Re: Server crash in /city/anthony/prison

[Jan Echternach <echter@informatik.uni-rostock.de>]

On Mon, Jun 19, 2000 at 09:11:46AM +0200, Neil Muller wrote:
> On Sun, Jun 18, 2000 at 09:04:32PM +0200, Jan Echternach wrote:
> There are two obvious advantages to implementing different error levels.

I still don't think these are worthwhile advantages because if you get
very few error messages, you almost never get error messages of
different types.

> >   llevCritical
> Never off, increments nroferrors, needs not be split into subsections

Why subsections?  Any llevCritical should be fixed immediately because
it is likely to corrupt the server's internal data structures,
producing spurious errors.

> >   llevError
> Switchable. (Incrementing nroferrors for these might be a
> commandline option.) I still think subsectioning would be useful.

Subsectioning might be useful here, assuming that there can be so many
error messages that you need to turn off some.  But I doubt that this
will happen more than once in a blue moon.

The only thing that really needs subdivision is llevDebug IMHO.

> >   llevInfo
> Info strikes me as additional information provided on the running of the
> server, either statistics or information on how the code is running. 
> Definately should not increment nroferrors. 

This is already implemented (and it does not increment nroferrors, of
course).  grep llevInfo crossfire/*/*.c for usage examples.

-- 
Jan