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