summaryrefslogtreecommitdiff
path: root/src/transport
diff options
context:
space:
mode:
authorChristian Grothoff <christian@grothoff.org>2019-05-20 00:00:40 +0200
committerChristian Grothoff <christian@grothoff.org>2019-05-20 00:00:40 +0200
commit426317707f34282830e5595193fcf2546c7d5b49 (patch)
tree3a18da0e6268dac4dabe9517d3a66e1644fc0423 /src/transport
parent6e2ca97fb0d3d85c1ef127bc0b9bd3e1e72aa9e8 (diff)
use _exit in signal handlers
Diffstat (limited to 'src/transport')
-rw-r--r--src/transport/gnunet-service-tng.c10
1 files changed, 10 insertions, 0 deletions
diff --git a/src/transport/gnunet-service-tng.c b/src/transport/gnunet-service-tng.c
index f009a491b..20285f94a 100644
--- a/src/transport/gnunet-service-tng.c
+++ b/src/transport/gnunet-service-tng.c
@@ -44,6 +44,16 @@
* - increment incoming_fc_window_size_used_kb when CORE is done
* with incoming packets!
*
+ * for DV)
+ * - send challenges via DV (when DVH is confirmed *and* we care about
+ * the target to get window size, or when DVH is unconfirmed (passive
+ * learning!) to confirm it!)
+ * - handle challenge responses in this case (note: validity period of addresses
+ * will be zero!)
+ * - if available, try to use DV paths when trying to establish
+ * virtual link for a `struct IncomingRequest`. (i.e. if DVH is
+ * unconfirmed, incoming requests also trigger challenge-via-DV!)
+ *
* - review retransmission logic, right now there is no smartness there!
* => congestion control, etc [PERFORMANCE-BASICS]
*