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 - core dump / jump to nowhere (valgrind) in udp test in transports
19 - TCP not used bi-directionally (especially important also for PONG!)
20 Need a way for plugin to tell to service that it can *reliably* transmit
21 PONGs. Need *good* way to ensure only one (plugin) session is active at
22 any given point in time; need systematic way to periodically probe latency/
23 transport cost changes
24 - disconnect notifications from TCP not always as nice as we might want
25 them to be (but how can we do this cleanly given that UDP is connectionless?)
26 - implement transport API to pretty-print transport address
27 + transport_api extension (API extension!)
28 + service-transport extension (protocol extension)
29 - instantly filter addresses from *other* peers that
30 are *equal* to our own address + port (i.e., localhost:2086). We
31 no longer filter those for outgoing (helps with loopback testing
32 and keeps the code clean), but we should filter strictly *impossible*
33 incoming addresses! This is for efficiency, not correctness.
34 - [./transport/gnunet-service-transport.c:173]: (style) struct or union member 'TransportPlugin::rebuild' is never used
35 - [./transport/plugin_transport_tcp.c:391]: (style) struct or union member 'Plugin::address_update_task' is never used
37 - SET_KEY is triggered every 2.5 minutes (after session is up); should
38 just do PING (not SET_KEY + PING)
39 - "Forcing disconnect of XXX due to inactivity" -- can happen every few MS!?
40 (disconnect does not really succeed, or what?)
42 - need function to test "equivalency" of HELLOs; use in topology!
45 + active reply route caching design & implementation of service,
47 - gnunet-publish cannot be aborted using CTRL-C
48 - on some systems, keyword search does not find locally published content
49 (need testcase of command-line tools!)
50 - 2-peer download is still too slow (why?)
51 - advanced FS API parts
52 + namespaces: fundamental namespace API
53 + search: SBlocks, KSBlocks, probes, notify FS-service of known results
55 + location URIs (publish, search, download)
56 + indexing: index-failure-cleanup
57 + download: management/bounding of parallel downloads (for recursive dl.)
58 + datastore reservation (publishing)
59 + gnunet-publish (URI args)
60 + gnunet-download (directory)
61 + gnunet-search (options, incl. namespace search)
62 + gnunet-download (many options)
63 + gnunet-directory (man page, options)
64 + gnunet-pseudonym (all of it)
65 + gnunet-service-fs (remove failing on-demand blocks, hot-path routing, load-based routing, nitpicks)
66 - implement adv. FS testcases
68 + insert: sblocks, loc uris
70 + search: sblocks, skblocks, loc uris
73 - implement FS performance tests
78 - [./fs/gnunet-service-fs.c:208]: (style) struct or union member 'LocalGetContext::results_bf_size' is never used
79 - [./fs/gnunet-service-fs.c:501]: (style) struct or union member 'PendingRequest::used_pids_size' is never used
80 - [./fs/gnunet-service-fs.c:654]: (style) struct or union member 'ConnectedPeer::last_client_replies' is never used
81 - [./fs/gnunet-service-fs.c:669]: (style) struct or union member 'ConnectedPeer::avg_delay' is never used
82 - [./fs/gnunet-service-fs.c:675]: (style) struct or union member 'ConnectedPeer::avg_priority' is never used
83 - [./fs/gnunet-service-fs.c:688]: (style) struct or union member 'ConnectedPeer::pending_requests' is never used
84 - [./fs/gnunet-service-fs.c:694]: (style) struct or union member 'ConnectedPeer::last_p2p_replies_woff' is never used
85 - [./fs/gnunet-service-fs.c:700]: (style) struct or union member 'ConnectedPeer::last_client_replies_woff' is never used
92 * Module features to implement:
94 + good to have for DV evaluation!
97 + implement DV service [Nate & CG]
98 + implement DV library [Nate]
99 + implement DV transport plugin [Nate & CG]
100 + implement testcases [Nate]
101 + implement performance tests [Nate]
103 - design & implement new setup tool
106 + how to integrate scheduler with GTK event loop!
108 - 'server' uses 'GNUNET_PEERINFO_iterate', should probably switch to notification API
109 (for more instant / up-to-date hostlists at lower cost) [OPTIMIZATION]
111 - does not seem to work with timeouts (especially if service is not running)
113 - trust: need *fast* way to check/update trust in peers
114 (async peerinfo would not be right)
115 - scheduler should change OS process priority based on task priority;
116 should make better use of task priorities in general
117 - only connect() sockets that are ready (select()) [Nils]
118 [On W32, we need to select after calling socket before
121 - add stats (# bytes available, # bytes used, # PUTs, # GETs, # GETs satisfied)
123 - persistence support (publish, unindex, search, download)
124 - non-anonymous FS service (needs DHT)
125 + basic DHT integration
126 + CS-DHT-functions (DHT-put of LOC)
127 + P2P-functions (DHT-get)
128 - active migration support (in fs or in datastore?)
130 - needs more testing (especially F2F topology)
131 - needs to re-try connecting after disconnect (currently, it
132 initially triggers a connection request, but if that connection
133 fails / goes down, it does not retry in a timely fashion;
134 cause seems to be the 'blacklist_after_attempt' being set to 1h,
135 which is rather long -- and should probably be adjusted based on
136 the number of connections / known peers)
137 - If the topology daemon crashes, peers that were put on the
138 blacklist with transport will never be removed from it (until
139 transport service dies); we should use the blacklist notification
140 API to learn about the exact set of blacklisted peers at all times
141 (FIXME: the transport_api implementation of blacklisting
142 also does not work nicely for this since it won't let us know about
143 disconnect-reconnect events and the implicit whitelisting
144 that might happen here; that's not so bad since we will
145 re-blacklist on pre-connect attempts anyway, so this is
147 - the code uses the term 'blacklist' for both peers that are forbidden
148 to connect (i.e. F2F mode) as well as peers that we currently
149 won't try to actively connect to ourselves (since we just tried);
150 This is confusing. We need two distinct terms.
151 - move code to use hash table instead of linked list
152 - instead of periodically discarding blacklisted entries,
153 simply add task that is triggered at the right time (earlier free,
155 - check if new HELLO learned is different from old HELLO
156 before resetting entire state!
159 * Module features to implement:
161 + good to have for DV/DHT evaluation!
163 + implement DHT service (needs DV, DATACACHE)
164 + implement DHT library
165 + implement testcases
166 + implement performance tests
169 * Plugins to implement:
170 - UDP backend (RC-pre2)
171 + Fragmentation library
173 - HTTP backend (RC-pre2)
175 - need to get rid of synchronous API for service starts (cause all kinds of problems)
176 [=> eliminate for need to tell ARM about service starts most of the time!] [Safey]
177 - better tracking of which config changes actually need to cause process restarts by ARM.
178 - listen for requests to discover dependencies between services (and avoid
179 having to explicitly program start requests)
180 - better crash management (attach debugging support, capture and analyze
181 debug output, detect random vs. deterministic crashes)
188 + see discussions @ FISL about integration with event loop!
189 * Determine RC bugs and fix those!
191 - modify configuration to allow controlling connections for non-local starts
192 - testbed creation with topology (needs working F2F topology) [Nate]
193 - testbed with churn [Nate]
194 - implement testcases for distributed testing [Nate]
195 - test basic peer re-configure [Nate]
196 - test topology creation [Nate]
197 - test churn generation [Nate]
198 - consider changing API for peer-group termination to
199 call continuation when done
205 - update webpage documentation
207 - expand bibliography
208 - convert documentation pages to books
209 - update books (especially for developers)
210 - create good Drupal theme for GNUnet
211 - make a NICE download page and figure out how to
212 enable developers to publish TGZs nicely
213 - port "contact" page
214 - add content type for "todo" items?
216 * Plugins to implement:
217 - MySQL database backends
220 - Postgres database backends
224 * Determine RC bugs and fix those!
229 - SMTP transport backend
230 - HTTPS transport backend
231 + improved HTTPS support in MHD
233 - Implement method of learning our external addresses from
234 other peers; need some kind of threshold-based
235 scheme, limiting both the total number of addresses that we accept
236 this way as well as requiring multiple confirmations; also, we
237 should possibly try to confirm that the given address works for
238 us ourselves (loopback-style) before adding it to the list
239 + we may be able to simplify WELCOME messages (no need to add
240 addresses there anymore, but may help to learn them there anyway...).
241 + we probably want some kind of voting/counting for learning IP addresses
242 (maybe including IP addresses in ads proportional to how often others
243 report them? we at least need some protection against >64k HELLOs!),
244 + provide a way to give the user a list of "learned" IP addresses and
245 a way to easily "veto" addresses off the list!
246 => If MiM attacker uses vetoed address, blacklist the specific IP for
247 the presumed neighbour!
248 - implement gnunet-transport (transport configurator / tester)
249 - UPnP-based IP detection
250 (Note: build library always, build service when libxml2/etc. are available)
252 - Remove KBlocks in gnunet-unindex (see discussion with Kenneth Almquist on gnunet-devs in 9/2009)
253 - Allow checking of presence of search results and/or content via command-line tools
254 (add options to gnunet-search / gnunet-download to limit search to local peer)
256 - should use BIO instead of mmap
258 - expire 'ancient' HELLOs (those without valid addresses AND that
259 we have not 'used' (for their public keys) in a while; need a way
260 to track actual 'use')
261 - make sue we also trigger notifications whenever HELLOs expire