> -----Original Message----- > From: Heath Borders [mailto:[EMAIL PROTECTED] > Subject: Re: Warning messages > Well, the reason I would want to skip it is because of the > performance hit it will cause. > For every UIInput, for instance, we would have to traverse up > the tree to find the containing UIForm.
Ah yes.. In that case doing isWarnEnabled() before looking up the form is definitely useful. Kalle > On Mon, 21 Mar 2005 14:03:18 -0800, Korhonen, Kalle > <[EMAIL PROTECTED]> wrote: > > > On Mon, 21 Mar 2005 21:06:46 +0100, Manfred Geiler > > > <[EMAIL PROTECTED]> wrote: > > > > Yes, or better isWarnEnabled(). Warn level is probably the best > > > > for this kind of warnings. > > > > ... and I wouldn't even bother doing isWarnEnabled() check. Those > > enabled() methods are basically only provided to skip over > unnecessary > > string concatenation. Really useful for debug() logging but > not so for > > warn messages, which there usually are a lot less and they > are almost > > always enabled anyways. > > > > Kalle > > > > > > > > > > Manfred > > > > > > > > On Mon, 21 Mar 2005 14:00:14 -0600, Heath Borders > > > > <[EMAIL PROTECTED]> wrote: > > > > > I guess if I just did a check for isDebugEnabled() > first before > > > > > going through the hassle of searching for a parent form, it > > > > > would have the same result. > > > > > > > > > > > > > > > On Mon, 21 Mar 2005 11:39:34 -0800, Korhonen, Kalle > > > <[EMAIL PROTECTED]> wrote: > > > > > > > -----Original Message----- > > > > > > > From: Heath Borders [mailto:[EMAIL PROTECTED] > > > > > > > Subject: Warning messages > > > > > > > Do you guys think we should have warning messages for > > > things we > > > > > > > know will cause components to not function properly, and > > > > > > > then have a variable in the web.xml to turn it off? > > > > > > > We could have warning messages for all components > that need > > > > > > > forms, or for components using the "for" attribute > > > that cannot > > > > > > > find their parent component, etc. > > > > > > > These could all be enabled by default. > > > > > > > Then, there could be a context-variable named > > > > > > > "org.apache.myfaces.warnings" which could be set to > > > "false" or "true" > > > > > > > to turn it off or on. > > > > > > > > > > > > Otherwise, I'm definitely for more logging and warning > > > messages, > > > > > > but please let's not add a new context-variable for > this. Even > > > > > > newbies should be able to configure java/commons/log4j > > > logging and > > > > > > if not, they are much better off with learning to > > > configure those > > > > > > properly than some other custom logging properties. I > > > can also see > > > > > > that that pretty soon your simple and harmless > warnings on/off > > > > > > would be changed to logging level, then a few more settings > > > > > > for redirecting etc... So let's keep logging > configuration in > > > > > > the standard logging properties. Surely we can provide > > > default logging > > > > > > properties where warnings are enabled (and usually > > > anything above info is logged by default anyways). > > > > > > > > > > > > Kalle > > > > > > > > > > > > > > > > -- > > > > > -Heath Borders-Wing > > > > > [EMAIL PROTECTED] > > > > > > > > > > > > > > > > > > -- > > > -Heath Borders-Wing > > > [EMAIL PROTECTED] > > > > > > > > -- > -Heath Borders-Wing > [EMAIL PROTECTED] >
