> Upon inspection, the code was correct but one of the #defines was
> causing the problem for European Q.931. I have a hard time believing the
> original #define was right...
It wasn't. The page at
http://freesoft.org/CIE/Topics/126.htm
for example, indicates that the first octet of an information element
has:
0x80: bit indicating whether the IE is single-octet or
multi-octet (presumably along the lines of "bit 8 set
means this is the last octet of the item to which it
belongs, bit 8 not set means more stuff follows")
0x70: for some single-octet IEs, the IE type
0x0F: for those single-octet IEs, the data
0x7F: for some single-octet IEs, the IE type (no data)
However, most tables of IE type values include bit 8 in the value, and
two IE type values that differ only in the setting of bit 8 refer to
different types of IEs.
> Plus, this diff fixes the time display.
If you make any ISDN calls after 1999-12-31 23:59:59, could you grab a
trace of them? Q.931 isn't particularly informative about the way to
interpret the fields of the Date IE, and I'm curious whether the year is
supposed to go from 99 to 00 or to 100....
Both changes have been checked in.
