
216 Also redefine the silc_select to perhaps return a separate
217 structure of the events that actually occurred, instead of
218 returning the events in the fd_list which is then traversed
219 in the generic code to find the changed events. This can be
220 made faster by having own struct which includes only the
221 changed events, thus the tarversing is faster since the whole
222 fd_list is not traversed anymore (it is still traversed in the
223 silc_select but at least it removes one extra tarversing later
226 Other task queues should be changed to use SilcList. (***DONE)
228 o Add SILC scheduler's internal routines into a table of implementation
229 function pointers, that the generic code then takes as extern from
230 implementation. These are the silc_schedule_internal_* routines.
233 o Change SILC_TASK_CALLBACK to non-static, and remove the macro
234 SILC_TASK_CALLBACK_GLOBAL. (***DONE)
236 o SILC Schedule API changes to WIN32.
239 lib/silcutil/silcasync.[ch] ****DONE****
240 ===========================
242 o Add SilcAsyncOperation to utility library. Any function that takes
243 callback as an argument must/should return SilcAsyncOperation.
247 lib/silcutil/silctime.[ch] ****DONE****
248 ===========================
250 o SilcTime. (***DONE)
252 o system time, universal, generalized. (***DONE)
255 lib/silcutil/silcfsm.[ch] ****DONE****
256 =========================
258 o SILC Finite State Machine API. Replaces SILC Protocol API (***DONE)
261 lib/silcutil/silcnet*, lib/silcutil/*/silc*net* ****PARTLY DONE****
262 ===============================================
264 o Add UDP interface (***DONE)
266 o Add UDP interface for WIN32
268 o New network interfaces (***DONE)
274 o Porting to use the new util library.
280 o lib/silcutil/epoc routines missing or not completed.
282 o The PKCS#1 also calls global RNG (even though it is not used
283 currently in SILC, the interface allows its use).
285 o Something needs to be thought to the logging globals as well,
286 like silc_debug etc. They won't work on EPOC. Perhaps logging
287 and debugging is to be disabled on EPOC.
293 o Import TFM. Talk to Tom to add the missing functions. Use TFM in
294 client and client library, but TMA in server, due to the significantly
295 increased memory consumption with TFM, and the rare need for public
296 key operations in server.
298 o Change LTM and TFM function names when importing to SILC tree to avoid
299 rare linking problems on system that has same named symbols already in
302 o The SILC MP API function must start returning indication of success
303 and failure of the operation.
305 o Do SilcStack support for silc_mp_init, silc_mp_init_size and other
306 any other MP function (including utility ones) that may allocate
311 o All utility functions should be made non-allocating ones.
314 lib/silcasn1 ****PARTLY DONE****
317 o ASN.1 library (***DONE)
319 o Header documentation missing. (***DONE)
321 o Some string encodings missing (copy/paste matter). (***DONE)
323 o Negative integer encoding
329 o PKIX implementation
335 o (Re)write commands/command replys.
337 o (Re)write notify handling.
339 o The SERVER_SIGNOFF notify handing is not optimal, because it'll
340 cause sending of multiple SIGNOFF notify's instead of the one
341 SERVER_SIGNOFF notify that the server received. This should be
342 optimized so that the only SERVER_SIGNOFF is sent and not
343 SIGNOFF of notify at all (using SIGNOFF takes the idea about
344 SERVER_SIGNOFF away entirely).
346 o Another SERVER_SIGNOFF opt/bugfix: Currently the signoff is
347 sent to a client if it is on same channel as the client that
348 signoffed. However, the entire SERVER_SIGNOFF list is sent to
349 the client, ie. it may receive clients that was not on the
350 same channel. This is actually against the specs. It must be
351 done per channel. It shouldn't receive the whole list just
352 because one client happened to be on same channel.
354 o MAYBE: The SilcChannelClientEntry can be:
358 where address is SilcClientEntry address XOR SilcChannelEntry.
359 You can get SilcClientEntry by doing client = chl->address XOR channel,
360 and SilcChannelEntry by doing channel = chl->address XOR client.
361 As long as the other pointer is always available when accessing the
362 structure this can be done.
364 o Add reference counters to all Silc*Entry structures
366 o SERVICEs support (plugin, SIM)
368 o If client's public key is saved in the server (and doing public key
369 authentication) then the hostname and the username information could
370 be taken from the public key. Should be a configuration option!
372 o Add a timeout to handling incoming JOIN commands. It should be
373 enforced that JOIN command is executed only once in a second or two
374 seconds. Now it is possible to accept n incoming JOIN commands
375 and process them without any timeouts. THis must be employed because
376 each JOIN command will create and distribute the new channel key
377 to everybody on the channel (Fix this to 0.9.x).
379 o Related to above. If multiple JOINs are received in sequence perhaps
380 new key should be created only once, if the JOINs are handeled at the same
381 time. Now we create multiple keys and never end up using them because
382 many JOINs are processed at the same time in sequence. Only the last
383 key ends up being used.
385 o The CMODE cipher & hmac change problem (#101).