[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Pengfork-devel] Logging and debug facilities
From: |
Jean-Charles Salzeber |
Subject: |
Re: [Pengfork-devel] Logging and debug facilities |
Date: |
Fri, 30 Aug 2002 01:30:27 +0200 |
User-agent: |
Mutt/1.5.1i |
On Thu, Aug 29, 2002 at 23:39, Nicolas Burrus wrote:
> On Thursday 29 August 2002 23:17, Jean-Charles Salzeber wrote:
> >
> > well, I've some remarks:
> > * the log function always write to stderr *AND* syslog, is this the
> > desired behaviour?
> > * I was thinking that log should write to syslog if in daemon mode and
> > to stdout/stderr in command line mode
>
> Ok, why not, as stderr, stdout, ... are disabled in daemon mode it would
> react
> the good way, but a test is cleaner. I don't know it is standard to syslog
> when running in interactive mode, but if it is I think it can be great to
> keep a trace even when not running in daemon mode.
There is a misunderstanding, I said
daemon mode => syslog
interactive mode => stdout/stderr
So we are agree :)
Syslog is not standard in interactive mode.
I don't know if it is useful to have log on both stderr and syslog in
daemon mode. You were using the LOG_PERROR as define, why not having a
parameter (ex: PARAM_FORCE_LOG_PERROR)?
>
> > * Log may write to stderr only if level >= LOG_WARNING
>
> Why not, and maybe write all syslog msg if PARAM_DEBUG_LEVEL is > 0 ?
I don't understand what you mean.
I was meaning: level >= LOG_WARNING -> stderr
level < LOG_WARNING -> stdout
>
> > * debug should write to stdout not stderr debug is normally not error ;)
>
> Right :)
>
> --
> Nicolas
JC
pgpADUrLEZRJv.pgp
Description: PGP signature