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:
9 * FRAGMENTATION [Ji Lu]
10 * HTTP transport [Matthias]
15 * MySQL / Postgres plugins (datastore, datacache)
19 - "Forcing disconnect of XXX due to inactivity" -- can happen every few MS!?
20 (disconnect does not really succeed, or what?)
21 Also, core does not seem to actually decrement the number of active
22 sessions even if no connections exist for a while! (likely related...)
23 - transport reports bw quota violations on slow systems (core or
26 - utilization can (easily, restart?) go out of control (very large), causing
27 content expiration job to go crazy and delete everything!
29 - gnunet-publish uses 'NULL' for the file name when
30 publishing a directory (!)
31 - on some systems, keyword search does not find locally published content
32 (need testcase of command-line tools! - also good to cover getopt API!)
33 [could be related to datastore issue above!]
34 - 2-peer download is still too slow (why?)
35 - advanced FS API parts
36 + pick correct filenames for recursive downloads (mkdir, .gnd)
37 + support recursive download even if filename is NULL and we hence
38 do not generate files on disk (use temp_filename)
39 + bound parallelism (# fs downloads)
40 + distinguish in performance tracking and event signalling between
41 downloads that are actually running and those that are merely in the queue
42 + gnunet-service-fs (hot-path routing, load-based routing, nitpicks)
43 - [gnunet-service-fs.c:208]: member 'LocalGetContext::results_bf_size' is never used
44 - [gnunet-service-fs.c:501]: member 'PendingRequest::used_pids_size' is never used
45 - [gnunet-service-fs.c:654]: member 'ConnectedPeer::last_client_replies' is never used
46 - [gnunet-service-fs.c:669]: member 'ConnectedPeer::avg_delay' is never used
47 - [gnunet-service-fs.c:675]: member 'ConnectedPeer::avg_priority' is never used
48 - [gnunet-service-fs.c:688]: member 'ConnectedPeer::pending_requests' is never used
49 - [gnunet-service-fs.c:694]: member 'ConnectedPeer::last_p2p_replies_woff' is never used
50 - [gnunet-service-fs.c:700]: member 'ConnectedPeer::last_client_replies_woff' is never used
57 - trust: need *fast* way to check/update trust in peers
58 (async peerinfo would not be right; certainly not with the
61 - scheduler should change OS process priority based on task priority;
62 should make better use of task priorities in general
63 - only connect() sockets that are ready (select()) [Nils]
64 [On W32, we need to select after calling socket before
67 - need function to test "equivalency" of HELLOs (or integrate with "merge"?); use in PEERINFO
69 - design & implement new setup tool
71 - good to have for transport/DV evaluation!
74 - implement DV service
75 - implement DV library
76 - implement DV transport plugin
78 - implement performance tests
80 - does not seem to work with timeouts (especially if service is not running)
82 - needs more testing (especially F2F topology)
83 - needs to re-try connecting after disconnect (currently, it
84 initially triggers a connection request, but if that connection
85 fails / goes down, it does not retry in a timely fashion;
86 cause seems to be the 'blacklist_after_attempt' being set to 1h,
87 which is rather long -- and should probably be adjusted based on
88 the number of connections / known peers)
89 - If the topology daemon crashes, peers that were put on the
90 blacklist with transport will never be removed from it (until
91 transport service dies); we should use the blacklist notification
92 API to learn about the exact set of blacklisted peers at all times
93 (FIXME: the transport_api implementation of blacklisting
94 also does not work nicely for this since it won't let us know about
95 disconnect-reconnect events and the implicit whitelisting
96 that might happen here; that's not so bad since we will
97 re-blacklist on pre-connect attempts anyway, so this is
98 a minor issue; OTOH, we might want to be more explicit about
99 allowing/forbidding connects on pre-connect to avoid
100 entering connect attempts to just be blacklisted shortly afterwards).
101 - the code uses the term 'blacklist' for both peers that are forbidden
102 to connect (i.e. F2F mode) as well as peers that we currently
103 won't try to actively connect to ourselves (since we just tried);
104 This is confusing. We need two distinct terms (greylist?).
105 - move code to use hash table instead of linked list
106 - instead of periodically discarding blacklisted entries,
107 simply add task that is triggered at the right time (earlier free,
109 - check if new HELLO learned is different from old HELLO
110 before resetting entire state!
112 - datastore reservation (publishing)
113 - search: availability probes
114 - persistence support (publish, unindex, search, download)
115 - active migration support (in fs or in datastore or new daemon?)
117 + active reply route caching design & implementation of service; gap extension!
119 - how to integrate scheduler with GTK event loop!?
123 - good to have for DV/DHT evaluation!
125 - implement DHT service (needs DV, DATACACHE)
126 - implement DHT library
127 - implement testcases
128 - implement performance tests
132 - need to get rid of synchronous API for service starts (cause all kinds of problems)
133 [=> eliminate for need to tell ARM about service starts most of the time!] [Safey]
134 - better tracking of which config changes actually need to cause process restarts by ARM.
135 - listen for requests to discover dependencies between services (and avoid
136 having to explicitly program start requests)
137 - better crash management (attach debugging support, capture and analyze
138 debug output, detect random vs. deterministic crashes)
141 - location URIs (publish, search, download)
142 - non-anonymous FS service (needs DHT)
143 + DHT integration for search
144 + CS-DHT-functions (DHT-put of LOC)
145 + P2P-functions (DHT-get)
146 - collection API & tests
147 + gnunet-pseudonym (collection support)
148 - implement FS performance tests
155 * Determine RC bugs and fix those!
157 - modify configuration to allow controlling connections for non-local starts
158 - testbed creation with topology (needs working F2F topology)
160 - implement testcases for distributed testing
161 - test basic peer re-configure
162 - test topology creation
163 - test churn generation
164 - consider changing API for peer-group termination to
165 call continuation when done
167 - finalize API design
170 - integration with transport service
171 * MYSQL database backends: [CG]
177 - expand bibliography
178 - convert documentation pages to books
179 - update books (especially for developers)
180 - create good Drupal theme for GNUnet
181 - make a NICE download page and figure out how to
182 enable developers to publish TGZs nicely
183 - port "contact" page
184 - add content type for "todo" items?
185 * POSTGRES database backends: [CG]
188 * Determine RC bugs and fix those!
192 - SMTP transport backend
193 - HTTPS transport backend
194 + improved HTTPS support in MHD
196 - Implement method of learning our external addresses from
197 other peers; need some kind of threshold-based
198 scheme, limiting both the total number of addresses that we accept
199 this way as well as requiring multiple confirmations; also, we
200 should possibly try to confirm that the given address works for
201 us ourselves (loopback-style) before adding it to the list
202 + we may be able to simplify WELCOME messages (no need to add
203 addresses there anymore, but may help to learn them there anyway...).
204 + we probably want some kind of voting/counting for learning IP addresses
205 (maybe including IP addresses in ads proportional to how often others
206 report them? we at least need some protection against >64k HELLOs!),
207 + provide a way to give the user a list of "learned" IP addresses and
208 a way to easily "veto" addresses off the list!
209 => If MiM attacker uses vetoed address, blacklist the specific IP for
210 the presumed neighbour!
211 - implement gnunet-transport (transport configurator / tester)
212 - UPnP-based IP detection
213 (Note: build library always, build service when libxml2/etc. are available)
215 - Remove KBlocks in gnunet-unindex (see discussion with Kenneth Almquist on gnunet-devs in 9/2009)
216 - Allow checking of presence of search results and/or content via command-line tools
217 (add options to gnunet-search / gnunet-download to limit search to local peer)
219 - expire 'ancient' HELLOs (those without valid addresses AND that
220 we have not 'used' (for their public keys) in a while; need a way
221 to track actual 'use')
222 - make sue we also trigger notifications whenever HELLOs expire
229 - should use hash map to look up sessions
231 - api creates many, many short-lived TCP connections; either some
232 clients should use the API differently or we need to change the
233 API to enable re-use of connections to the service
235 - should use BIO instead of mmap
237 - need to periodically probe latency/transport cost changes & possibly switch transport
238 - instantly filter addresses from *other* peers that
239 are *equal* to our own address + port (i.e., localhost:2086). We
240 no longer filter those for outgoing (helps with loopback testing
241 and keeps the code clean), but we should filter strictly *impossible*
242 incoming addresses! This is for efficiency, not correctness.
243 - should use hash map to look up Neighbours
245 - 'server' uses 'GNUNET_PEERINFO_iterate', should probably switch to notification API
246 (for more instant / up-to-date hostlists at lower cost) [OPTIMIZATION]
250 - repeatedly resolve hostname and look up interfaces to determine our own IP
251 - [./transport/plugin_transport_tcp.c:391]: (style) struct or union member 'Plugin::address_update_task' is never used (related to issue above)
253 - [./transport/gnunet-service-transport.c:173]: (style) struct or union member 'TransportPlugin::rebuild' is never used (related to TCP not refreshing external addresses?)
255 - add stats (# bytes available, # bytes used, # PUTs, # GETs, # GETs satisfied)
257 - support inline data in directories for recursive file downloads (fs_download)