On Wed, Jan 26, 2000 at 07:03:45PM +0100, [EMAIL PROTECTED] wrote:
> > Oh OK.  Actually, I haven't heard of anyone using inactivity watchers.
> > Maybe they should watch a group of watchers instead of event
> > priorities...
> 
> Watching of watchers sounds really interesting!

Oh?  Would you use it like an extra timeout?

> > It is restricted to alphanumerics.  I don't want to allow fancy stuff
> > like (?{ ... }).  Does anyone know what is reasonable to allow in a REx?
> 
> I do not know such rules and found none in my Perl books (I do not own Friedls 
>regexp book, may be
> there ...). But - in my opinion - is it really useful to restrict a user of such a 
>tool? I think no
> real server should open such a port. It should always run under developers control. 
>If someone can
> suspend running watchers and change their priority at his will, what worse could he 
>do with regular
> expressions?

Fair enough.  Maybe I am being unrealistic about safety.  Ah well, I
unrestricted them in my copy.

> > > ... The "t" control accepts only certain values (e.g. second value
> > > 100 becomes 120, 150 becomes 180, but 54 and 39 work ...).
> >
> > Yah, that's just how the stats are collected.
> 
> Ah, ok. Possibly a short doc note to avoid confusion ...?

Done.

> So don't I. If you refresh the screen, the command line contents is still under the 
>(telnet) clients
> control, so the server has no access to it, I assume ... hm!

Yup.  I'd need to write a client which turned off line buffering, and
that would spoil the easy of use to some extent.

> > Hm.  I don't want to change anything until you are sure you need it...
> 
> So let's see first what the currently running debugging and ..::ProcessTop sessions 
>will show, and I
> will ask again for these new levels when I see no other way to solve the load 
>balance problem.

Sounds good.  Bear in mind that high loads cannot always be scheduled
properly because they are, well, too loaded.  But of course you already
knew that.  :-)

-- 
"Never ascribe to malice that which can be explained by stupidity."
                            via, but not speaking for Deutsche Bank

Reply via email to