- Make flush() block less, by waiting only for "accept" into the
streaming queue rather than "completion" (i.e. ACK from the far end).
This prevents complete window stalls when flushing, and should help performance
of apps that use flush(), like i2psnark (and SAM?).
close() still does a flush that waits for completion, as i2ptunnel
doesn't like a fast return from close().
- flush/close javadocs and comments
* i2ptunnel:
- Now that streaming flush() is fixed, use it in IRCClient, and
for initial data in I2PTunnelRunner, to avoid the 250 ms
passive flush delay
- Expire unreachable routers quickly, even if they don't have introducers,
so we don't have old data on routers that ran out of introducers.
- Fix rare NPEs at shutdown
- Add new option i2cp.messageReliability=none, which prevents the
router from sending MessageStatusMessages back in reply to an
outbound SendMessageMessage. Since the streaming lib always ignored
the MSMs anyway, make it the default for streaming.
This will reduce the I2CP traffic significantly.
MSM handling now avoided, but it is still fairly broken, see
comments in I2PSessionImpl2.
- Cleanups, javadoc, rate reduction
- Convert all ArrayList caching to LBQs in YKGenerator,
HMACGenerator, and AESKeyCache.
- Change DSAEngine params from Hash to new SHA1Hash, since
these were really 20 byte hashes, not 32 byte Hashes.
- Add stats to track YKGenerator caching success
- Fix YKGenerator precalculation to be much more useful by
increasing the cache size and dramatically shortening the delay
- Option cleanups
- YKGenerator cleanups
- Mark HMAC256Generator unused
* Base64: comment out some unused methods
* Remove huge whitespace in CryptoConstants
* ElGamalAESEngine:
- Reduce rates
- Check number of tags earlier
