Todo's sorted by priority:

BNS
=========================
(1) Reconnect of "Broadcast Ephemeris" stream upload to caster does not work.
    If we shut down the caster, wait a minute and then restart it, stream
    RTCM3EPH does not reconnect.
(2) GLONASS-IOD 7-bit-Problem, explained by email from Andrea Stuerze.

BNC
=========================
(1) Change time tag for GLONASS ephemeris from GPS time to GLONASS time (UTC).

(2) We may also have a date problem in the GLONASS Ephemeris time tag.
    We once saw a wrong date.

(3) Oscar Colombo describes the following bug in the BRDC and CLK* time tags:
    "I have found an odd situation concerning the correspondence between the epochs
    of data in "RTCM3EPH" and of those of the data in "CLK10". 
    In the file extract below you can see that the "CLK10" corrections to the orbits
    and clocks are for an epoch close to Greenwich midnight of November 5, 2009
    (when I was collecting it with BNC 1.7), but the orbits and clocks with the
    same IODE are the broadcast values for 15:59 hs and 16:00 hs: some 8 hours
    earlier. So those are midnight corrections to the broadcast orbits and clocks
    from 8 hours earlier."
    File extract:
    ....
    BRDCRR  1060 0 1556 430940.0 G28    44      9.498  -46.343 -155.593  147.576
    BRDCRR  1060 0 1556 430940.0 G29    22     -3.045  -58.524  133.230  116.716
    BRDCRR  1060 0 1556 430940.0 G30    51      3.554  -18.734   22.571 -174.303
    BRDCRR  1060 0 1556 430940.0 G31    29     -5.229  116.870   77.207 -113.687
    BRDCRR  1060 0 1556 430940.0 G32    79    -12.534  179.447  -90.711  144.278
    RINAVH 027 09 11  5 15 59 44.0 0.973059795797E-04 0.318323145621E-11 0.000000000000E+00            
    RINAVR 0    0.190000000000E+02-0.167312500000E+03 0.398159460957E-08 0.242476843284E+01            
    RINAVR 0   -0.858306884766E-05 0.215127451811E-01 0.515021383762E-05 0.515372610092E+04            
    RINAVR 0    0.403184000000E+06 0.279396772385E-06-0.194402304722E+00-0.670552253723E-07            
    RINAVR 0    0.976590072451E+00 0.287843750000E+03-0.152511337870E+01-0.761210294797E-08            
    RINAVR 0   -0.272154188075E-09 0.100000000000E+01 0.155600000000E+04 0.000000000000E+00            
    RINAVR 0    0.200000000000E+01 0.000000000000E+00-0.419095158577E-08 0.190000000000E+02            
    RINAVR 0    0.396018000000E+06 0.400000000000E+01 0.000000000000E+00 0.000000000000E+00 
    ....

(4) Ntrip version "U" (for UDP) works with Dirk Stoecker's caster at
    www.dgpsonline.eu:2101 but not with our own Caster at
    euref-ip.bkg.bund.de:2101. Check situation with mountPoints string:
    mountPoints=//bkg:realtime06@141.74.33.12:2101/ZOUF0 RTCM_2.3 46.56 12.97 no U,
               //weber:bkg@www.dgpsonline.eu:2101/WILD_RTK3 RTCM_3 52.36 13.56 no U

(5) Bandwidth graphics: one column per stream showing the current 
    bandwidth needed. Please consider this as "nice to have" - it's not a must.
    What we have is marked with "Begin new Perlt" and "End new Perlt" in bncwindow.cpp 
    and bncwindow.h. If you don't like it or if its to difficult to implement
    it then: please delete it form the source code.

(6) GW: Keep an eye on www.igs-ip.net/PENC0.

RTNET
=========================
(1) SV missing in solution after temporary exclusion (e.g. SV not in IGU orbits or in SAT_2009.CRX), only on clock-ip

