[Vortex] Is vortex assuming that msg numbers on channels start at 1 and increase by 1/

Francis Brosnan Blazquez francis at aspl.es
Thu Nov 20 10:51:50 CET 2008


Hi Sam,

> Just to be clear, beepcore-c hasn't actually reused any msg numbers,
> it just decided to start channel zero messsage numbers at 101!

Sure, this was really clear from your previous mail Sam. To me this has
to do with the RFC definition that started all this issue, I mean,
reading:

"""
2.2.1.1 Frame Header
  [...]
  The message number ("msgno") must be a non-negative integer (in the
  range 0..2147483647) and have a different value than all other "MSG"
  messages on the same channel for which a reply has not been
  completely received.
"""

..it is clear that current vortex and beepcore-c approaches are valid to
generate MSG numbers, however, from all valid approaches, I wouldn't
start with 101 as beepcore-c do, but that isn't the point. 

The root issue for me is that the RFC states it is possible to reuse
message numbers, which is not allowed by Vortex at this moment.

> Btw, this comment from your last release:
> """
> Added support to allow buggy BEEP peers to use 0 as first MSG no
>  number on channel 0 (can be controlled via API), support to disable
>  SEQ frame generation (useful for tuning profiles).
> """
> 
> This, though, doesn't sound like buggy beep peers, it just sounds like
> msgno reuse.

Certainly, you are right. We have been assuming that message number
can't be reused, so we supposed that a peer behaving this way can be
considered buggy. This is wrong.

> The "greeting" msg on connnection comes across as RPY, chan=0,
> msgno=0. That means msgno=0 is now no longer in use, so it can be used
> for a new message.

Right. Cheers!

> Sam
-- 
Francis Brosnan Blazquez <francis at aspl.es>
Advanced Software Production Line, S.L.



More information about the Vortex mailing list