This is a very minor point but it needs clearing up.
From review by Elwyn Davies:
Possible issue:
[I say 'possible' because I am not an ssh expert but there is an
apparent inconsistency with other ssh documents which makes me wonder].
s2: para 4: The text says 'If the BREAK-length parameter is 0 *or not
present*, the BREAK SHOULD be interpreted...'. As far as I can see no
other ssh message has optional parameters in this way. Although it
would obviously be possible to cope with both cases, it seems to be
unusual and makes parsing the message more complex than it needs to
be, given that this message is going to be a relative rarity. Was this
intended? If so I think it would be desirable to add an explicit note
closer to the message definition to point out that the parameter is
optional. Otherwise just delete 'or not present'.
BC: As I read the spec, the channel request always includes
uint32 break-length in milliseconds
so the case where the break-length parameter is absent simply doesn't
exist. If that's correct, indeed just delete 'or not present'.