DO NOT REPLY TO THIS EMAIL, BUT PLEASE POST YOUR BUG RELATED COMMENTS THROUGH THE WEB INTERFACE AVAILABLE AT <http://issues.apache.org/bugzilla/show_bug.cgi?id=28237>. ANY REPLY MADE TO THIS MESSAGE WILL NOT BE COLLECTED AND INSERTED IN THE BUG DATABASE.
http://issues.apache.org/bugzilla/show_bug.cgi?id=28237 [PATCH] Use the commons logging LogFactory also in Fop.java ------- Additional Comments From [EMAIL PROTECTED] 2004-04-18 19:28 ------- Glen, On Sat, Apr 17, 2004 at 08:36:13AM -0000, [EMAIL PROTECTED] wrote: > ------- Additional Comments From [EMAIL PROTECTED] 2004-04-17 08:36 ------- > Simon (and others), > > For the remaining three issues remaining, please take a look at the patch I just > added. > > 1.) I removed the -debug and -quiet options, because they just cover over the > commons-logging settings. Also I'm leery of switching to SimpleLog as a > default, seeing the Commons Logging will normally use JDK1.4 logging. > > Rather than use an API that covers up Commons-Logging functionality (i.e., -d > and -q), I would rather have us make it very simple for the user to use/learn > the exact commons-logging method, which brings up: I thought that we should retain backward compatibility, but if that is not (yet) the case for the development branch, I am most happy to see these options be dropped. > 2.) In fop.bat and fop.sh, (I only tested the former, I'll need someone to > check the latter), I created commented-out lines that will allow the user to > choose a different logger from the default JDK1.4 and the logging level (the > latter, iff they're using SimpleLog). That is very helpful. Frankly I think that users who are not satisfied with the defaults really should study log configuration. But giving them a headstart does not hurt us. > 3.) Per your last comment below, on using .info() instead of .debug() in the > dumpConfiguration() nee debug() method--I agree, that should be the message > level instead because that is what the user is specifically requesting. My > patch does this, but without switching to a SimpleLog() to guarantee output. > The drawback is that if the user specifically chooses "error" or "fatal", etc., > as a level, that will take precedence over setting this field, and as a > consequence this information won't be shown. OTOH, the logger that the user > specifically chose externally won't be overridden. I suspected the latter was > the lesser of two evils, and so went with that. Your thoughts? With the info level, it is OK not to check whether the user has info level enabled; it is the default. I am quite happy with your patch. Peter, I generally agree with Glen's point of view. With this solution we have chosen a good and well-respected logging solution, which has the good sense to act as a middle man who gives the user a choice of logging implementation. Answering queries about logging behaviour and features is no longer on our shoulders. Regards, Simon
