Mercurial > dovecot > original-hg > dovecot-1.2
annotate TODO @ 8158:ad3ac0cb4573 HEAD
Updated TODO and NEWS.
author | Timo Sirainen <tss@iki.fi> |
---|---|
date | Fri, 05 Sep 2008 19:42:58 +0300 |
parents | f8de4a912916 |
children | a43b7da805f8 |
rev | line source |
---|---|
8158 | 1 - new primes code: are hash tables now being resized too often? |
2 - auth_log_prefix setting similar to mail_log_prefix | |
3 - LDAP attrs: uid=foo,uid=bar doesn't work | |
4 | |
5 - thread indexes: if we expunge a duplicate message-id: and we have a sibling | |
6 with identical message-id:, we can probably just move the children? | |
7 (unless there are non-sibling duplicates) | |
8 - SEARCH INTHREAD requires no thread sorting by date - don't do it | |
9 - virtual: keep MD5 of search parameters. if it changes, rebuild all. | |
10 - CONDSTORE: use per-flag/per-keyword conflict checking | |
11 - should modseq header be updated even when modseqs aren't used? otherwise | |
12 log reads always at beginning of file? are there other performance | |
13 issues? | |
14 - QRESYNC: Should we send HIGHESTMODSEQ in tagged reply for EXPUNGE | |
15 even when it was already returned in FETCH MODSEQs? | |
16 - QRESYNC: Drop expunges from the middle of given seq sets if possible | |
17 - use universal hash functions? | |
18 | |
19 - UIDVALIDITY changed while saving -> sync errors | |
20 - mbox: copy to Trash, manually delete copied msg, change uidvalidity, | |
21 set nextuid=1, copy again -> error | |
22 - recent_uids assert at least with mbox | |
23 - quota fs: Should values returned by quota be divided by the actual | |
24 filesystem block size instead of hardcoded DEV_BSIZE? not with AIX.. | |
25 - mailbox list fs: Listing subscriptions with children return options doesn't | |
26 work unless iter_is_mailbox() returns the children flags | |
7502 | 27 - sieve-cmu.c crash: i_assert(buf->used - 1 == part->body_size.physical_size); |
7449 | 28 - convert plugin: Create a r/w lock for a file. It's read-locked if |
29 conversion isn't wanted and released when process dies. If conversion is | |
30 wanted and write-lock succeeds, conversion is done, if write-lock doesn't | |
31 succeed it fallbacks to using the old storage. When process is exiting it | |
32 again tries to write-lock and do the conversion. Add a parameter that | |
33 specifies if conversion should be done. | |
7259 | 34 - lucene: handle replacement chars? |
35 - squat: | |
7380 | 36 - wrong indexid |
37 - fts_build_init() assertion failed: (last_uid < last_uid_locked) | |
7259 | 38 - is locking done right? it reads header without file being locked? |
39 - split after ~8 bytes? | |
7391 | 40 - expunges are delayed until more mails are added |
7259 | 41 - test replacement chars (SEARCH / SORT / Squat) |
6987 | 42 |
7575 | 43 - dbox: |
44 - "File unexpectedly lost" doesn't get fixed by itself | |
45 - Fix support for multi-message files | |
7603 | 46 - Delete dovecot-keywords and dovecot-uidlist after all maildir files |
47 have been converted to native dbox | |
7449 | 48 - DEBUG: buffer overflow checking code probably doesn't handle a successful |
49 t_try_realloc() or pool_alloconly_realloc() properly | |
6987 | 50 - cache: compress when we can drop temporary fields. |
51 - ACL: "foo/bar" in public namespace -> LIST "" % doesn't show "foo" | |
52 - new %modifier for reverse DNS lookups with a cache | |
53 - auth_gssapi_hostname = %Xl | |
54 - proxying would also want DNS lookups, but not reverse.. | |
7072 | 55 - ldap domain lookups which set the base for user lookup |
7502 | 56 - ldap: same attribute can't be used for multiple values. |
57 - ldap: multiple attributes can't be merged to a single value. | |
5621 | 58 |
7575 | 59 - Per-user options: |
60 - Deny deleting non-empty mailboxes | |
61 - Disable IDLE "still here" notifications | |
62 | |
6105 | 63 - maildir+pop3/deliver fast updates: |
64 - with locking enabled, pop3 could just keep the one and same sync lock and | |
65 do the whole thing using sync transaction | |
66 - don't update dovecot-uidlist if dovecot.index.cache doesn't exist / | |
67 there's nothing to cache | |
68 - if all messages are expunged and there are no unknown extensions in index, | |
69 unlink dovecot.index and rotate log and add some initial useful info to | |
70 the log (uidvalidity, nextuid) | |
5621 | 71 |
6105 | 72 - maildir |
6987 | 73 - don't allow more than 26 keywords |
6105 | 74 - physical separator could be configurable |
75 - setting flags to same as they already are causes unnecessary | |
76 rename()s with identical parameters | |
77 - deliver+maildir: if new mails are in new/ or cur/ they're not added to | |
78 dovecot-uidlist but newly saved mails are, so UIDs will be in wrong order | |
79 - maildir_copy_with_hardlinks: We're currently first hardlinking to tmp/ and | |
80 then rename()ing. This wouldn't be necessary if uidlist syncing noticed | |
81 that someone else already had added them to uidlist, and the existing UIDs | |
82 could be assigned to them in the index. | |
6754 | 83 - copying should copy already-cached data |
6105 | 84 - maildir_copy_preserve_filename=yes has a race condition causing "Append with |
85 UID n, but next_uid = y" errors when quota plugin is loaded. Practically | |
86 won't happen except in stress testing. | |
5621 | 87 |
6105 | 88 - mbox |
89 - UID renumbering doesn't really work after all? | |
90 - still problems with CRLF mboxes.. | |
91 | |
6987 | 92 - mail_log: log appends. support from and subject fields |
93 - proxy: If remote server disconnects on login: | |
94 login: tried to change state 2 -> 2 | |
6105 | 95 - logging consistency: |
96 http://www.dovecot.org/list/dovecot/2007-April/021532.html | |
97 - EXPUNGE command in read-only mailbox should give an error message if | |
98 there are messages marked as \Deleted? | |
99 - dovecot -o setting=something overriding | |
5557 | 100 - file_cache: we're growing the mmap in page size blocks, which is horribly |
101 slow if mremap() doesn't exist. | |
102 - login_max_processes_count shouldn't count proxying processes | |
103 | |
104 - Allow %variables in mail_chroot setting | |
105 - something should be done to umask setting. we should be creating files with | |
106 0666 or 0777 and rely on umask, but we shouldn't do that unless we're sure | |
107 that the umask is the wanted one (ie. imap/pop3 process wasn't started | |
108 directly) | |
109 - ssl_verify_client_cert isn't working if the SSL cert doesn't have CRL | |
110 | |
3087 | 111 - keywords: |
112 - add some limits to how many there can be | |
113 - don't return \* in PERMANENTFLAGS when we're full | |
114 - remove unused keywords? | |
2774 | 115 |
116 - caching | |
3904 | 117 - force bits should be used only for nonregistered fields |
118 - change envelope parsing not to use get_headers() so imap.envelope can | |
119 actually be cached without all the headers.. | |
6105 | 120 - if there's no other pressure for compression, we should do it when |
121 enough temp fields are ready to be dropped | |
3904 | 122 - we could try compressing same field values into a single |
123 location in cache file. | |
7259 | 124 - place some maximum limit of fields to cache file? maybe some soft and |
125 hard limits, so when soft limit is reached drop fields that have | |
126 been used only once. when hard limit is reached drop any fields to get | |
127 more space. all this to avoid cache file growing infinitely. | |
2774 | 128 |
3087 | 129 - mbox |
4449 | 130 - dirty state should be stored to index (with mbox_very_dirty_syncs) |
131 - after some locking timeouts: mbox-lock.c: line 518 (mbox_lock): | |
132 assertion failed: (lock_type == F_RDLCK || mbox->mbox_lock_type != F_RDLCK) | |
133 - controldir for mboxes too and place subscriptions file there? | |
3904 | 134 - syncing existing indexes takes 4x longer than creating new one, why? |
135 - how well does dirty sync + status work? it reads the last mail every | |
136 time? not very good.. | |
137 - always add empty line. make the parser require it too? syncing should | |
138 make sure there always exists two LFs at end of file. raw-mbox-stream | |
139 should make sure the last message ends with LF even if it doesn't exist | |
140 in the file | |
141 - Quote "From ", unquote ">From " | |
4449 | 142 - COPY doesn't work to itself (lock assert crash, for now just disallowed) |
2143 | 143 - if we can't create dotlock file for mbox, make sure it still can be |
144 selected in read-only state | |
145 | |
146 - index | |
3087 | 147 - read-only support for mailboxes where we don't have write-access |
6105 | 148 - index file format changes: |
149 - pack UIDs to beginning of file with UID ranges | |
150 - use squat-like compressed uid ranges everywhere | |
6782 | 151 - write first extension intros in dovecot.index.log always with names |
152 - or better yet, drop the intro concept completely as it is now | |
153 - add "transaction boundaries" so we know which records belong to a | |
154 single transaction. | |
155 - only after that we can remove the transaction log offset | |
156 overwriting (otherwise we can get partial transactions in views) | |
2143 | 157 |
4449 | 158 - namespaces |
159 - namespaces: add new "auto_disable" flag so if the mailbox can't be opened | |
160 (eg. file doesn't exist), just ignore the problem and disable the | |
161 namespace | |
162 | |
163 - lib-storage | |
5557 | 164 - rename: allow moving between storages, as long as they're of same type |
4449 | 165 - x search charset asdf all -> should fail |
3087 | 166 - should we allow following symlinks in mbox/maildirs? they are now. |
167 - if we implement shared mailboxes with shared indexes, never do that or | |
168 others could symlink your personal mailboxes and see the indexes | |
169 created for it which may contain envelope etc. data | |
170 - this allows circular mailbox hierarchies which should be prevented by | |
171 eg. allowing max. 20 hierarchies. | |
172 | |
173 - login | |
4449 | 174 - imap-login: Master sent reply with unknown tag 1. client closed |
175 connection at the exact same time master was logging it in? | |
176 see master_request_abort() | |
3087 | 177 - Digest-MD5: support integrity protection, and maybe crypting. Do it |
178 through login process like SSL is done? | |
179 - x login foo bar | |
180 x NO Authentication failed. | |
181 x login cras pass | |
182 * BYE Disconnected for inactivity. | |
183 ^ but it's not disconnecting! (buggy dovecot-auth not replying) | |
3744 | 184 probably because userdb lookup didn't reply, and fd was already sent |
185 for master.. should imap-login be handling it anymore?.. | |
3087 | 186 - imap-login: Authenticate PLAIN failed: Authentication failed: |
187 Authentication server isn't connected, try again later.. [127.0.0.1] | |
188 ^ NO Authentication failed. (should be Temporary login failure!) | |
189 - if auth process dies, login process should retry authentication if | |
190 possible. or if not, disconnect the client so it doesn't think the auth | |
191 failed. | |
192 - send client IP immediately after accept() to master process. make sure | |
193 master shows the IP if login dies unexpectedly. master should probably | |
194 also kill the login process if it doesn't kill itself soon enough.. or | |
195 maybe just log the IP immediately. | |
474 | 196 |
3087 | 197 - auth |
4449 | 198 - with blocking passdb we're not caching lookups if the password was wrong |
5557 | 199 - non-plaintext authentication doesn't support all features: |
200 - multiple passdbs don't work, only the first one is used | |
201 - auth cache's last_success password change check doesn't exist | |
6987 | 202 - allow returning multiple gids |
3087 | 203 - SIGHUP restarts auth processes .. but does it wait until they've finished |
204 with all requests? no. | |
205 - does dovecot-auth really break when it runs out of fds? | |
4449 | 206 - dovecot-auth should limit how fast authentication requests are allowed |
207 from login processes. especially if there's one login/connection the speed | |
3087 | 208 should be something like once/sec. also limit how fast to accept new |
209 connections. | |
210 - support read-only logins. user could with alternative password get only | |
211 read-access to mails so mails could be read relatively safely with | |
212 untrusted computers. Maybe always send [ALERT] about the previous | |
213 read-only login time with IP? | |
4449 | 214 - dovecot-auth workers: create a separate dovecot-pam worker which shares |
215 pretty much all code with dovecot-auth worker but isn't linked against | |
216 any libraries. or..? this might be difficult to do, especially because the | |
217 workers currently can handle any kind of passdb/userb requests. perhaps | |
218 there should be a completely separate simple PAM authenticator binary. | |
526 | 219 |
3087 | 220 - master |
221 - configurable syslog prefix | |
3450 | 222 - if there are duplicate settings, complain about it |
3087 | 223 |
224 - quota | |
225 - if dovecot-uidlist can't be written, assume the new mails have UIDs | |
226 beginning from uidlist.next_uid. Whenever mails are expunged, overwrite | |
227 the next_uid field with the current highest next_uid. Whenever we have | |
228 assumed UIDs and uidlist gets updated, throw the client out with | |
229 "inconsist mailbox". | |
230 | |
231 - ssl | |
232 - add setting: ssl_options = bitmask. by default we enable all openssl | |
233 workarounds, this could be used to disable some of them | |
234 - gnutls support isn't working | |
965 | 235 |
3087 | 236 - search |
6232 | 237 - message header search: we should ignore LWSP between two MIME blocks(?) |
238 - message_search_init() could accept multiple search keywords so we | |
1250 | 239 wouldn't need to call it separately for each one (so we wouldn't need |
240 to parse the message multiple times). | |
241 - could optionally support scanning inside file attachments and use | |
242 plugins to extract text out of them (word, excel, pdf, etc. etc.) | |
243 - Create our own extension: When searching with TEXT/BODY, return | |
244 the message text surrounding the keywords just like web search engines | |
245 do. like: SEARCH X-PRINT-MATCHES TEXT "hello" -> * SEARCH 1 "He said: | |
246 Hello world!" 2 "Hello, I'm ...". This would be especially useful with | |
247 the above attachment scanning. | |
0 | 248 |
3087 | 249 - general |
4449 | 250 - stop using atol(), atoi(), strtoul() etc. in places where we actually |
251 care about what they return, and rather create our own function which | |
252 checks if the input overflows the integer, and if so call i_fatal() | |
253 - LMTP server | |
254 - ability to build plugins statically into the binaries | |
3087 | 255 - ~/.dovecotrc to override system wide settings. namespace settings should |
256 override all the previous namespace settings instead of adding new. | |
257 - things break if next_uid gets to 2^32 | |
376
fd1fc4cf11b7
updated with lots of new capability comments
Timo Sirainen <tss@iki.fi>
parents:
340
diff
changeset
|
258 |
fd1fc4cf11b7
updated with lots of new capability comments
Timo Sirainen <tss@iki.fi>
parents:
340
diff
changeset
|
259 capabilities: |
3087 | 260 - THREAD=ORDEREDSUBJECT - although pretty useless I'd think. |
5621 | 261 - acl (rfc4314) |
376
fd1fc4cf11b7
updated with lots of new capability comments
Timo Sirainen <tss@iki.fi>
parents:
340
diff
changeset
|
262 - mailbox-referrals (rfc2193) |
fd1fc4cf11b7
updated with lots of new capability comments
Timo Sirainen <tss@iki.fi>
parents:
340
diff
changeset
|
263 - this is useful whenever we would otherwise need to make the |
fd1fc4cf11b7
updated with lots of new capability comments
Timo Sirainen <tss@iki.fi>
parents:
340
diff
changeset
|
264 connection ourself. for example load balancing and shared mailboxes |
fd1fc4cf11b7
updated with lots of new capability comments
Timo Sirainen <tss@iki.fi>
parents:
340
diff
changeset
|
265 requiring another UID to run. |
fd1fc4cf11b7
updated with lots of new capability comments
Timo Sirainen <tss@iki.fi>
parents:
340
diff
changeset
|
266 - this rfc defines no exact way for server to detect if client |
fd1fc4cf11b7
updated with lots of new capability comments
Timo Sirainen <tss@iki.fi>
parents:
340
diff
changeset
|
267 supports referrals or not. I don't think there's much point in |
fd1fc4cf11b7
updated with lots of new capability comments
Timo Sirainen <tss@iki.fi>
parents:
340
diff
changeset
|
268 supporting only referrals, as most clients don't support them. |
fd1fc4cf11b7
updated with lots of new capability comments
Timo Sirainen <tss@iki.fi>
parents:
340
diff
changeset
|
269 Instead we should return referrals when we know that client |
fd1fc4cf11b7
updated with lots of new capability comments
Timo Sirainen <tss@iki.fi>
parents:
340
diff
changeset
|
270 supports them, otherwise do the connecting ourself. If client |
fd1fc4cf11b7
updated with lots of new capability comments
Timo Sirainen <tss@iki.fi>
parents:
340
diff
changeset
|
271 issues RLIST or RLSUB command, it's safe to assume it supports |
fd1fc4cf11b7
updated with lots of new capability comments
Timo Sirainen <tss@iki.fi>
parents:
340
diff
changeset
|
272 referrals. |
fd1fc4cf11b7
updated with lots of new capability comments
Timo Sirainen <tss@iki.fi>
parents:
340
diff
changeset
|
273 - for load balancing this works just fine, but what about shared |
fd1fc4cf11b7
updated with lots of new capability comments
Timo Sirainen <tss@iki.fi>
parents:
340
diff
changeset
|
274 mailboxes which require different UID? If we login with our own |
fd1fc4cf11b7
updated with lots of new capability comments
Timo Sirainen <tss@iki.fi>
parents:
340
diff
changeset
|
275 username, we end up with our own UID instead of what we wanted. |
fd1fc4cf11b7
updated with lots of new capability comments
Timo Sirainen <tss@iki.fi>
parents:
340
diff
changeset
|
276 IMAP URLs don't support separated authorization id which would |
fd1fc4cf11b7
updated with lots of new capability comments
Timo Sirainen <tss@iki.fi>
parents:
340
diff
changeset
|
277 have made this very easy.. We could give the "userid@group" as |
fd1fc4cf11b7
updated with lots of new capability comments
Timo Sirainen <tss@iki.fi>
parents:
340
diff
changeset
|
278 userid, but clients probably treat it as different userid and |
fd1fc4cf11b7
updated with lots of new capability comments
Timo Sirainen <tss@iki.fi>
parents:
340
diff
changeset
|
279 ask the password again. |
fd1fc4cf11b7
updated with lots of new capability comments
Timo Sirainen <tss@iki.fi>
parents:
340
diff
changeset
|
280 - problems, problems, .. maybe not worth the trouble. |
fd1fc4cf11b7
updated with lots of new capability comments
Timo Sirainen <tss@iki.fi>
parents:
340
diff
changeset
|
281 - drafts: |
fd1fc4cf11b7
updated with lots of new capability comments
Timo Sirainen <tss@iki.fi>
parents:
340
diff
changeset
|
282 - http://www.imc.org/ids.html |
fd1fc4cf11b7
updated with lots of new capability comments
Timo Sirainen <tss@iki.fi>
parents:
340
diff
changeset
|
283 - annotate (draft-ietf-imapext-annotate) |
fd1fc4cf11b7
updated with lots of new capability comments
Timo Sirainen <tss@iki.fi>
parents:
340
diff
changeset
|
284 - per-message annotations. this will be major change. especially |
fd1fc4cf11b7
updated with lots of new capability comments
Timo Sirainen <tss@iki.fi>
parents:
340
diff
changeset
|
285 because currently there's no suitable storage for them, and |
fd1fc4cf11b7
updated with lots of new capability comments
Timo Sirainen <tss@iki.fi>
parents:
340
diff
changeset
|
286 they'll probably change all the time.. maybe if we moved into |
fd1fc4cf11b7
updated with lots of new capability comments
Timo Sirainen <tss@iki.fi>
parents:
340
diff
changeset
|
287 berkeley db to store the .data file and these annotations. |
1178 | 288 - this is separate problem from index files. indexes are treated as |
289 temporary files, annotations are permanent data. we'd have to | |
290 support non-db way to do this too, which would probably be just a | |
291 simple (slow) text file. | |
5621 | 292 - use lib-dict probably |
293 - metadata (draft-daboo-imap-annotatemore) | |
376
fd1fc4cf11b7
updated with lots of new capability comments
Timo Sirainen <tss@iki.fi>
parents:
340
diff
changeset
|
294 - server and per-mailbox annotations. much easier than |
fd1fc4cf11b7
updated with lots of new capability comments
Timo Sirainen <tss@iki.fi>
parents:
340
diff
changeset
|
295 per-message annotations, but they'd be easier to place into |
fd1fc4cf11b7
updated with lots of new capability comments
Timo Sirainen <tss@iki.fi>
parents:
340
diff
changeset
|
296 db as well. |
5621 | 297 - lib-dict, again |
376
fd1fc4cf11b7
updated with lots of new capability comments
Timo Sirainen <tss@iki.fi>
parents:
340
diff
changeset
|
298 - binary (draft-nerenberg-imap-binary) |