"Schaefer, Andreas" wrote:
>
> My suggestion is to make both because:
> - to make it final allows the user to suppress any logging if he/she likes
> to
No it doesn't: 'users' are going to grab the binary and need to be able
to turn on debugging to troubleshoot.
> and therefore increases the performance
> - make it settable from JMX to direct the logging
If it's final it isn't settable from JMX. Are you suggesting 2 flags for
the same thing?
>
> Both works fine together except that when you take out some loggings with
> the part one then you don't get this messages even when you set it with JMX.
OK, yes you are suggesting 2 flags for the same thing.
I suppose, but if we're talking about performance here, we need to
quantify the boost we get from having this thing be final. Also, if we
do this scheme, the binary distribution will really really need to be
built with the final variable true (debugging enabled)
> Maybe the JMX part can reject any changes to values which are not usefull
> because
> of part one.
>
> Have fun - Mad Andy
>
> -----Original Message-----
> From: danch [mailto:[EMAIL PROTECTED]]
> Sent: Friday, March 09, 2001 9:15 AM
> To: JBoss-Dev
> Subject: Re: [jBoss-Dev] [VOTE] debug variable - final or not?
>
> No, make it settable from JMX. That way when you want it true, you make
> it true, when you want it false, make it false. No recompile neccesary!
>
> Hiram Chirino wrote:
> >
> > I vote that we make the debug variable true by default. If someone wants
> to
> > optimize his execution, let him recompile. Notice that you should be able
> > to still reduce the amount of logging you produce at run time with an
> MBean
> > (It's just that it the logging calls will not be optimized out)
> >
> > Regards,
> > Hiram
> >
> > ----- Original Message -----
> > From: "Kimpton,C (Chris)" <[EMAIL PROTECTED]>
> > To: "'JBoss-Dev'" <[EMAIL PROTECTED]>
> > Sent: Friday, March 09, 2001 4:50 AM
> > Subject: [jBoss-Dev] [VOTE] debug variable - final or not?
> >
> > > Hi,
> > >
> > > >
> > > > I agree completely and vehemently with Tom. Please don't go backwards,
> > > > in deployed systems it must be as easy as possible to turn
> > > > trace output
> > > > on and off. Recompiling application code is not even an option,
> > > > recompiling your app server just to turn on trace is
> > > > something that just
> > > > can't happen.
> > > >
> > > > Please put this back. Please. Better yet, make it
> > > > configurable from JMX.
> > > >
> > > > Tom Cook wrote:
> > > > >
> > > > > On Thu, 8 Mar 2001, jBoss CVS Development wrote:
> > > > >
> > > > > [snip]
> > > > > > Log:
> > > > > > add final clause to debug var
> > > > >
> > > > > Just out of interest - why? To make it faster? If you
> > > > don't make this
> > > > > final then you could potentially change it from the JMX web
> > > > admin thingy,
> > > > > which would be useful in debugging/support...
> > > > >
> > >
> > >
> > > Oops - looks like a controversial one...
> > >
> > >
> > > I was assuming that since the debug output was controlled by variable
> and
> > > not just by the jboss.conf logging options that it was very much a
> > > development only kind of output. And thus we would want to optimise it
> > out
> > > at the earliest opportunity in a prod environment - to make that CMP
> stuff
> > > really fly ;-)
> > >
> > > Maybe I was wrong.
> > >
> > > I can see that changing it at runtime has some advantages - and I didn't
> > > 'twig' that I could just set the flag from my code too - which might be
> > > useful...
> > >
> > > Please vote and I'll amend it accordingly.
> > >
> > > Chris
> > >
> > >
> >
> ============================================================================
> > ====================
> > > This electronic message (email) and any attachments to it are subject to
> > copyright and are sent for the personal attention of the addressee.
> Although
> > you may be the named recipient, it may become apparent that this email and
> > its contents are not intended for you and an addressing error has been
> made.
> > This email may include information that is legally privileged and exempt
> > from disclosure. If you have received this email in error, please advise
> us
> > immediately and delete this email and any attachments from your computer
> > system.Rabobank International is the trading name of Coöperatieve Centrale
> > Raiffeisen-Boerenleenbank B.A. which is incorporated in the Netherlands.
> > Registered with the Registrar of Companies for England & Wales No.
> BR002630
> > and regulated by the SFA for the conduct of investment business in the UK.
> > >
> > > The presence of this footnote also confirms that this email has been
> > automatically checked by Rabobank International for the presence of
> computer
> > viruses prior to it being sent, however, no guarantee is given or implied
> > that this email is virus free upon delivery.
> > >
> > >
> > >
> > >
>
> --
> Dan Christopherson (danch)
> nVisia Technical Architect (www.nvisia.com)
>
> Opinions expressed are mine and do not neccessarily reflect any
> position or opinion of nVISIA.
>
> ---------------------------------------------------------------------------
> If you're a capitalist and you have the best goods and they're
> free, you don't have to proselytize, you just have to wait.
> -Eben Moglen
--
Dan Christopherson (danch)
nVisia Technical Architect (www.nvisia.com)
Opinions expressed are mine and do not neccessarily reflect any
position or opinion of nVISIA.
---------------------------------------------------------------------------
If you're a capitalist and you have the best goods and they're
free, you don't have to proselytize, you just have to wait.
-Eben Moglen
