OK, let's step back a sec. JRun does call exit() from its main method, OK?
And that's the only place. You seem to have implied that JRun calls exit()
whenever it gets an exception, which is not the case. The original
complaint was that a user wanted to be able to shutdown an embedded JRun
without JRun shutting down the JVM. This is do-able, but it will require
that the main method in the JRun class be changed so that JRun can be
stopped and clean up all its resources without calling System.exit().
Anyway, I logged this as a bug on Friday, so someone will take a look at it.
Scott Stirling
Allaire Corporation
http://www.allaire.com/developer/jrunreferencedesk/
> -----Original Message-----
> From: Paul Fink [mailto:[EMAIL PROTECTED]]
> Sent: Monday, August 28, 2000 10:27 AM
> To: [EMAIL PROTECTED]
> Subject: Re: Another Example of the problem with JRun3.0
> exit() behavior
>
>
> Look my point was that if this has to be explained to developers then
> I don't want to buy Allaire!
>
> OK, OK, I'll calm down.
>
> If I'm embedding the server its because I want to do
> something special with
> it such
> as a special control console, special logging, or may even
> other threads
> handling network connections. So I have lots of threads doing
> things that as
> a vendor you don't know about. Now JRun gets an exception. I want to
> do something like send a notice to operations, gently inform
> the users,
> log the error, or maybe complete my network connections, log
> my state and
> exit()
> Or maybe I want to write a fault tolerant wrapper that will
> let me detect
> the exception
> and restart the server. By calling exit() you don't let me do
> any of these
> things.
>
>
>
>
> ----- Original Message -----
> From: "Scott Stirling" <[EMAIL PROTECTED]>
> To: <[EMAIL PROTECTED]>
> Sent: Monday, August 28, 2000 9:18 AM
> Subject: RE: Another Example of the problem with JRun3.0
> exit() behavior
>
>
> > What's the actual impact in a real world situation? I am
> (still) just
> > looking for some practical example(s), not academic
> "shoulds." It will
> only
> > help me when I take this to the engineer who's going to
> have to fix it if
> I
> > can show him at least one reason why this is a real problem.
> >
> > Scott Stirling
> > Allaire Corporation
> > http://www.allaire.com/developer/jrunreferencedesk/
> >
> > > -----Original Message-----
> > > From: Paul Fink [mailto:[EMAIL PROTECTED]]
> > > Sent: Monday, August 28, 2000 10:02 AM
> > > To: [EMAIL PROTECTED]
> > > Subject: Re: Another Example of the problem with JRun3.0
> > > exit() behavior
> > >
> > >
> > > Scott,
> > > Java methods should never call "exit()" except in the
> > > controlling thread,
> > > i.e. the main.
> > > This is a good design pattern for single threaded C programs
> > > and a necessity
> > > for multi-threaded
> > > Java programs. Method are suppose to throw exceptions that
> > > can be caught and
> > > handle in the
> > > "proper" manner. You could ask "What other threads may be
> > > running and how do
> > > they want to
> > > deal with the error?" After what is the point of making
> the server
> > > embeddable?
> > >
> > >
> > > This is Java 101 stuff and it should be a "oh $#!+" and fixed
> > > with out any
> > > discussion.
> > >
> > > ----- Original Message -----
> > > From: "Scott Stirling" <[EMAIL PROTECTED]>
> > > To: <[EMAIL PROTECTED]>
> > > Sent: Friday, August 25, 2000 3:26 PM
> > > Subject: RE: Another Example of the problem with JRun3.0
> > > exit() behavior
> > >
> > >
> > > > I'm willing to log this as a bug. But can you tell me
> why this is a
> > > problem
> > > > that can't be worked around by other means? Seriously,
> I think your
> > > request
> > > > is reasonable, but on the other hand it'd help to have a
> > > good reason why
> > > > you'd want to stop an embedded instance of JRun without
> > > exiting the JVM.
> > > >
> > > > Scott Stirling
> > > > Allaire Corporation
> > > > http://www.allaire.com/developer/jrunreferencedesk/
> > > >
> > > > > -----Original Message-----
> > > > > From: Hammon, Randy [mailto:[EMAIL PROTECTED]]
> > > > > Sent: Friday, August 25, 2000 2:37 PM
> > > > > To: '[EMAIL PROTECTED]'
> > > > > Subject: Another Example of the problem with JRun3.0
> > > exit() behavior
> > > > >
> > > > >
> > > > > Last month I discovered a very uncool behavior of JRun3.0, in
> > > > > which it calls
> > > > > exit() when
> > > > > it stops. This is a huge problem for OEMS trying to
> > > embedding JRun3.0.
> > > > >
> > > > > I just ran into another example of why it's a bad idea for
> > > > > JRun3.0 to call
> > > > > exit().
> > > > >
> > > > > An internal customer had accidently started an instance
> > > of JRun, then
> > > > > proceeded to
> > > > > boot our application which internally starts JRun. After
> > > taking a few
> > > > > minutes to boot,
> > > > > our app started JRun, JRun found it's listening ports were
> > > > > being used, then
> > > > > it called
> > > > > exit and brought down OUR app along with it. Ouch!
> > > > >
> > > > > Allaire. Please fix this in the next release.
> > > > >
> > > > > Thanks,
> > > > > -Randy Hammon
> > > > > Evolve Software, Inc
> > > > > 1400 65th St. Suite 100
> > > > > Emeryville, CA 94608
------------------------------------------------------------------------------
Archives: http://www.egroups.com/group/jrun-interest/
Unsubscribe: http://www.houseoffusion.com/index.cfm?sidebar=lists&body=lists/jrun_talk
or send a message to [EMAIL PROTECTED] with 'unsubscribe' in the
body.
