A new release for community has hit the demonsaw.com downloads section.
Here are the latest updates since 3.0 went was launched at DEFCON 24:
[3.0.2]
* queue unauthorized packets then emoty queue once re-authenticated with xfr router
* transfer routers should not be overwritten when refreshing with msg router (keep in memory and replace only). msg router sends back all xfr routers to clients (not just success ones)
* transfer service unauthorized will drop transfer request (to restart). Re-queue and/or don’t restart, just re-group & handshake? Add to post-auth queue and then process upon successful re-connect?
* restart of client should NOT restart transfer routers (allow transfers to continue)
* turn yellow if you have 0 xfr routers
* use transfer router as long as not error status
* max/min transfer routers that are used in transfer? (1 – 8)
* transfer router open fail should abort? Or.. use TTL and fail when we run out of retries?
* right click download, Stop doesn’t appear cross rows (if yes/no)
* fix browse sort crash bug
* wrap all service/acceptor handlers in exception handling
* TwoFish crashes on linux??? (Just use AES for session?)
* 5 sec progress update change from green to blue
* removed early socket close function calls (should fix router crashes)
* return 501 if router uses non AES for session
* router socket->close() – let destructor do it
* porn’s weird file parse issue – i think an exception is thrown tbh
* Only allow folder browsing if folder size > 0
* alpha 1 sets button at 255 alpha?
* auto rejoin client to router if going from 0 -> 1 or 1 -> 0 shares
* minus button remove all (download, upload, etc.)
[3.0.3]
* Stack crash was occurring on Linux due to 4K worker thread stacks
* Increased stack size to 256K
* Added twofish algorithm back
* tooltip for dock widget
* Added tooltips back
* Updated retry to 1 sec (100 ms was DDOSing the router?)
* retry is set to 0-1 sec now…
* Add retries to services & acceptors
* Remove unnecessary error messages
* Unable to write needs to eventually time out?
* retry should be 8
* bring over random retry code from message acceptor to clients (0 – 1000)
* tunnel sockets to 1
[3.0.4]
* Restart ignores previous queued events
* unable to write still doesn’t timeout – it does, just takes 1 min
* playing while downloading corrupted file, resume fixed?
* 1 tunneled socket
[3.0.5]
* 1 thread for message routers
* Display “Upgrade Now” message in chat when incompatible router found