PE>> Fortunately I am rarely wrong.
 PE>> A single 00 marks the end.

 BL> Unfortunately, this is one of those rare occasions when you are
 BL> wrong. FTS-0001 dictates a null terminator for each message,
 BL> but the end of the packet adds another two 00's.

 PE> Bloody hell, I didn't even make it through the next paragraph
 PE> with my reputation for flawlessness intact. It was a genuine
 PE> mistake though.

  ROFL! But it made me a very happy old builder . I love it when 
someone else makes mistakes after sweeping statements.

 PE> As for being illogical, the idea of the packet is that you read
 PE> it in sequentially. If you had read the fixed header bit first,
 PE> which is 50 bytes (or whatever), then you search for a NUL 4
 PE> times (or whatever), that is the proper way to go from message
 PE> to message. The idea is not to hop into the middle and try to
 PE> fight your way out. Read one message at a time, and any you
 PE> want to delete, delete. BFN. Paul.

  Yair. That's alright for you plutocrats with DX8/200 machines, but I
have to make do with my old steam-driven 386 with a slipping belt. If
I just go looking for the twit and fight my way out, I can do a 150K
packet in 1.5 seconds. If I read every message it takes 22 seconds. I
tried it the sensible way first, but my way is more fun.

  But I stand by what I said. It is illogical to use 20 00 as the
identifier. They should have used 4 bytes, and anything other than 00.
Personally, I would have use KcUf as the identifier, and fixed-length
headers.

Regards,
Bob
  
___ Blue Wave/QWK v2.12
@EOT:

---