jrandom authored 20 years ago and zzz committed 20 years ago

* wrote basic license stuff

jrandom authored 20 years ago and zzz committed 20 years ago

* in the multirouter (sim), give each router a randomly distributed clock skew (within the acceptable period)
* in the multirouter (sim), disable the timestamper (so the clock skew 'sticks')
* logging

jrandom authored 20 years ago and zzz committed 20 years ago

allow explicit overriding any logical constraints on the clock skew (useful for simulating strange things)

hypercubus authored 20 years ago and zzz committed 20 years ago

* shell process spawned by the installer on Windows will not create a visible command window

jrandom authored 20 years ago and zzz committed 20 years ago

* removed some failsafe code that had been preventing any messages from being sent down alternate tunnels in a leaseSet [oops]
this may have unintended consequences, as we need to deal with messages received from skewed clocks, but I believe the two pathways
used here are safe (we leave the larger timeout thresholds intact for dealing with remotely generated message times)

hypercubus authored 20 years ago and zzz committed 20 years ago

* removed 'Shutdown I2P router' from systray menu; preferred method of shutdown is now router console

jrandom authored 20 years ago and zzz committed 20 years ago

if a netDb refetch of a lease we were able to fetch is going slowly, short circuit it by reinjecting the old (dropped) one after 10 seconds so we can attempt a resend

hypercubus authored 20 years ago and zzz committed 20 years ago

changed deprecated 'tail -1' to 'tail -n 1'; the Java Service Wrapper team will probably change it soon, but we need this for 0.4

jrandom authored 20 years ago and zzz committed 20 years ago

don't fail the tunnels used to send/receive a request on a lost reply, as the potential that they're to blame is only 1 out of 5.
(and if other people's tunnels suck, that leads us to kill all of our tunnels.  which is Bad)

jrandom authored 20 years ago and zzz committed 20 years ago

cleaned up peer selection so we don't have to repeatedly ask the profileOrganizer the same thing over and over
(instead, have the profileOrganizer check the netDb to see if the profiled peer is reachable)
cleaned up the threshold calculation a bit more

jrandom authored 20 years ago and zzz committed 20 years ago

drop the links to install/uninstall the service from this page.  not worth the effort atm (they can just run the shell scripts)

jrandom authored 20 years ago and zzz committed 20 years ago

(this fixes the longstanding "-318 bytes read" bug)

jrandom authored 20 years ago and zzz committed 20 years ago

* never drop the threshold under the baseline (the peer selection algorithm can handle there being no fast peers, etc)
* revert some of the overly zealous peer distribution code - select randomly from the fast peers, not according to a strict LRU
(which was causing lots of queued up tunnel requests, as well as tunnel failures when they all failed)
need to think some more thunks about how to address this right now.  a few different algorithms available
to deal with different scenarios and #s of users, but nothing that by itself strikes me as 'ideal', yet.   perhaps its
time for another trip to the pub to see what inspiration can be found there ;)

jrandom authored 20 years ago and zzz committed 20 years ago

load the body of the index.jsp from ./readme.html (so we can put in some intro crap w/ links and the user can change it later)

jrandom authored 20 years ago and zzz committed 20 years ago

crazy paranoia to deal with errors referencing bad dll/.so files for the service wrapper / systray / etc

jrandom authored 20 years ago and zzz committed 20 years ago

* removed all of the "temporary" threads used for adding timeouts to blocking socket operations:
 - use the ConnectionEstablisher's thread + a SimpleTimer.TimedEvent callback to timeout socket create
 - added a pool of socket handler threads (size=3 atm) for receiving any inbound sockets, which are
   pulled off a queue, after which a handshake occurs to verify the other side is I2NP (along side
   another SimpleTimer.TimedEvent callback in case that blocks)
this should get the last of the temporary threads (Jetty has its own thread pool for dealing with
HTTP requests, so we can ignore that thread created in the AdminRunner).  The only significant
reduction in threads left is to go with either NIO or UDP, but neither are happening in the immediate
future.

jrandom authored 20 years ago and zzz committed 20 years ago

oh, you mean we should actually stop waiting for something on a thread that has nothing left to do?  what a concept!
(this should kill the leaking thread issue)