1 PHASE #1: (Goal: settle key design questions)
4 * testcase fails the first time when run with fresh /tmp\
7 * improve disk API [Nils] (Nils, is this done? -Christian)
8 * Windows: use events instead of pipes to signal select()s [Nils]
9 * only connect() sockets that are ready (select()) [Nils]
10 [On W32, we need to select after calling socket before
14 For these functions, it would be nice if we had testcases ("make check")
15 that would cause them to be executed and check that they are working:
16 * gnunet-service-peerinfo:
17 - change_host_trust / flush_trust
19 - discard_hosts_helper / cron_clean_data_hosts
20 * gnunet-service-transport:
21 - try_unvalidated_addresses
22 - lookup_address_callback
23 - lookup_hello_callback
24 - plugin_env_lookup_address
25 - notify_clients_disconnect
26 - list_validated_addresses
28 - disconnect_neighbour
30 * plugin_transport_tcp.c:
32 - tcp_plugin_address_pretty_printer / append_port
33 - tcp_plugin_set_receive_quota
36 - GNUNET_TRANSPORT_set_qutoa / send_set_quota
40 - remove_from_any_list / remove_neighbour
41 - GNUNET_TRANSPORT_notify_transmit_ready_cancel
44 - solicit_traffic / copy_and_free
45 - GNUNET_CORE_peer_configure / produce_configure_message
46 * gnunet-service-core:
49 - handle_client_request_configure
52 - handle_transport_notify_disconnect
53 * hostlist (everything)
54 * topology (everything)
58 PHASE #2: (Goal: recover basic file-sharing functionality)
61 * implement sqlite-based sqstore/datastore service
62 + implement service (datastore + talks to plugin)
63 + implement library (talks to service)
64 + implement sqlite plugin (talks to DB)
65 + fix testcases (make them use CPS, complete their inner workings...)
66 * implement sqlite-based dstore services
68 + implement library (talks to service)
69 + implement service (talks to plugin)
70 + implement sqlite plugin (talks to DB)
78 PHASE #3: (Goal: ready for pre-release) [completion-goal: end of 2009]
88 * UDP backend (need LIBRARY to support (de)fragmentation!)
91 + Testcases for set_quota, timeouts, disconnects, transmit_ready_cancel
92 + Better coverage of gnunet-service-transport (hello validation)
93 + direct test of plugins compliance to plugin API
96 * sqlite, mysql & postgres backend
102 - crypto_hash: use libgcrypt (supports SHA-512 since 2003)
103 - container_bloomfilter: improve efficiency (see FIXME)
105 - inefficient memmove
107 - transport_api: support forcing disconnects through low quotas!
108 (required for working F2F support!)
109 - API: consider having core provide deadline information for each message
110 (likely important for DV plugin which wants to loop back!)
111 - implement transport API to pretty-print transport address
112 + transport_api extension (API extension!)
113 + service-transport extension (protocol extension)
114 - add calls to statistics in various places
115 - implement gnunet-transport (transport configurator / tester)
116 - UPnP-based IP detection
117 (Note: build library always, build service when libxml2/etc. are available)
118 - instantly filter addresses from *other* peers that
119 are *equal* to our own address + port (i.e., localhost:2086). We
120 no longer filter those for outgoing (helps with loopback testing
121 and keeps the code clean), but we should filter strictly *impossible*
122 incoming addresses! This is for efficiency, not correctness.
123 - We currently are happy to take any address told to us in a WELCOME
124 to our set of addresses; we should have some minimal threshold-based
125 scheme, limiting both the total number of addresses that we accept
126 this way as well as requiring multiple confirmations; also, we
127 should possibly try to confirm that the given address works for
128 us ourselves (loopback-style) before adding it to the list
130 + we may be able to simplify WELCOME messages (no need to add
131 addresses there anymore, but may help to learn them there anyway...).
132 + we probably want some kind of voting/counting for learning IP addresses
133 (maybe including IP addresses in ads proportional to how often others
134 report them? we at least need some protection against >64k HELLOs!),
135 + provide a way to give the user a list of "learned" IP addresses and
136 a way to easily "veto" addresses off the list!
137 => If MiM attacker uses vetoed address, blacklist the specific IP for
138 the presumed neighbour!
139 - not sure current way of doing ACKs works well-enough
140 with unreliable transports where the ACK maybe lost;
141 the "is_new" check would then possibly prevent future
142 ACKs to be delivered, all while we're happily
143 receiving messages from that peer! Worse, the other
144 peer won't generate another ACK since it thinks we're
145 connected just fine...
147 + How necessary is ACKing in the first place? (alternatives?)
148 + Should we transmit ACKs in response to every HELLO? (would that
149 fully address the problem?)
150 - latency measurements implemented in the transport
151 plugins makes it only work for bi-di transports
152 and results in code replication
153 - should latency be included in the ReceiveCallback and
154 NotifyConnect or passed on request?
156 - auto-generate "defaults.conf" using gnunet-setup from "config.scm"
157 - integrate all options into "config.scm"
158 - change config-file writing to exclude options set to default values
160 - implement exponential back-off for service restarts
161 - better tracking of which config changes actually need to cause process restarts by ARM.
162 - have way to specify dependencies between services (to manage ARM restarts better)
163 - client-API is inefficient since it opens a TCP connection per service that is started
164 (instead of re-using connections).
166 - code currently notifies clients about "encrypted" connections being up well before
167 we get the encrypted PONG; sometimes this may be OK (for topology killing
168 unwanted connnections), but of course not in general. I suspect we want
169 to signal on PONG and have topology hook directly into transport to
170 kill plaintext connections before they have a chance to become encrypted
171 (may require minor hack in transport API)
173 - have gnunet-peerinfo print actual host addresses again
174 - add option to gnunet-peerinfo to modify trust value
176 - finish postgres implementation; simplify other SQLs using new stats
181 - active reply route caching design & implementation of service,
184 - implement advertising of hostlist URL
185 - implement learning of hostlist URLs
191 PHASE #4: [completion-goal: mid 2010]
199 * Features eliminated from util:
200 - threading (goal: good riddance!)
201 - complex logging features [ectx-passing, target-kinds] (goal: good riddance!)
202 - complex configuration features [defaults, notifications] (goal: good riddance!)
203 - network traffic monitors (goal: eliminate)
204 - IPC semaphores (goal: d-bus? / eliminate?)
206 - DNS lookup (goal: have async service; issue: still need synchronous resolution in places, current code may not be portable)
207 => code shrunk from 61 files to 34, 22k LOC to 15k LOC, 470k to 330k (with symbols)
208 * New features in util:
210 - service and program boot-strap code
211 * Major changes in util:
212 - more expressive server (replaces selector)
214 - how to integrate scheduler with GTK event loop!