-* topology
- - valgrind crash (in topology test case):
-==31107== Invalid read of size 8
-==31107== at 0x401DA1: free_peer (gnunet-daemon-topology.c:480)
-==31107== by 0x4030A2: cleaning_task (gnunet-daemon-topology.c:1212)
-==31107== by 0x545DCBD: run_ready (scheduler.c:482)
-==31107== by 0x545E0CD: GNUNET_SCHEDULER_run (scheduler.c:609)
-==31107== by 0x545B00F: GNUNET_PROGRAM_run (program.c:223)
-==31107== by 0x403477: main (gnunet-daemon-topology.c:1346)
-==31107== Address 0x737e828 is 8 bytes inside a block of size 144 free'd
-==31107== at 0x4C21DBC: free (vg_replace_malloc.c:325)
-==31107== by 0x544240B: GNUNET_xfree_ (common_allocation.c:144)
-==31107== by 0x401E13: free_peer (gnunet-daemon-topology.c:487)
-==31107== by 0x4030A2: cleaning_task (gnunet-daemon-topology.c:1212)
-==31107== by 0x545DCBD: run_ready (scheduler.c:482)
-==31107== by 0x545E0CD: GNUNET_SCHEDULER_run (scheduler.c:609)
-==31107== by 0x545B00F: GNUNET_PROGRAM_run (program.c:223)
-==31107== by 0x403477: main (gnunet-daemon-topology.c:1346)
-==31107==
- - considers peers 'connected' well before they actually are
- (since core notifies about it too early?);
- - (forced) disconnect does not work (also CORE API issue)
- - needs testing (not sure the current testcase does much...)
-* hostlist
- - test fails (looks like it works, but that's because of a bad
- connectivity notification; somehow core is unable to send
- messages successfully via transport)
-* CORE:
- - core notifies about connects "too early" (when we have not yet
- succeeded with the full key exchange) [see also: BUGS]