3 - "Forcing disconnect of XXX due to inactivity" -- can happen every few MS!?
4 (disconnect does not really succeed, or what?)
5 Also, core does not seem to actually decrement the number of active
6 sessions even if no connections exist for a while! (likely related...)
7 - transport reports bw quota violations on slow systems (core or
8 transport issue? tracking issue or actual violation?)
10 - Need more work on testcase of command-line tools!
11 - 2-peer download is still too slow (why? testcases fail on slow systems!)
13 - Get IPv6 hooked up [AK, after April 12th]
14 - change DNS [CG, need DNS]
18 - trust: need *fast* way to check/update trust in peers
19 (async peerinfo would not be right; certainly not with the
22 - scheduler should change OS process priority based on task priority;
23 should make better use of task priorities in general
24 - only connect() sockets that are ready (select()) [Nils]
25 [On W32, we need to select after calling socket before
28 - need function to test "equivalency" of HELLOs (or integrate with "merge"?); use in PEERINFO
30 - design & implement new setup tool
32 - good to have for transport/DV evaluation!
34 - write DV API (need to move declarations from dv_api.c to gnunet_dv_service.h!)
35 - implement DV service
36 - implement DV library (looks done)
37 - implement DV transport plugin
39 - implement performance tests
41 - does not seem to work with timeouts (especially if service is not running)
43 - needs more testing (especially F2F topology)
44 - needs to re-try connecting after disconnect (currently, it
45 initially triggers a connection request, but if that connection
46 fails / goes down, it does not retry in a timely fashion;
47 cause seems to be the 'blacklist_after_attempt' being set to 1h,
48 which is rather long -- and should probably be adjusted based on
49 the number of connections / known peers)
50 - If the topology daemon crashes, peers that were put on the
51 blacklist with transport will never be removed from it (until
52 transport service dies); we should use the blacklist notification
53 API to learn about the exact set of blacklisted peers at all times
54 (FIXME: the transport_api implementation of blacklisting
55 also does not work nicely for this since it won't let us know about
56 disconnect-reconnect events and the implicit whitelisting
57 that might happen here; that's not so bad since we will
58 re-blacklist on pre-connect attempts anyway, so this is
59 a minor issue; OTOH, we might want to be more explicit about
60 allowing/forbidding connects on pre-connect to avoid
61 entering connect attempts to just be blacklisted shortly afterwards).
62 - the code uses the term 'blacklist' for both peers that are forbidden
63 to connect (i.e. F2F mode) as well as peers that we currently
64 won't try to actively connect to ourselves (since we just tried);
65 This is confusing. We need two distinct terms (greylist?).
66 - move code to use hash table instead of linked list
67 - instead of periodically discarding blacklisted entries,
68 simply add task that is triggered at the right time (earlier free,
70 - check if new HELLO learned is different from old HELLO
71 before resetting entire state!
73 - support recursive download even if filename is NULL and we hence
74 do not generate files on disk (use temp_filename)
75 - bound parallelism (# fs downloads)
76 - distinguish in performance tracking and event signalling between
77 downloads that are actually running and those that are merely in the queue
78 - persistence support (publish, unindex, search, download)
79 - active migration support (in fs or in datastore or new daemon?)
80 - gnunet-service-fs (hot-path routing, load-based routing, nitpicks)
81 - [gnunet-service-fs.c:208]: member 'LocalGetContext::results_bf_size' is never used
82 - [gnunet-service-fs.c:501]: member 'PendingRequest::used_pids_size' is never used
83 - [gnunet-service-fs.c:654]: member 'ConnectedPeer::last_client_replies' is never used
84 - [gnunet-service-fs.c:669]: member 'ConnectedPeer::avg_delay' is never used
85 - [gnunet-service-fs.c:675]: member 'ConnectedPeer::avg_priority' is never used
86 - [gnunet-service-fs.c:688]: member 'ConnectedPeer::pending_requests' is never used
87 - [gnunet-service-fs.c:694]: member 'ConnectedPeer::last_p2p_replies_woff' is never used
88 - [gnunet-service-fs.c:700]: member 'ConnectedPeer::last_client_replies_woff' is never used
90 + active reply route caching design & implementation of service; gap extension!
92 - API lacks cancellation methods (needed? or is disconnect enough?);
93 may also want to integrate request queuing here instead of gnunet-service-fs_drq.c
95 - how to integrate scheduler with GTK event loop!?
99 - good to have for DV/DHT evaluation!
101 - implement DHT service
102 - implement DHT library
103 - implement testcases
104 - implement performance tests
108 - need to get rid of synchronous API for service starts (cause all kinds of problems)
109 [=> eliminate for need to tell ARM about service starts most of the time!] [Safey]
110 - better tracking of which config changes actually need to cause process restarts by ARM.
111 - listen for requests to discover dependencies between services (and avoid
112 having to explicitly program start requests)
113 - better crash management (attach debugging support, capture and analyze
114 debug output, detect random vs. deterministic crashes)
117 - datastore reservation (publishing)
118 - search: availability probes
119 - location URIs (publish, search, download)
120 - non-anonymous FS service (needs DHT)
121 + DHT integration for search
122 + CS-DHT-functions (DHT-put of LOC)
123 + P2P-functions (DHT-get)
124 - collection API & tests
125 + gnunet-pseudonym (collection support)
126 - implement FS performance tests
133 * Determine RC bugs and fix those!
135 - modify configuration to allow controlling connections for non-local starts
136 - testbed creation with topology (needs working F2F topology)
138 - implement testcases for distributed testing
139 - test basic peer re-configure
140 - test topology creation
141 - test churn generation
142 - consider changing API for peer-group termination to
143 call continuation when done
145 - finalize API design
148 - integration with transport service
149 * MYSQL database backends: [CG]
155 - expand bibliography
156 - convert documentation pages to books
157 - update books (especially for developers)
158 - create good Drupal theme for GNUnet
159 - make a NICE download page and figure out how to
160 enable developers to publish TGZs nicely
161 - port "contact" page
162 - add content type for "todo" items?
163 * POSTGRES database backends: [CG]
166 * Determine RC bugs and fix those!
170 - SMTP transport backend
171 - HTTPS transport backend
172 + improved HTTPS support in MHD
174 - Implement method of learning our external addresses from
175 other peers; need some kind of threshold-based
176 scheme, limiting both the total number of addresses that we accept
177 this way as well as requiring multiple confirmations; also, we
178 should possibly try to confirm that the given address works for
179 us ourselves (loopback-style) before adding it to the list
180 + we may be able to simplify WELCOME messages (no need to add
181 addresses there anymore, but may help to learn them there anyway...).
182 + we probably want some kind of voting/counting for learning IP addresses
183 (maybe including IP addresses in ads proportional to how often others
184 report them? we at least need some protection against >64k HELLOs!),
185 + provide a way to give the user a list of "learned" IP addresses and
186 a way to easily "veto" addresses off the list!
187 => If MiM attacker uses vetoed address, blacklist the specific IP for
188 the presumed neighbour!
189 - implement gnunet-transport (transport configurator / tester)
190 - UPnP-based IP detection
191 (Note: build library always, build service when libxml2/etc. are available)
193 - Remove KBlocks in gnunet-unindex (see discussion with Kenneth Almquist on gnunet-devs in 9/2009)
194 - Allow checking of presence of search results and/or content via command-line tools
195 (add options to gnunet-search / gnunet-download to limit search to local peer)
197 - expire 'ancient' HELLOs (those without valid addresses AND that
198 we have not 'used' (for their public keys) in a while; need a way
199 to track actual 'use')
200 - make sue we also trigger notifications whenever HELLOs expire
207 - should use hash map to look up sessions
209 - api creates many, many short-lived TCP connections; either some
210 clients should use the API differently or we need to change the
211 API to enable re-use of connections to the service
213 - should use BIO instead of mmap
215 - need to periodically probe latency/transport cost changes & possibly switch transport
216 - instantly filter addresses from *other* peers that
217 are *equal* to our own address + port (i.e., localhost:2086). We
218 no longer filter those for outgoing (helps with loopback testing
219 and keeps the code clean), but we should filter strictly *impossible*
220 incoming addresses! This is for efficiency, not correctness.
221 - should use hash map to look up Neighbours
223 - 'server' uses 'GNUNET_PEERINFO_iterate', should probably switch to notification API
224 (for more instant / up-to-date hostlists at lower cost) [OPTIMIZATION]
226 - check for duplicates on insertion (currently, same content is frequently
227 stored again [seen with KBLOCKS and SBLOCKS]!)
231 - repeatedly resolve hostname and look up interfaces to determine our own IP
232 - [./transport/plugin_transport_tcp.c:391]: (style) struct or union member 'Plugin::address_update_task' is never used (related to issue above)
234 - [./transport/gnunet-service-transport.c:173]: (style) struct or union member 'TransportPlugin::rebuild' is never used (related to TCP not refreshing external addresses?)
236 - add stats (# bytes available, # bytes used, # PUTs, # GETs, # GETs satisfied)
238 - support inline data in directories for recursive file downloads (fs_download)