Le samedi 06 novembre 2004 à 20:41 +0100, Gilles Espinasse a écrit :
> I had to adapt the GUI for the 4 differents CMV selection.
hum, you are more advanced than us ;-)
> During installation I rename CMVep.txt and CMVei.txt to CMVepWO.txt,
> CMVEiWO.txt.
> This let CMVep.txt and CVMei.txt used by the driver free to ease symlink
> creation with those names.
I had this in mind as well (to use symlinks), though changing eagle-usb.conf is
the reason why it exists.
Currently, using a symlink seems more flexible though.
> I make 4 options available in a LINE variable :
> WO (world : others countries)
> ES ESPANA
> FR FRANCE
> IT ITALIA
I still do not understand the reason why it can be country related.
For me, it's DSLAM related but I may be in error.
> And now I just symlink for the selected CMVep only since there only one
> CMVei.txt
> /bin/ln -f -s /etc/eagle-usb/CMVep$LINE.txt /etc/eagle-usb/CMVep.txt
> /bin/ln -f -s /etc/eagle-usb/CMVeiWO.txt /etc/eagle-usb/CMVei.txt
> and it should work
hum, so we should put on CVS a CMVe{i|p}WO.txt rather than CMVei.txt and
CMVep.txt (it can be done at installation...)
If we choose WO, tell me how to erase CMVe{i|p}.txt from CVS (I know how to add
a file, not how to remove it...)
> Actually CMVei.txt use old values, it could be updated with Win2.031 values
> ( Win2.2.31 has same values)
> ; ISDN for EAGLE
> ; Number of CMV writes made in this file MUST not exceed 16
> ; version DTAG pour 40e4be96
>
> CW cntl 0 1
> CW optn 0 80000066
> CW optn 2 23600000
> CW optn 3 00003028
> CW optn 4 00000001
> CW optn 5 00000500
> CW optn 8 C0000000
> CW optn 15 09090909
> CW optn 18 82602180
> CW optn 20 00020038
> CW optn 21 FFFFFE2F
> CW optn 22 020600c0
> CW info 9 0400000A
> CW info 21 5341
> CW info 22 474D0000
> CW cntl 0 2
>
> ;For Bridge ONLY CW DIAG 0 23870030
> ;For Bridge ONLY CW FLAG 0 00000820
current version is (was) :
CW optn 0 80020066
CW optn 2 23600000
CW optn 3 00000026
CW optn 4 0
CW optn 5 500
CW optn 15 09090909
CW cntl 0 2
I thought that OPTN0 with 80020066 corresponds to a "testing"
requirement that enables compliant DSLAM to work better (I may be in
error ;-) of interpretation), non compliant DSLAM would just do what
they can (but still work).
Other different options are OPTN3 & OPTN4 : as long as we do not know
what they represent or control ... well we can just guess. That's the
kind of information that is valuable from ADI. It permits to choose
better parameters for the end-user.
I've updated CVS with your latest CMV file (as it will not yet be
used... in 2.0.0)
Though, I'd prefer to understand the reasons of change rather than doing
them blindly ;-) "it works better" would be a start but not that really
convincing... I would really appreciate ADI's feedback or Sagem's return
of experience ;-) (keep in mind that it makes it work better for the
end-user if you are still reluctant to provide currently-not-authorized
information ;-) ).
@++
Ben'. aka baud123
