From mboxrd@z Thu Jan 1 00:00:00 1970 From: Pete Zaitcev Subject: Re: [PATCH 0/6 v2] logging refactoring Date: Tue, 15 Dec 2009 15:47:33 -0700 Message-ID: <20091215154733.39cfb54f@redhat.com> References: <1260317516-30991-1-git-send-email-cmccabe@alumni.cmu.edu> <436f52800912132323paf46ea2j6392ce81e13a7981@mail.gmail.com> <4B26A1C0.2080401@garzik.org> <436f52800912141818s1dea5f9bg2d156597210c7bb8@mail.gmail.com> <20091215111904.05b1b260@redhat.com> <4B280FDA.9010705@garzik.org> Mime-Version: 1.0 Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <4B280FDA.9010705@garzik.org> Sender: hail-devel-owner@vger.kernel.org List-ID: Content-Type: text/plain; charset="us-ascii" To: Jeff Garzik Cc: Colin McCabe , Project Hail List On Tue, 15 Dec 2009 17:38:18 -0500 Jeff Garzik wrote: > > In fact, does anyone even understand why we continue to split > > LOG_WARNING and LOG_ERR? We really have three log levels: > > - debug > > - info (which is actually an error) > > - error (not really all that different from info actually) > > Might as well create HAIL_DEBUG, HAIL_INFO, and HAIL_ERR (to be > > used before exit(1), with values of 0,1,2 and remap to syslog > > values where it goes to syslog. > > It would be nice to disconnect from syslog, yes. But the 'warning' log > level has value. Can you offer any guidelines that are easy to follow and enforce? I am often confused and select the level randomly. The only rule I follow is: if and only if it's under if (debugging), then LOG_DEBUG. Otherwise... hmm... -- Pete