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