2 * TRANSPORT: inbound sessions are not properly PING-PONG validated,
3 connect-event is then not properly signalled (peer behind NAT
4 never gets a 'connect' message) and finally we DO transmit
5 data messages over the invalidated, only half-connect-signalled
6 channel (which confuses the core state machine, or now causes
7 transport-api assertion failures). [CG]
8 => now fixed, but needs testing!
9 * FS-acceptance testing [CG]
19 - trust: need *fast* way to check/update trust in peers
20 (async peerinfo would not be right; certainly not with the
23 - scheduler should change OS process priority based on task priority;
24 should make better use of task priorities in general
25 - only connect() sockets that are ready (select()) [Nils]
26 [On W32, we need to select after calling socket before
29 - design & implement new setup tool
31 - good to have for transport/DV evaluation!
33 - write DV API (need to move declarations from dv_api.c to gnunet_dv_service.h!)
34 - implement DV service
35 - implement DV library (looks done)
36 - implement DV transport plugin
38 - implement performance tests
40 - does not seem to work with timeouts (especially if service is not running)
42 - needs more testing (especially F2F topology)
43 - needs to re-try connecting after disconnect (currently, it
44 initially triggers a connection request, but if that connection
45 fails / goes down, it does not retry in a timely fashion;
46 cause seems to be the 'blacklist_after_attempt' being set to 1h,
47 which is rather long -- and should probably be adjusted based on
48 the number of connections / known peers)
49 - If the topology daemon crashes, peers that were put on the
50 blacklist with transport will never be removed from it (until
51 transport service dies); we should use the blacklist notification
52 API to learn about the exact set of blacklisted peers at all times
53 (FIXME: the transport_api implementation of blacklisting
54 also does not work nicely for this since it won't let us know about
55 disconnect-reconnect events and the implicit whitelisting
56 that might happen here; that's not so bad since we will
57 re-blacklist on pre-connect attempts anyway, so this is
58 a minor issue; OTOH, we might want to be more explicit about
59 allowing/forbidding connects on pre-connect to avoid
60 entering connect attempts to just be blacklisted shortly afterwards).
61 - the code uses the term 'blacklist' for both peers that are forbidden
62 to connect (i.e. F2F mode) as well as peers that we currently
63 won't try to actively connect to ourselves (since we just tried);
64 This is confusing. We need two distinct terms (greylist?).
65 - move code to use hash table instead of linked list
66 - instead of periodically discarding blacklisted entries,
67 simply add task that is triggered at the right time (earlier free,
69 - check if new HELLO learned is different from old HELLO
70 before resetting entire state!
72 - support recursive download even if filename is NULL and we hence
73 do not generate files on disk (use temp_filename)
74 - bound parallelism (# fs downloads)
75 - distinguish in performance tracking and event signalling between
76 downloads that are actually running and those that are merely in the queue
77 - persistence support (publish, unindex, search, download)
78 - active migration support (in fs or in datastore or new daemon?)
79 - gnunet-service-fs (hot-path routing, load-based routing, nitpicks)
80 - [gnunet-service-fs.c:208]: member 'LocalGetContext::results_bf_size' is never used
81 - [gnunet-service-fs.c:501]: member 'PendingRequest::used_pids_size' is never used
82 - [gnunet-service-fs.c:654]: member 'ConnectedPeer::last_client_replies' is never used
83 - [gnunet-service-fs.c:669]: member 'ConnectedPeer::avg_delay' is never used
84 - [gnunet-service-fs.c:675]: member 'ConnectedPeer::avg_priority' is never used
85 - [gnunet-service-fs.c:688]: member 'ConnectedPeer::pending_requests' is never used
86 - [gnunet-service-fs.c:694]: member 'ConnectedPeer::last_p2p_replies_woff' is never used
87 - [gnunet-service-fs.c:700]: member 'ConnectedPeer::last_client_replies_woff' is never used
89 + active reply route caching design & implementation of service; gap extension!
91 - API lacks cancellation methods (needed? or is disconnect enough?);
92 may also want to integrate request queuing here instead of gnunet-service-fs_drq.c
94 - use g_main_context_set_poll_func to integrate GTK with GNUnet Scheduler!? (YUCK!)
95 - OR: add scheduler API to enable integration with GTK main loop instead of doing our own select
96 - use g_main_context_pending, g_main_context_query / g_main_context_check / g_main_context_dispatch
97 and NEVER g_main_loop_run (can this be done? might be the clean way to do this! But how
98 to integrate this with "tak_main"? Docu says:
99 "It's OK to use the GLib main loop directly instead of gtk_main(), though it involves
100 slightly more typing. See GMainLoop in the GLib documentation."
101 => so maybe it "just works"?
105 - good to have for DV/DHT evaluation!
107 - implement DHT service
108 - implement DHT library
109 - implement testcases
110 - implement performance tests
114 - need to get rid of synchronous API for service starts (cause all kinds of problems)
115 [=> eliminate for need to tell ARM about service starts most of the time!] [Safey]
116 - better tracking of which config changes actually need to cause process restarts by ARM.
117 - listen for requests to discover dependencies between services (and avoid
118 having to explicitly program start requests)
119 - better crash management (attach debugging support, capture and analyze
120 debug output, detect random vs. deterministic crashes)
123 - datastore reservation (publishing)
124 - search: availability probes
125 - location URIs (publish, search, download)
126 - non-anonymous FS service (needs DHT)
127 + DHT integration for search
128 + CS-DHT-functions (DHT-put of LOC)
129 + P2P-functions (DHT-get)
130 - collection API & tests
131 + gnunet-pseudonym (collection support)
132 - implement FS performance tests
139 * Determine RC bugs and fix those!
141 - modify configuration to allow controlling connections for non-local starts
142 - testbed creation with topology (needs working F2F topology)
144 - implement testcases for distributed testing
145 - test basic peer re-configure
146 - test topology creation
147 - test churn generation
148 - consider changing API for peer-group termination to
149 call continuation when done
151 - finalize API design
154 - integration with transport service
155 * MYSQL database backends: [CG]
161 - expand bibliography
162 - convert documentation pages to books
163 - update books (especially for developers)
164 - create good Drupal theme for GNUnet
165 - make a NICE download page and figure out how to
166 enable developers to publish TGZs nicely
167 - port "contact" page
168 - add content type for "todo" items?
169 * POSTGRES database backends: [CG]
172 * Determine RC bugs and fix those!
176 - SMTP transport backend
177 - HTTPS transport backend
178 + improved HTTPS support in MHD
180 - Implement method of learning our external addresses from
181 other peers; need some kind of threshold-based
182 scheme, limiting both the total number of addresses that we accept
183 this way as well as requiring multiple confirmations; also, we
184 should possibly try to confirm that the given address works for
185 us ourselves (loopback-style) before adding it to the list
186 + we may be able to simplify WELCOME messages (no need to add
187 addresses there anymore, but may help to learn them there anyway...).
188 + we probably want some kind of voting/counting for learning IP addresses
189 (maybe including IP addresses in ads proportional to how often others
190 report them? we at least need some protection against >64k HELLOs!),
191 + provide a way to give the user a list of "learned" IP addresses and
192 a way to easily "veto" addresses off the list!
193 => If MiM attacker uses vetoed address, blacklist the specific IP for
194 the presumed neighbour!
195 - implement gnunet-transport (transport configurator / tester)
196 - UPnP-based IP detection
197 (Note: build library always, build service when libxml2/etc. are available)
199 - Remove KBlocks in gnunet-unindex (see discussion with Kenneth Almquist on gnunet-devs in 9/2009)
200 - Allow checking of presence of search results and/or content via command-line tools
201 (add options to gnunet-search / gnunet-download to limit search to local peer)
203 - expire 'ancient' HELLOs (those without valid addresses AND that
204 we have not 'used' (for their public keys) in a while; need a way
205 to track actual 'use')
206 - make sue we also trigger notifications whenever HELLOs expire
213 - should use hash map to look up sessions
215 - api creates many, many short-lived TCP connections; either some
216 clients should use the API differently or we need to change the
217 API to enable re-use of connections to the service
219 - should use BIO instead of mmap
221 - need to periodically probe latency/transport cost changes & possibly switch transport
222 - instantly filter addresses from *other* peers that
223 are *equal* to our own address + port (i.e., localhost:2086). We
224 no longer filter those for outgoing (helps with loopback testing
225 and keeps the code clean), but we should filter strictly *impossible*
226 incoming addresses! This is for efficiency, not correctness.
227 - should use hash map to look up Neighbours
229 - 'server' uses 'GNUNET_PEERINFO_iterate', should probably switch to notification API
230 (for more instant / up-to-date hostlists at lower cost) [OPTIMIZATION]
232 - check for duplicates on insertion (currently, same content is frequently
233 stored again [seen with KBLOCKS and SBLOCKS]!)
237 - repeatedly resolve hostname and look up interfaces to determine our own IP
238 - [./transport/plugin_transport_tcp.c:391]: (style) struct or union member 'Plugin::address_update_task' is never used (related to issue above)
240 - [./transport/gnunet-service-transport.c:173]: (style) struct or union member 'TransportPlugin::rebuild' is never used (related to TCP not refreshing external addresses?)
242 - add stats (# bytes available, # bytes used, # PUTs, # GETs, # GETs satisfied)
244 - support inline data in directories for recursive file downloads (fs_download)