In article <[log in to unmask]>, Jim
Phillips <[log in to unmask]> writes
>Bruce Toback writes:
>
>>1. If method #1 ever loses sync -- that is, if it ever misinterprets data
>>as a length -- the chance of recovering is tiny, and the chance of
>>recovering without damaging data somewhere is tiny squared.
>
>
>Of course if tiny < 1 then:
>
>Tiny**2 < Tiny
>
>and the saying is correct.
>
tiny is a probability. As 1 is certainty, tiny cannot be > 1.

So tiny**2 <= tiny.

'Tiny', in this context, is an undefined variable in some nightmare
computer language which is case-sensitive. (Anybody know of one?)


Back on topic, what about using method 2 to guard against problems with
method 1?

i.e. use method 1 for efficiency, but check the final character sent,
which should be the defined terminator?

If it isn't, then method 1 knows that it has lost sync, and 'tiny' gets
back up near 1.

--
Roy Brown               Phone : (01684) 291710     Fax : (01684) 291712
Affirm Ltd              Email : [log in to unmask]
The Great Barn, Mill St 'Have nothing on your systems that you do not
TEWKESBURY GL20 5SB (UK) know to be useful, or believe to be beautiful.'