* Improved the bandwidth throtting on tunnel participation, especially for
low bandwidth peers.
* Improved failure handling in SSU with proactive reestablishment of
failing idle peers, and rather than shitlisting a peer who failed too
much, drop the SSU session and allow a new attempt (which, if it fails,
will cause a shitlisting)
* Clarify the cause of the shitlist on the profiles page, and include
bandwidth limiter info at the bottom of the peers page.
* Improved the bandwidth throtting on tunnel participation, especially for
low bandwidth peers.
* Improved failure handling in SSU with proactive reestablishment of
failing idle peers, and rather than shitlisting a peer who failed too
much, drop the SSU session and allow a new attempt (which, if it fails,
will cause a shitlisting)
* Clarify the cause of the shitlist on the profiles page, and include
bandwidth limiter info at the bottom of the peers page.
* In Syndie, propogate the subject and tags in a reply, and show the parent
post on the edit page for easy quoting. (thanks identiguy and CofE!)
* Streamline some netDb query handling to run outside the jobqueue -
which means they'll run on the particular SSU thread that handles the
message. This should help out heavily loaded netDb peers.
* Defer netDb searches for newly referenced peers until we actually want
them
* Ignore netDb references to peers on our shitlist
* Set the timeout for end to end client messages to the max delay after
finding the leaseSet, so we don't have as many expired messages floating
around.
* Add a floor to the streaming lib window size
* When we need to send a streaming lib ACK, try to retransmit one of the
unacked packets instead (with updated ACK/NACK fields, of course). The
bandwidth cost of an unnecessary retransmission should be minor as
compared to both an ACK packet (rounded up to 1KB in the tunnels) and
the probability of a necessary retransmission.
* Adjust the streaming lib cwin algorithm to allow growth after a full
cwin messages if the rtt is trending downwards. If it is not, use the
existing algorithm.
* Increased the maximum rto size in the streaming lib.
* Load balancing bugfix on end to end messages to distribute across
tunnels more evenly.
* Defer netDb searches for newly referenced peers until we actually want
them
* Ignore netDb references to peers on our shitlist
* Set the timeout for end to end client messages to the max delay after
finding the leaseSet, so we don't have as many expired messages floating
around.
* Add a floor to the streaming lib window size
* When we need to send a streaming lib ACK, try to retransmit one of the
unacked packets instead (with updated ACK/NACK fields, of course). The
bandwidth cost of an unnecessary retransmission should be minor as
compared to both an ACK packet (rounded up to 1KB in the tunnels) and
the probability of a necessary retransmission.
* Adjust the streaming lib cwin algorithm to allow growth after a full
cwin messages if the rtt is trending downwards. If it is not, use the
existing algorithm.
* Increased the maximum rto size in the streaming lib.
* Load balancing bugfix on end to end messages to distribute across
tunnels more evenly.
* Add a blocking fetch to EepGetScheduler and RemoteArchiveBean and use them from the updater, to prevent race conditions with multiple archive fetches.
* Integrated GNU-Crypto's Fortuna PRNG, seeding it off /dev/urandom and
./prngseed.rnd (if they exist), and reseeding it with data out of
various crypto operations (unused bits in a DH exchange, intermediary
bits in a DSA signature generation, extra bits in an ElGamal decrypt).
The Fortuna implementation under gnu.crypto.prng has been modified to
use BouncyCastle's SHA256 and Cryptix's AES (since those are the ones
I2P uses), and the resulting gnu.crypto.prng.* are therefor available
under GPL+Classpath's linking exception (~= LGPL). I2P's SecureRandom
wrapper around it is, of course, public domain.
* Fix bug in ircclient that prevented it to use its own dest (i.e. was
always shared. (thx for info Ragnarok)
* Fix crash in Sucker with some bad html.
* Workaround a bug in GCJ's Calendar implementation
* Propery throw an exception in the streaming lib if we try to write to a
closed stream. This will hopefully help clear some I2Phex bugs (thanks
GregorK!)
* Ported the snark bittorrent client to I2P such that it is compatible
with i2p-bt and azneti2p. For usage information, grab an update and run
"java -jar lib/i2psnark.jar". It isn't currently multitorrent capable,
but adding in support would be fairly easy (see PeerAcceptor.java:49)
* Don't allow leaseSets expiring too far in the future (thanks postman)
libjbigi.so. Yeah, this means that i2psnark uses pure java modPow, but it doesn't do
any real heavy lifting anyway, except a DSA signature every 5-10 minutes. whoop de do.
* Ported the snark bittorrent client to I2P such that it is compatible
with i2p-bt and azneti2p. For usage information, grab an update and run
"java -jar lib/i2psnark.jar". It isn't currently multitorrent capable,
but adding in support would be fairly easy (see PeerAcceptor.java:49)
* Don't allow leaseSets expiring too far in the future (thanks postman)
The build in tracker has been removed for simplicity.
Example usage:
java -jar lib/i2psnark.jar myFile.torrent
or, a more verbose setting:
java -jar lib/i2psnark.jar --eepproxy 127.0.0.1 4444 \
--i2cp 127.0.0.1 7654 "inbound.length=2 outbound.length=2" \
--debug 6 myFile.torrent
* Bugfix for the auto-update code to handle different usage patterns
* Decreased the addressbook recheck frequency to once every 12 hours
instead of hourly.
* Handle dynamically changing the HMAC size (again, unless your nym is
toad or jrandom, ignore this ;)
* Cleaned up some synchronization/locking code
* Allow an env prop to configure whether we want to use the backwards
compatible (but not standards compliant) HMAC-MD5, or whether we want
to use the not-backwards compatible (but standards compliant) one. No
one should touch this setting, unless your name is toad or jrandom ;)
* Added some new dummy facades
* Be more aggressive on loading up the router.config before building the
router context
* Added new hooks for apps to deal with previously undefined I2NP message
types without having to modify any code.
* Demo code for using a castrated router for SSU comm (SSUDemo.java)
* Bundled dust's Sucker for pulling RSS/Atom content into SML, which can
then be injected into Syndie with the Syndie CLI.
* Bundled ROME and JDOM (BSD and Apache licensed, respectively) for
RSS/Atom parsing.
2005-10-13 jrandom
* SSU retransmission choke bugfix (== != !=)
* Include initial transmissions in the retransmission choke, so that
if we are already retransmitting a message, we won't send anything
to that peer other than that message (or ACKs, if necessary)
* SSU retransmission choke bugfix (== != !=)
* Include initial transmissions in the retransmission choke, so that
if we are already retransmitting a message, we won't send anything
to that peer other than that message (or ACKs, if necessary)
* Choke SSU retransmissions to a peer while there is already a
retransmission in flight to them. This currently lets other initial
transmissions through, since packet loss is often sporadic, but maybe
this should block initial transmissions as well?
* Display the retransmission bytes stat on peers.jsp (thanks bar!)
* Filter QUIT messages in the I2PTunnelIRCClient proxy
* Piggyback the SSU partial ACKs with data packets. This is backwards
compatible.
* Syndie RSS renderer bugfix, plus now include the full entry instead of
just the blurb before the cut.
* Piggyback the SSU explicit ACKs with data packets (partial ACKs aren't
yet piggybacked). This is backwards compatible.
* SML parser cleanup in Syndie
* Implemented a new I2PTunnelIRCClient which locally filters inbound and
outbound IRC commands for anonymity and security purposes, removing all
CTCP messages except ACTION, as well as stripping the hostname from the
USER message (while leaving the nick and 'full name'). The IRC proxy
doesn't use this by default, but you can enable it by creating a new
"IRC proxy" tunnel on the web interface, or by changing the tunnel type
to "ircclient" in i2ptunnel.config.
2005-10-10 jrandom
* I2PTunnel http client config cleanup and stats
* Minor SSU congestion tweaks and stats
* Reduced netDb exploration period
* Implemented a new I2PTunnelIRCClient which locally filters inbound and
outbound IRC commands for anonymity and security purposes, removing all
CTCP messages except ACTION, as well as stripping the hostname from the
USER message (while leaving the nick and 'full name'). The IRC proxy
doesn't use this by default, but you can enable it by creating a new
"IRC proxy" tunnel on the web interface, or by changing the tunnel type
to "ircclient" in i2ptunnel.config.
2005-10-10 jrandom
* I2PTunnel http client config cleanup and stats
* Minor SSU congestion tweaks and stats
* Reduced netDb exploration period
* Now that the streaming lib works reasonably, set the default inactivity
event to send a 0 byte keepalive payload, rather than disconnecting the
stream. This should cut the irc netsplits and help out with other long
lived streams. The default timeout is now less than the old timeout as
well, so the keepalive will be sent before earlier builds fire their
fatal timeouts.
