Text 30316, 55 rader
Skriven 2012-06-29 13:23:57 av mark lewis (1:3634/12.0)
  Kommentar till text 30302 av Robert Bashe (2:2448/44)
Ärende: binkd (was: Link requ... blah blah)
===========================================
[large snip]

 RB> My understanding was always that the _last_ node definition in the
 RB> binkd.cfg file overwrote everything previous, but the above
 RB> experience doesn't appear to confirm that... 

1. what version of binkd are you running?

my config has my connections, some with passwords and some without, listed
first and then the distributed BINKD.TXT is listed second. i'm password
protected with rossC and have no problems like you describe...

 RB> unless it really is _vital_ to have the number of fields matched 
 RB> in all definitions (either four or five fields).

i suspect that if you have one entry with a password first and then another
entry for that same node with a - in the password position that the - will
override the initial password setting... IIRC, the - means "nothing set for
this"...

 RB> I might note that binkd.txt, the file you referred to, also has
 RB> only _four_ fields:

right... as i pointed out before ;)

 RB> To be honest, I don't really understand what's happening, since the
 RB> binkd reference page states the node definition to be:

 RB> node [ftn_addr [options] [{hostlist|-} [{password|-} [flavour
 RB> [{obox|-} [{ibox|-}]]]]]

 RB> In other words, five fields.

the above definition is confusing, too... i see 7 fields and in my manual
configuration, i use all 7 because i use fileboxes and require the last two...

 RB> Since switching the order as indicated, I haven't had any incoming
 RB> polls, but will report on the results when I can.

 RB> At the moment, it looks like the number of fields must be matched
 RB> in _all_ node difinitions for the system to operate correctly.

i don't think so... i think it has to do what having an existing line that gets
overridden by following ones and if you are using a generic list with 5 fields
in it, then you are overwritting the 5th (password) field...

but then again, i'm used to the first definition being the one used and others
skipped... however, i have coded and used tools that do take following
definitions as updates or alterations to the one already built in memory... so
i can see how your list of passworded connections may be being overridden by
the later inclusion of a list that has a dash in the password field...

)\/(ark

 * Origin:  (1:3634/12)