That's the one area that's handicapped by sticking to
the "you touch it, you maintain it" approach to
developing off Ofbiz.  I try to do everything without
touching ofbiz's code base.  

Your personal approach wouldn't help in this debug
scenario because I want the log message "service xyz
started" "service xyz ended" for 98% of the services
that are running.  The problem is the remaining 2% of
the services create 98% of the log.

The only way to comment out the log entry is to
disable it for every service. 

--- David E Jones <[EMAIL PROTECTED]>
wrote:

> 
> Well, personally in development I often comment out
> or add log  
> statements to make things easier to track down. It's
> pretty easy and  
> quick to change, build, and run.
> 
> -David
> 
> 
> On Dec 31, 2006, at 2:28 PM, Chris Howe wrote:
> 
> > I like the log message in general.  It's probably
> the
> > most useful debug message when trying to figure
> out
> > where a problem may lie in development. It's just
> > certain common services I may run a thousand or
> more
> > times when iterating through a list and that's
> going
> > to end up being 2000 lines of debug messages
> (roughly
> > half a meg) where I'm fairly confident the problem
> > doesn't lie.  If I could find a way to prevent
> those
> > couple of services from announcing their presence,
> > life would be grand :)
> >
> > The environment where this is a concern, is in
> > development.
> >
> > --- David E Jones <[EMAIL PROTECTED]>
> > wrote:
> >
> >>
> >> There should be "info" level messages and I'd say
> >> turn them off in
> >> production and live with them in
> >> development/testing.
> >>
> >> Or if enough people don't like them we can always
> >> get rid of them...
> >>
> >> -David
> >>
> >>
> >> On Dec 31, 2006, at 1:40 PM, Chris Howe wrote:
> >>
> >>> I find myself using certain services quite a lot
> >> (ie
> >>> partyNameFromDate when iterating through a large
> >> data
> >>> set) and would like it not to clutter up the log
> >> with
> >>> all of its service start/service end log
> entries.
> >> At
> >>> the same time, I need to be able to see when
> other
> >>> services start/end.
> >>>
> >>> I've thought of a number of ways to handle this.
> >> What
> >>> are your thoughts (community)?
> >>>
> >>> 1. attribute for service tag defining the
> service
> >>> start/end level
> >>> 2. attribute for element calling it (in simple
> >> method
> >>> <call-service/> or in screens <service/>)
> >>> 3. attribute for either the service or the
> service
> >>> call that raises or lowers (+1, -1) the level of
> >>> debugging for the entire service.
> >>> 4. learn to continue living with the noisy log
> >>>
> >>> Thanks!
> >>
> >>
> >
> 
> 

Reply via email to