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 - SET_KEY is triggered every 2.5 minutes (after session is up); should
19 just do PING (not SET_KEY + PING)
20 - "Forcing disconnect of XXX due to inactivity" -- can happen every few MS!?
21 (disconnect does not really succeed, or what?)
22 Also, core does not seem to actually decrement the number of active
23 sessions even if no connections exist for a while! (likely related...)
24 - transport reports bw quota violations on big-endian systems (core or
27 - utilization can (easily, restart?) go out of control (very large), causing
28 content expiration job to go crazy and delete everything!
30 ! fs_search test fails (hangs) since blocktype was changed from "ANY" to "KBLOCK"!
31 (insert blocktype wrong? or what!?)
32 ! fs_namespace test fails (hangs), but local fs service has found results; somehow
33 they are not passed to the client and/or the shutdown sequence is not working
34 - gnunet-publish cannot be aborted using CTRL-C
35 - on some systems, keyword search does not find locally published content
36 (need testcase of command-line tools! - also good to cover getopt API!)
37 - 2-peer download is still too slow (why?)
38 - advanced FS API parts
39 + search: SBlocks, NBlocks, probes, notify FS-service of known results
40 + indexing: index-failure-cleanup
41 + gnunet-publish (URI args)
42 + gnunet-download (directory)
43 + gnunet-search (options, incl. namespace search)
44 + gnunet-download (many options)
45 + gnunet-directory (man page, options)
46 + gnunet-pseudonym (all of it)
47 + gnunet-service-fs (remove failing on-demand blocks, hot-path routing,
48 load-based routing, nitpicks)
49 - implement adv. FS testcases
50 - [./fs/gnunet-service-fs.c:208]: (style) struct or union member 'LocalGetContext::results_bf_size' is never used
51 - [./fs/gnunet-service-fs.c:501]: (style) struct or union member 'PendingRequest::used_pids_size' is never used
52 - [./fs/gnunet-service-fs.c:654]: (style) struct or union member 'ConnectedPeer::last_client_replies' is never used
53 - [./fs/gnunet-service-fs.c:669]: (style) struct or union member 'ConnectedPeer::avg_delay' is never used
54 - [./fs/gnunet-service-fs.c:675]: (style) struct or union member 'ConnectedPeer::avg_priority' is never used
55 - [./fs/gnunet-service-fs.c:688]: (style) struct or union member 'ConnectedPeer::pending_requests' is never used
56 - [./fs/gnunet-service-fs.c:694]: (style) struct or union member 'ConnectedPeer::last_p2p_replies_woff' is never used
57 - [./fs/gnunet-service-fs.c:700]: (style) struct or union member 'ConnectedPeer::last_client_replies_woff' is never used
63 * Module features to implement:
65 + good to have for DV evaluation! [MW]
68 + implement DV service [Nate & CG]
69 + implement DV library [Nate]
70 + implement DV transport plugin [Nate & CG]
71 + implement testcases [Nate]
72 + implement performance tests [Nate]
74 - design & implement new setup tool
77 + how to integrate scheduler with GTK event loop!
79 - does not seem to work with timeouts (especially if service is not running)
81 - trust: need *fast* way to check/update trust in peers
82 (async peerinfo would not be right; certainly not with the
85 - scheduler should change OS process priority based on task priority;
86 should make better use of task priorities in general
87 - only connect() sockets that are ready (select()) [Nils]
88 [On W32, we need to select after calling socket before
91 - datastore reservation (publishing)
92 - persistence support (publish, unindex, search, download)
93 - download: management/bounding of parallel downloads (for recursive dl.)
94 - non-anonymous FS service (needs DHT)
95 + basic DHT integration
96 + CS-DHT-functions (DHT-put of LOC)
97 + P2P-functions (DHT-get)
98 - active migration support (in fs or in datastore or new daemon?)
100 + active reply route caching design & implementation of service; gap extension!
102 - needs more testing (especially F2F topology)
103 - needs to re-try connecting after disconnect (currently, it
104 initially triggers a connection request, but if that connection
105 fails / goes down, it does not retry in a timely fashion;
106 cause seems to be the 'blacklist_after_attempt' being set to 1h,
107 which is rather long -- and should probably be adjusted based on
108 the number of connections / known peers)
109 - If the topology daemon crashes, peers that were put on the
110 blacklist with transport will never be removed from it (until
111 transport service dies); we should use the blacklist notification
112 API to learn about the exact set of blacklisted peers at all times
113 (FIXME: the transport_api implementation of blacklisting
114 also does not work nicely for this since it won't let us know about
115 disconnect-reconnect events and the implicit whitelisting
116 that might happen here; that's not so bad since we will
117 re-blacklist on pre-connect attempts anyway, so this is
119 - the code uses the term 'blacklist' for both peers that are forbidden
120 to connect (i.e. F2F mode) as well as peers that we currently
121 won't try to actively connect to ourselves (since we just tried);
122 This is confusing. We need two distinct terms.
123 - move code to use hash table instead of linked list
124 - instead of periodically discarding blacklisted entries,
125 simply add task that is triggered at the right time (earlier free,
127 - check if new HELLO learned is different from old HELLO
128 before resetting entire state!
130 - need function to test "equivalency" of HELLOs (or integrate with "merge"?); use in PEERINFO
133 * Module features to implement:
135 + good to have for DV/DHT evaluation!
137 + implement DHT service (needs DV, DATACACHE)
138 + implement DHT library
139 + implement testcases
140 + implement performance tests
143 * Plugins to implement:
144 - UDP backend (RC-pre2)
145 + Fragmentation library
147 - HTTP backend (RC-pre2)
149 - need to get rid of synchronous API for service starts (cause all kinds of problems)
150 [=> eliminate for need to tell ARM about service starts most of the time!] [Safey]
151 - better tracking of which config changes actually need to cause process restarts by ARM.
152 - listen for requests to discover dependencies between services (and avoid
153 having to explicitly program start requests)
154 - better crash management (attach debugging support, capture and analyze
155 debug output, detect random vs. deterministic crashes)
158 - location URIs (publish, search, download)
159 - collection API & tests
160 - implement FS performance tests
170 + see discussions @ FISL about integration with event loop!
171 * Determine RC bugs and fix those!
173 - modify configuration to allow controlling connections for non-local starts
174 - testbed creation with topology (needs working F2F topology) [Nate]
175 - testbed with churn [Nate]
176 - implement testcases for distributed testing [Nate]
177 - test basic peer re-configure [Nate]
178 - test topology creation [Nate]
179 - test churn generation [Nate]
180 - consider changing API for peer-group termination to
181 call continuation when done
187 - update webpage documentation
189 - expand bibliography
190 - convert documentation pages to books
191 - update books (especially for developers)
192 - create good Drupal theme for GNUnet
193 - make a NICE download page and figure out how to
194 enable developers to publish TGZs nicely
195 - port "contact" page
196 - add content type for "todo" items?
198 * Plugins to implement:
199 - MySQL database backends
202 - Postgres database backends
206 * Determine RC bugs and fix those!
211 - SMTP transport backend
212 - HTTPS transport backend
213 + improved HTTPS support in MHD
215 - Implement method of learning our external addresses from
216 other peers; need some kind of threshold-based
217 scheme, limiting both the total number of addresses that we accept
218 this way as well as requiring multiple confirmations; also, we
219 should possibly try to confirm that the given address works for
220 us ourselves (loopback-style) before adding it to the list
221 + we may be able to simplify WELCOME messages (no need to add
222 addresses there anymore, but may help to learn them there anyway...).
223 + we probably want some kind of voting/counting for learning IP addresses
224 (maybe including IP addresses in ads proportional to how often others
225 report them? we at least need some protection against >64k HELLOs!),
226 + provide a way to give the user a list of "learned" IP addresses and
227 a way to easily "veto" addresses off the list!
228 => If MiM attacker uses vetoed address, blacklist the specific IP for
229 the presumed neighbour!
230 - implement gnunet-transport (transport configurator / tester)
231 - UPnP-based IP detection
232 (Note: build library always, build service when libxml2/etc. are available)
234 - Remove KBlocks in gnunet-unindex (see discussion with Kenneth Almquist on gnunet-devs in 9/2009)
235 - Allow checking of presence of search results and/or content via command-line tools
236 (add options to gnunet-search / gnunet-download to limit search to local peer)
238 - expire 'ancient' HELLOs (those without valid addresses AND that
239 we have not 'used' (for their public keys) in a while; need a way
240 to track actual 'use')
241 - make sue we also trigger notifications whenever HELLOs expire
247 - should use hash map to look up sessions
249 - api creates many, many short-lived TCP connections; either some
250 clients should use the API differently or we need to change the
251 API to enable re-use of connections to the service
253 - should use BIO instead of mmap
255 - need to periodically probe latency/transport cost changes & possibly switch transport
256 - instantly filter addresses from *other* peers that
257 are *equal* to our own address + port (i.e., localhost:2086). We
258 no longer filter those for outgoing (helps with loopback testing
259 and keeps the code clean), but we should filter strictly *impossible*
260 incoming addresses! This is for efficiency, not correctness.
261 - should use hash map to look up Neighbours
263 - 'server' uses 'GNUNET_PEERINFO_iterate', should probably switch to notification API
264 (for more instant / up-to-date hostlists at lower cost) [OPTIMIZATION]
268 - repeatedly resolve hostname and look up interfaces to determine our own IP
269 - [./transport/plugin_transport_tcp.c:391]: (style) struct or union member 'Plugin::address_update_task' is never used (related to issue above)
271 - [./transport/gnunet-service-transport.c:173]: (style) struct or union member 'TransportPlugin::rebuild' is never used (related to TCP not refreshing external addresses?)
273 - add stats (# bytes available, # bytes used, # PUTs, # GETs, # GETs satisfied)