1 Implementable right now (but not necessarily important), with caveats
2 (unavailable components that will limit what can be implemented right
3 away), in order in which they will likely be done:
4 * TESTING-TESTBED [Nate]
9 * FRAGMENTATION [Ji Lu]
11 * MySQL / Postgres plugins (datastore, datacache)
18 - "Forcing disconnect of XXX due to inactivity" -- can happen every few MS!?
19 (disconnect does not really succeed, or what?)
20 Also, core does not seem to actually decrement the number of active
21 sessions even if no connections exist for a while! (likely related...)
22 - transport reports bw quota violations on big-endian systems (core or
25 - utilization can (easily, restart?) go out of control (very large), causing
26 content expiration job to go crazy and delete everything!
28 - gnunet-publish cannot be aborted using CTRL-C
29 - on some systems, keyword search does not find locally published content
30 (need testcase of command-line tools! - also good to cover getopt API!)
31 [could be related to datastore issue above!]
32 - 2-peer download is still too slow (why?)
33 - advanced FS API parts
34 + search: SBlocks, NBlocks, probes, notify FS-service of known results
35 + gnunet-publish (URI args)
36 + gnunet-download (directory)
37 + gnunet-search (options, incl. namespace search)
38 + gnunet-download (many options)
39 + gnunet-directory (man page, options)
40 + gnunet-pseudonym (listing of namespaces, advertisement uris, testing)
41 + gnunet-service-fs (remove failing on-demand blocks, hot-path routing,
42 load-based routing, nitpicks)
43 + indexing: index-failure-cleanup
44 - implement adv. FS testcases
45 - [./fs/gnunet-service-fs.c:208]: (style) struct or union member 'LocalGetContext::results_bf_size' is never used
46 - [./fs/gnunet-service-fs.c:501]: (style) struct or union member 'PendingRequest::used_pids_size' is never used
47 - [./fs/gnunet-service-fs.c:654]: (style) struct or union member 'ConnectedPeer::last_client_replies' is never used
48 - [./fs/gnunet-service-fs.c:669]: (style) struct or union member 'ConnectedPeer::avg_delay' is never used
49 - [./fs/gnunet-service-fs.c:675]: (style) struct or union member 'ConnectedPeer::avg_priority' is never used
50 - [./fs/gnunet-service-fs.c:688]: (style) struct or union member 'ConnectedPeer::pending_requests' is never used
51 - [./fs/gnunet-service-fs.c:694]: (style) struct or union member 'ConnectedPeer::last_p2p_replies_woff' is never used
52 - [./fs/gnunet-service-fs.c:700]: (style) struct or union member 'ConnectedPeer::last_client_replies_woff' is never used
58 * Module features to implement:
60 + good to have for DV evaluation! [MW]
63 + implement DV service [Nate & CG]
64 + implement DV library [Nate]
65 + implement DV transport plugin [Nate & CG]
66 + implement testcases [Nate]
67 + implement performance tests [Nate]
69 - design & implement new setup tool
72 + how to integrate scheduler with GTK event loop!
74 - does not seem to work with timeouts (especially if service is not running)
76 - trust: need *fast* way to check/update trust in peers
77 (async peerinfo would not be right; certainly not with the
80 - scheduler should change OS process priority based on task priority;
81 should make better use of task priorities in general
82 - only connect() sockets that are ready (select()) [Nils]
83 [On W32, we need to select after calling socket before
86 - datastore reservation (publishing)
87 - persistence support (publish, unindex, search, download)
88 - download: management/bounding of parallel downloads (for recursive dl.)
89 - non-anonymous FS service (needs DHT)
90 + basic DHT integration
91 + CS-DHT-functions (DHT-put of LOC)
92 + P2P-functions (DHT-get)
93 - active migration support (in fs or in datastore or new daemon?)
95 + active reply route caching design & implementation of service; gap extension!
97 - needs more testing (especially F2F topology)
98 - needs to re-try connecting after disconnect (currently, it
99 initially triggers a connection request, but if that connection
100 fails / goes down, it does not retry in a timely fashion;
101 cause seems to be the 'blacklist_after_attempt' being set to 1h,
102 which is rather long -- and should probably be adjusted based on
103 the number of connections / known peers)
104 - If the topology daemon crashes, peers that were put on the
105 blacklist with transport will never be removed from it (until
106 transport service dies); we should use the blacklist notification
107 API to learn about the exact set of blacklisted peers at all times
108 (FIXME: the transport_api implementation of blacklisting
109 also does not work nicely for this since it won't let us know about
110 disconnect-reconnect events and the implicit whitelisting
111 that might happen here; that's not so bad since we will
112 re-blacklist on pre-connect attempts anyway, so this is
114 - the code uses the term 'blacklist' for both peers that are forbidden
115 to connect (i.e. F2F mode) as well as peers that we currently
116 won't try to actively connect to ourselves (since we just tried);
117 This is confusing. We need two distinct terms.
118 - move code to use hash table instead of linked list
119 - instead of periodically discarding blacklisted entries,
120 simply add task that is triggered at the right time (earlier free,
122 - check if new HELLO learned is different from old HELLO
123 before resetting entire state!
125 - need function to test "equivalency" of HELLOs (or integrate with "merge"?); use in PEERINFO
128 * Module features to implement:
130 + good to have for DV/DHT evaluation!
132 + implement DHT service (needs DV, DATACACHE)
133 + implement DHT library
134 + implement testcases
135 + implement performance tests
138 * Plugins to implement:
139 - UDP backend (RC-pre2)
140 + Fragmentation library
142 - HTTP backend (RC-pre2)
144 - need to get rid of synchronous API for service starts (cause all kinds of problems)
145 [=> eliminate for need to tell ARM about service starts most of the time!] [Safey]
146 - better tracking of which config changes actually need to cause process restarts by ARM.
147 - listen for requests to discover dependencies between services (and avoid
148 having to explicitly program start requests)
149 - better crash management (attach debugging support, capture and analyze
150 debug output, detect random vs. deterministic crashes)
153 - location URIs (publish, search, download)
154 - collection API & tests
155 + gnunet-pseudonym (collection support)
156 - implement FS performance tests
166 + see discussions @ FISL about integration with event loop!
167 * Determine RC bugs and fix those!
169 - modify configuration to allow controlling connections for non-local starts
170 - testbed creation with topology (needs working F2F topology) [Nate]
171 - testbed with churn [Nate]
172 - implement testcases for distributed testing [Nate]
173 - test basic peer re-configure [Nate]
174 - test topology creation [Nate]
175 - test churn generation [Nate]
176 - consider changing API for peer-group termination to
177 call continuation when done
183 - update webpage documentation
185 - expand bibliography
186 - convert documentation pages to books
187 - update books (especially for developers)
188 - create good Drupal theme for GNUnet
189 - make a NICE download page and figure out how to
190 enable developers to publish TGZs nicely
191 - port "contact" page
192 - add content type for "todo" items?
194 * Plugins to implement:
195 - MySQL database backends
198 - Postgres database backends
202 * Determine RC bugs and fix those!
207 - SMTP transport backend
208 - HTTPS transport backend
209 + improved HTTPS support in MHD
211 - Implement method of learning our external addresses from
212 other peers; need some kind of threshold-based
213 scheme, limiting both the total number of addresses that we accept
214 this way as well as requiring multiple confirmations; also, we
215 should possibly try to confirm that the given address works for
216 us ourselves (loopback-style) before adding it to the list
217 + we may be able to simplify WELCOME messages (no need to add
218 addresses there anymore, but may help to learn them there anyway...).
219 + we probably want some kind of voting/counting for learning IP addresses
220 (maybe including IP addresses in ads proportional to how often others
221 report them? we at least need some protection against >64k HELLOs!),
222 + provide a way to give the user a list of "learned" IP addresses and
223 a way to easily "veto" addresses off the list!
224 => If MiM attacker uses vetoed address, blacklist the specific IP for
225 the presumed neighbour!
226 - implement gnunet-transport (transport configurator / tester)
227 - UPnP-based IP detection
228 (Note: build library always, build service when libxml2/etc. are available)
230 - Remove KBlocks in gnunet-unindex (see discussion with Kenneth Almquist on gnunet-devs in 9/2009)
231 - Allow checking of presence of search results and/or content via command-line tools
232 (add options to gnunet-search / gnunet-download to limit search to local peer)
234 - expire 'ancient' HELLOs (those without valid addresses AND that
235 we have not 'used' (for their public keys) in a while; need a way
236 to track actual 'use')
237 - make sue we also trigger notifications whenever HELLOs expire
243 - should use hash map to look up sessions
245 - api creates many, many short-lived TCP connections; either some
246 clients should use the API differently or we need to change the
247 API to enable re-use of connections to the service
249 - should use BIO instead of mmap
251 - need to periodically probe latency/transport cost changes & possibly switch transport
252 - instantly filter addresses from *other* peers that
253 are *equal* to our own address + port (i.e., localhost:2086). We
254 no longer filter those for outgoing (helps with loopback testing
255 and keeps the code clean), but we should filter strictly *impossible*
256 incoming addresses! This is for efficiency, not correctness.
257 - should use hash map to look up Neighbours
259 - 'server' uses 'GNUNET_PEERINFO_iterate', should probably switch to notification API
260 (for more instant / up-to-date hostlists at lower cost) [OPTIMIZATION]
264 - repeatedly resolve hostname and look up interfaces to determine our own IP
265 - [./transport/plugin_transport_tcp.c:391]: (style) struct or union member 'Plugin::address_update_task' is never used (related to issue above)
267 - [./transport/gnunet-service-transport.c:173]: (style) struct or union member 'TransportPlugin::rebuild' is never used (related to TCP not refreshing external addresses?)
269 - add stats (# bytes available, # bytes used, # PUTs, # GETs, # GETs satisfied)