[j-nsp] allow-command question

[Guy Davies]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

> On Thu, Jul 31, 2003 at 08:59:37AM +0200, Hannes Gredler wrote:
> > On Mon, Jul 28, 2003 at 11:09:51AM -0400, Jared Mauch wrote:
> > | 	IMHO, you should never disallow quit.
> > | 
> > | 	ie: juniper should make it always available.
> > 
> > behave differnt than the regexp does dictate ?
> > 
> > i was under the impression that customer do
> > prefer predictability over fool-protection ? 
> 
> 	I agree with that in everything except the ability to 'quit'
> 
> 	You shouldn't be able to telnet/ssh/whatnot into a device but
> not be able to exit gracefully.

I'm with Jared on this one.  There are no circumstances I can think of under
which it would be necessary to *prevent* someone from being able to quit.  I
don't know how trivial (or otherwise) it would be to make quit and exception
within the regexp behaviour.  I would have thought that some implicit regexp
entries which exclude quit from a "deny-commands" would do the trick.

In all other aspects, I want determinism.  This is just the situation where
common sense should take precedence over the technically pure solution.

Regards,

Guy

-----BEGIN PGP SIGNATURE-----
Version: PGP 8.0

iQA/AwUBPykcGo3dwu/Ss2PCEQK1jwCgnXGWMIFbljr5SQPgcwtJMsfsuhsAn2h7
M2OHhoYTM1O1PhObQ3BTWFB4
=Fpe4
-----END PGP SIGNATURE-----