signal transmission in ssh2

Markus Friedl markus at
Sat Feb 2 02:41:28 EST 2002

On Thu, Jan 31, 2002 at 11:19:13PM +0100, Peter Stuge wrote:
> It's part of the draft, that's reason enough to put it in, IMO.

but it's not a 'must'.

> I like it.
> If anyone feels like it, the sig2name/name2sig could be made more complex to
> allow for arbitrary signal passing from a given system to a remote system of
> the same kind, which could be useful.
> I'm thinking of the draft text in section 4.10 of the connection protocol;

this is the part that i don't like:

> Additional signal names MAY be sent in the format "sig-name at xyz", where
> `sig-name' and `xyz' may be anything a particular implementor wants (except
> the `@' sign).  However, it is suggested that if a `configure' script is
> used, the non-standard signal names it finds be encoded as
> "SIG at xyz.config.guess" where `SIG' is the signal name without the "SIG"
> prefix, and `xyz' be the host type, as determined by `config.guess'.

More information about the openssh-unix-dev mailing list