Going to put this on some test routers very soon.
A long “changelog” on the download page:
What’s new in 2.9rc7:
*) fixed bug - wireless scan sometimes did not work;
*) fixed bug - wireless security-profile changes did not reset affected
interfaces;
*) fixed bug - Ethernet ports & miniPCI slots on daughterboard was locking up;
*) fixed bug - some ppp users where not removed from active list on lossy links;
*) removed replay checking when AES-CCM or TKIP used with static keys;
*) added back rip and ospf to routing-test package;
*) fixed nstreme2 transmit power setup;
*) fixed wild card url in web-proxy access list to be a substring;
*) fixed bug - web-proxy access list entries were not used in correct order;
*) added default web-proxy access rule to block dst-port=23-25;
*) added direct access list for built-in http proxy;
*) added support for multiple HTTP page sets for the same hotspot server;
*) fixed bug - hotspot sent only 1st RADIUS Class attribute in
accounting packets;
*) added option to select Nas-Port-Type value for hotspot RADIUS;
*) improved hotspot not to account idle-timeout as part of session time;
*) added ability to bypass whole ip subnet in hotspot;
*) improved upload/download speed of files to/from WinBox;
*) fixed bug - WinBox loader from v2.8 did not work well with v2.9;
*) fixed bug - PPPoE/PPTP/L2TP ignored profile rate-limit when authenticating over RADIUS;
*) fixed bugs with IPsec that made active policies invalid;
Hi Stephen,
The RC7 Works like a lightning with the special wierless npk file !.
test it 
While it might be fast, I’m seeing issues with the wireless-test package and WPA-PSK/TKIP connections.
I tested the wireless-test package on a RB500 in ap-bridge mode against WPA-PSK/TKIP Windows-XP
clients and could not get the clients to connect (the RB500 was logging errors about “group key timeouts”).
I’ll investigate this in more detail and contact support@ with the usual supout etc.
–Tom
Thanks all
Tom - do post your problems/results here too.
If we’re going to test this seriously it’s best that we pool results, and that’ll help MT get the problems resolved.
IMHO this “CPU-friendly” Nstreme is the way to go for MT, but needs to be debugged swiftly.
Regards
Well, OK, could not get the WPA-PSK/TKIP stuff to work.
Here’s what is logged by the router
09:32:33 wireless,info wlan1: 00:04:23:76:D7:6C connected
09:32:34 wireless,info wlan1: disconnected 00:04:23:76:D7:6C, decided to deauth: group key update timeout (16)
09:32:34 wireless,info wlan1: 00:04:23:76:D7:6C connected
09:32:34 wireless,info wlan1: disconnected 00:04:23:76:D7:6C, decided to deauth: group key update timeout (16)
09:32:41 wireless,info wlan1: data from unknown device 00:04:23:76:D7:6C, sent deauth
09:32:45 wireless,info wlan1: 00:04:23:76:D7:6C connected
09:32:46 wireless,info wlan1: disconnected 00:04:23:76:D7:6C, decided to deauth: group key update timeout (16)
09:32:53 wireless,info wlan1: data from unknown device 00:04:23:76:D7:6C, sent deauth
09:32:59 wireless,info wlan1: 00:04:23:76:D7:6C connected
09:32:59 wireless,info wlan1: disconnected 00:04:23:76:D7:6C, decided to deauth: group key update timeout (16)
09:33:06 wireless,info wlan1: data from unknown device 00:04:23:76:D7:6C, sent deauth
this continues forever.
I’ve tested this with the builtin Intel card of a “Centrino” Notebook using the Intel PROset drivers as well as with an Orinocco PCMCIA card with Odyssey client for WPA. Makes no difference, both lead to the same error shown above.
–Tom