aboutsummaryrefslogtreecommitdiff
path: root/README.1st
blob: c25992351f54602ba4d9219c8ae171e712c58ad7 (plain) (blame)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
WARNING!
=======

The following is a list of issues with GNUnet 0.11.0 that will need
to be addressed before we might consider GNUnet deployable to larger
audiences. Please keep this in mind when trying out GNUnet 0.11.0!

GNUnet may however work fine for applications that aren't impeded by
these known deficiencies. Help would of course be welcome to reduce
this list, so an estimate of how much work is needed and the main 
impact are given with each item.


ats:
* We currently select one transport per peer, but we should allow the
use of multiple channels concurrently (i.e. UDP + TCP at the same
time). Neither ATS nor transport support this today, and this requires
a major change in the ATS plugins and the ATS API and the overall
ATS/transport logic. [6-12 PM, robustness, performance]

* The current proportional ATS heuristic does not devalue increasing
the number of connections once we have reached saturation point,
possibly resulting in an excessive number of connections.
[1-3 PM, performance]

* The mlp/ril ATS heuristics are highly unstable and keep crashing or
worse, and have horrific code quality. [3-12 PM, stability,
performance]


transport: [12-48 PM overall]
* We should move plugins into separate processes to break an overly
complicated subsystem into more manageable bits. [maintainability]

* Transport plugins are currently expected to be bi-directional; the
API should be changed to make them uni-directional, so we could have
say UDP for sending but receive via say SMTP, simply because NAT
punching and other transports (especially SAT) simply are not
bi-directional. Fragmentation and ACKs should then not be done in UDP
plugin but at transport level.  This should result in significantly:
- better NAT traversal
- faster transports (especially with the multi-transport of ATS)
- simplified transport plugins
[stability, robustness, performance]
 
* Transport currently does not encrypt. This has the disadvantage that
TCP/UDP traffic is easily identified as "GNUnet"-traffic.  It would be
better to_also use a simple cipher (ECDHE+AEAD) in a plugin-specific
way (i.e. HTTPS is fine already) to minimize information leakage, even
if for efficiency that cipher is is not replay-protected at this level
(and leave true replay-protected OTR to CORE). [privacy, censorship-resistance]

* transport's 'manipulation' functions should be moved into
a plugin-proxy, simplifying the code. This may have
modest implications for testbed due to the API change.
[maintainability, security]

* testcases are plenty but insufficiently systematic, the changes
described above should also enable us to create more systematic tests.
[maintainability, correctness]


hello:
* The current code may leak LAN IPs (in particular IPv6 with
MAC) globally. We have started to put in some provisions to tag
addresses as loopback/LAN/WAN, but need to systematically ensure
that addresses are only propagated in a useful scope and avoid
leaking "sensitive" address data globally.
[3-6 PM, privacy]


core:
* core needs to be able to communicate to other peers whether this
peer is high-bandwidth or on battery and thus either great for
relaying or really bad for relaying. Higher-level subsystems
could then bias their peer selection to more capable peers --
this is key for going on battery-operated systems.
[1-3 PM, performance]

* we currently use timestamps (based on roughly-synchronized clocks),
challenge-response AND sequence numbers to protect against
replay-protection. This is overkill, and the use of timestamps
causes issues on OpenWRT where RTCs are often unavailable. We
might want to increase the nonce for challenges and get rid of
the roughly-synchronized clocks assumption, at least for CORE
(we can't avoid it for NSE). [1-3 PM, usability]

* once transport encrypts (ECDH+AES), we should simplify CORE level
encryption to use ECDH+Twofish(+SHA512) instead of the current
ECDH+AES+Twofish double-encryption.

* Mobile peers currently allow adversaries to track their users as the
peer's identity does not change with locations. We need a mechanism to
notify a mobile peer about a location change and then systematically
change the public key we use depending on our location. [6-12 PM,
privacy]

* Core should support protocol versioning for higher-level services
[3 PM]


cadet:
* needs more systematic testing, ideally with a mock up of DHT
and core. [6-12 PM, correctness]

* flow and congestion control implementation remains incomplete;
should borrow more from net2o [3-6 PM, performance]

* Various optimizations, such as key material pre-computation
and avoiding unnecessary DHT queries should be implemented.
[3-9 PM, performance]

* Once transport/core use AES/Twofish for encryption, we should
switch CADET to use KECCAK-AEAD.

* Generally needs more benchmarking to identify performance
bottlenecks tune accordingly. [3-12 PM, performance]


nat:
* many known NAT traversal methods are not yet supported
[6-12 PM, connectivity, usability]


dv:
* dv is still not working, but currently based on SET and we
should simplify the design to not use SET. [3-6 PM, connectivity,
correctness]


util:
* Event loop instantiations for various event loops (like glib,
libev, libevent, etc.) should be created (and tested).  This
allows applications to be built with GNUnet using those various
styles of event loop handling.  General support for abstracting
the event loop is new in 0.11.0, we just need to make broader use
of it. [1-3 PM, performance, usability]

* gnunet-service-resolver should use asynchronous DNS queries if
available.  This will ensure that name resolution is not super-slow
when GNUnet makes various parallel DNS queries (rare, but DNS
queries can happen to resolve peer IP addresses for visualization).
[1 PM, performance, usability]


dht:
* various minor tweaks should be explored, but first we need to
define and find a good benchmark to see where our performance
bottlenecks really are today; datacache (IO subsystem) might
be one of them.  Here, picking appropriate transaction scopes
and isolation levels should be an easy way to boost performance,
but we also need more benchmarking logic. [3-12 PM, performance]


set:
* Current implementation is not yet Byzantine fault-tolerant
against stuffing attack.  [1-3 PM, fault-tolerance]


topology:
* quite a few different functions are squished together in one
subsystem here, will require some changes to address changes
to transport and hello discussed above [1-3 PM, correctness]

* there are more robust known ways for maintaining a random mesh, we
might want to explore using those [3-6 PM, robustness]


fs:
* The service should be split into two, one for user-specific and one
for system-specific components [3-6 PM, security, usability]

* Currently, data and files are stored in a disassociated way,
resulting in search results that then fail to download; need OR/RPS
(see below) before we can think about fixing this properly. [6-18 PM,
usability]

* datastore (IO subsystem) is bad at keeping the database within the
  quota set by the user. [1-3 PM, correctness]


conversation:
* dropping 'silence' on large-buffers during replay should be
used to avoid accumulating latency (standard VoIP technique,
crucial to get it to work for real calls) [1-3 PM, usability]

* needs ring tones [1 PM, usability]


gns:
* should support hijacking arbitrary gtlds [1 PM, usability]

* should import ".fr" ccTLD and ensure implementation performs [3-6 PM]


multicast:
* implementation currently "trivial" (aka doing round-robin, not multicast)
  and under-tested

* still uses old NTR-style API, needs to transition to MQ API [DONE]


psyc:
* implementation under-tested, database performance likely in
need for optimization [3-12 PM]


social:
* code in need for clean up and more systematic testing [6-12 PM]


rps:
* Open research, still experimental. [3-24 PM]


xolotl/lake/or:
* Open research, still non-existent. Symmetric cipher should
be CHACHA. [48-120 PM]


No serious problems are known in:
* statistics
* nse
* curl
* revocation
* scalarproduct
* tun
* testbed
* block
* fragmentation
* consensus
* secretsharing
* pt
* vpn
* gns
* gnsrecord
* arm
* dns
* exit
* identity
* json
* sq
* my
* namecache
* namestore
* testing
* hostlist
* zonemaster


We reserve judgement (due to lack of data) on:
* reclaim
* credential
* abe
* rest
* messenger